A novel development of optimized hybrid MPPT controller for fuel cell systems with high voltage transformation ratio DC–DC converter

Abstract

The world is moving towards the utilization of hydrogen vehicle technology because its advantages are uniformity in power production, more efficiency, and high durability when compared to fossil fuels. So, in this work, the Proton Exchange Membrane Fuel Stack (PEMFS) device is selected for producing the energy for the hydrogen vehicle. The merits of this fuel technology are the possibility of operating less source temperature, and more suitability for stationery and transportation applications. Also, it provides a high amount of power density for heavy-duty electric vehicle applications. However, the major issue of the fuel stack technology is excessive current generation. Here, in the first objective, a Single Switch Wide Voltage Supply Converter (SSWVSC) is proposed to optimize the current levels of the fuel device thereby reducing the energy conduction losses of the entire system. In the 2nd objective, the duty cycle generation for the converter and handling of nonlinear energy generation of the fuel device has been done by introducing the Greywolf Optimization-dependent Adaptive neuro-fuzzy inference system (ANFIS). The features of this hybridization concept are less iteration number needed, less disturbance in MPP position, low stabilizing time of the fuel module production voltage, and more reliability. Here, the fuel module interfaced DC-DC circuit is studied by utilizing the MATLAB software and the introduced converter is tested only with programmable DC-Source.

Introduction

From the literature study, renewable energy technologies are more suitable for hydrogen vehicle applications because of their huge availability, less noise production, and better suitability for large and heavy-duty transportation systems. Based on their existence and availability, renewable voltage sources are differentiated as windmills, fuel cell methodology, sunlight systems, biomass technology, plus hydropower systems¹. In the article², the scholars investigated the various types of hydropower stations which are low-level water-head hydro stations, moderate level water head-based hydro energy systems, and excessive water-head hydro energy stations. Among all of that, the moderate-level water head stations are popularly used in the hill area stations³. The advantages of hydro energy power stations are they clean sources, have no need for fossil fuels requirement, are more reliable for domestic applications, reduce flood risks, and good dispatchable. However, the drawbacks of these systems are limited availability, and higher starting and ecological costs. Also, the world is running out of sites and expensive up-front costs⁴. The demerits of hydrological stations are compensated by applying biomass technology. The direct combustion methodology is utilized in biomass stations for obtaining highly efficient electrical power. Here, the biomass is burned completely in the boiler chamber to obtain the steam which is forwarded to the biomass turbine chamber for operating the electrical generator⁵.

The features of biomass technology are carbon neutrality, less dependence on petroleum products, more useful for all domestic applications, low energy generation cost, less waste production, and easy availability⁶. However, the disadvantages of this biomass technology are the less clean source, less economically efficient, creates the deforestation effect, requires a huge amount of space, low energy density, requires huge water, and requires high-level management. So, the power corporation industries are working on windmills to capture the kinetic energy of the air⁷. Here, the atmospheric air turns the propeller of windmills and the propeller helps the electric generator to produce the electrical voltage. The merits of wind systems are cost cost-effective source, clean, and renewable power sources. Also, it is best from an economical point of view and less installation space is needed. In addition to that, this source may not create any air and water pollution. However, the demerits of wind mills are high intermittent, more noise production, high effect on wildlife animals, more risk in maintenance, and recyclability⁸. In the article⁹, sunlight technology limits the disadvantages of wind energy production systems. Sunlight is very natural and more available in the atmosphere. From the literature investigation, each sunlight cell’s available potential is 0.97 V to 0.98 V, and it has very little value for local domestic consumer applications¹⁰.

So, the multiple potential value sunlight cells are integrated to enhance the entire sunlight system power production efficiency. Here, the working nature of any sunlight cell is exactly equal to the diode performance¹¹. The manufacturing of the sunlight cells has been done by selecting various materials which are amorphous silicon, cadmium telluride, gallium arsenide, and powder-coated aluminum. In these materials, the cadmium telluride material is utilized in most of the sunlight power production industry for developing sunlight cells because of the high thermal conductivity and more sunlight energy conversion efficiency¹². However, the sunlight technology drawbacks are more expensive, capturing more amount of space, unable to deliver energy at cloudy and night times. These drawbacks of sunlight technologies are limited by introducing the fuel stack technology in the automotive industry. From the literature study, the fuel cells are differentiated based on the electrolyte materials utilized which are Phosphoric Acid Electrolyte Material Fuel Module (PAMFM), Alkaline Electrolyte Material Fuel Module (AEMFM), Solid Oxide Electrolyte Material Fuel Module (SOEFM), Molten Carbonate based Electrolyte Fuel Module (MCEFM), and Direct Methanol Electrolyte based Fuel Cell (DMEFC)¹³.

In the article¹⁴, the scholars referred to the phosphoric electrolyte-involved fuel module for stationary 400 kW power production application. In this cell, the hydrogen and pure air chemical reactions happened at 150 °C to 220 °C to obtain the free electrons in the fuel module. Here, phosphoric acid is one type of corrosive acid and it gives 3-types of salts which are named dibasic phosphates, primarily available phosphates, plus tribasic phosphates¹⁵. Also, it is soluble in H₂O content. As a result, it is highly compatible with the resilient caustics. This fuel cell produces heat and electricity at a time. So, it is applied to the military applications. Also, this H₃PO₄ cell shape is very simple and has less electrolyte volatility. However, the disadvantages of this cell are less power density of electrolytes and chemically aggressive electrolytes. Also, it directly affects the human lounges through the inhalation of mist. This serious issue of phosphoric acid is limited by utilizing the Proton Exchange Membrane Fuel Cell (PEMFC). In this work, the polymer membrane electrolyte-based fuel cell was selected for lightweight automotive systems because its features are low-value operational temperature, compact in size, quick functioning state, more conversion efficiency, and lightweight. The simulative design of the polymer membrane fuel module is mentioned in Fig. 1 and the demand of the fuel cell utilization is mentioned in Fig. 2.

Fig. 1.

Schematic representation of PEMFC-based GWO-ANFIS power point tracking controller.

Fig. 2.

Utility of fuel modules for every year all over the world¹⁶.

From Fig. 1, the polymer electrolyte fuel module produces completely fluctuated nonlinear voltage. So, the obtaining of peak power from the fuel module is a highly complex task. From the literature discussion, there are different artificial intelligence, and nature-inspired methodologies are used for the extraction of the peak power from the fuel module. Along with the artificial intelligence controllers, there are a few more conventional MPPT controllers available in the literature which are Kalman filter, Perturb and Observe, sliding controller, hill climb, and incremental conductance controllers. In¹⁷, scholars referred to the Kalman filter technology as utilized to optimize the fluctuations of the fuel stack generated currents under various water membrane conditions. The design of this Kalman filter technology is very easy and needs fewer mathematical formulas for accurate tracking of the MPP of the proposed system. However, the drawbacks of this Kalman controller are more system integration costs and a high-level intelligence system needed for monitoring the entire system¹⁸. To limit this problem, the slider technology is implemented in the article¹⁹ for continuous identification of the peak power position of the fuel system at different hydrogen decomposition conditions. The slider collects the boundary conditions of the utilized converter and fuel system for reducing the overall system steady-state error and optimizing the overall system size²⁰.

The disadvantages of the slider concept are instability issues when the system complexity is more and less accurate for complete nonlinear systems. The slider limitations are compensated by using the artificial neural network block in the fuel stack integrated automotive systems²¹. The artificial neural concept is developed from the human brain’s functioning nature. In this, each neuron is identified as one node, and each node exchanges its associated information with the other neural nodes to find the required object of the solid oxide electrolyte-dependent electric vehicle system. The advantages of neural networks are less statistical training required, and the ability to capture the nonlinear relation between the independent and dependent variables²². However, the drawbacks of the above methodologies are overcome by using the hybrid MPPT controller. The second objective of this article is the development of a GWO-based ANFIS technique for tracking the actual functioning point of the proposed fuel module. Here, the hybridization of the nature optimization algorithm and ANFIS are utilized for limiting the tracking time of MPP. Due to this hybridization, the introduced algorithm running time, and oscillations of the converter output voltage are reduced excessively. The only challenge involved in this algorithm is more difficulty in understanding.

Another problem with the fuel system is more current generation which is five times more than its output voltage²³. So, there are various types of power DC–DC converters are selected in the literature for the optimization of the fuel system-generated current. In the literature study, the converters are differentiated by considering the isolation and non-isolation technology. In Ref.²⁴, scholars referred to the flyback isolated topology for the alkaline fuel module to supply power to emergency household applications. The features of the flyback converter circuit are capable of supplying multiple load voltages and provide good isolation between the power supply system and domestic loads to protect consumer electronic products. Also, it provides a wide range of supply voltages²⁵. However, the drawbacks of this flyback converter are the high amount of output current ripples, more noise-included converter power, and excessive complexity of feedback control circuits. The forward circuit-based DC–DC converter is applied to the molten carbonate electrolyte fuel cell system to limit the issues of the flyback converter circuit. This forward converter circuit helps to improve and decrease the source voltage of the fuel module under various oxygen decompositions, and water membrane conditions²⁶. The features of a forward converter are higher energy conversion efficiency, floating output power, switched mode power supply, good power adaptability, and good reliability.

The drawbacks of this forward converter are increased implementation cost due to the utilization of additional rectifiers, and filter circuits²⁷. The half and full bridge-oriented power DC–DC converters are used for the wide voltage conversion ratio of the fuel stack and their features are minimal fluctuations of converter generated voltage, efficient energy transmission in the distribution network, supplied current is equally transferred to the load, and better dynamic response. The drawbacks of the bridge model DC–DC converter are more circulating inrush currents, more costly switches needed for doing the rectification, and the possibility of distorted voltage generation²⁸. So, the non-isolated model DC–DC circuits are reviewed in the article²⁹ for heavy-duty electric vehicle applications. In this work, a single switch good fuel stack source voltage conversion ratio wide voltage supply DC–DC converter is introduced for improving the source power conversion efficiency of the proposed system. The features of this converter are less passive components utilization when associated with the interleaved boost DC–DC circuit, widely utilized for automotive heavy-duty electric vehicles, low-level voltage across the switching devices, and more reliability for all operating fuel module temperature conditions. The introduced converter circuit is illustrated in Fig. 1. The remaining part of the article is organized as the challenges and outcomes of the hybrid power point identifying controllers and its implementation are mentioned in “Challenges and outcomes of the hybrid MPPT controllers” and “Performance analysis of PEM fuel cell module” sections. The development of the single switch circuit and its simulative analysis are described in “Implementation and analysis of various MPPT controllers” and “Development of a new single switch converter circuit” sections. Finally, the comparative analysis and the outcome of the article are defined in “Simulative validation of the SSWVSC-FED PEMFC system” and “Experimental evaluation of SSWVSC” sections.

Challenges and outcomes of the hybrid MPPT controllers

All the fuel cell methodologies provide distorted nonlinear current characteristics. So, the delivery of the fuel module peak voltage is a very difficult task. So, the MPPT blocks are interconnected with the converter-fed polymer membrane electrolyte-based fuel module system. From the literature study, the P&O and other conventional controllers may not provide the accurate MPP location of the fuel module because the drawbacks are high distortions across the functioning point of the fuel module, and less reliability³⁵. The fractional current-oriented MPPT is included with the fuzzy logic block for reducing the tracing time of the fuel module MPP. The drawback of this controller is the selection of proportionality variables which are identified by using the linear approximation method. As a result, the working efficiency of this hybrid MPPT method is reduced. The slider-optimized fuzzy methodology is applied to the fuel module three-phase dual-active DC-DC bridge converter for four-wheeler electric vehicle applications³⁶. Here, the four switches’ duty signal is controlled by applying the steady state variables of the three-phase converter thereby optimizing the energy distribution losses of the overall fuel cell integrated microgrid system.

An improved beta-dependent fuzzy controller is developed in the article³⁷ for a polymer electrolyte-based fuel module system for obtaining the switching signals of the quadratic high voltage gain converter DC–DC circuit. Here, the DC–DC conventional circuit is interconnected with the quadratic circuit to increase the power quality of the renewable energy system. The features of this topology are easy operation, good understanding, more reliability, and good dynamic performance. However, this circuit takes more design and development costs³⁸. The drawbacks of this converter are limited by using the particle swarm optimization-artificial neural controller. Here, the neuron training and its weight updating are done by applying the swarm intelligence method. In this method, the multilayer neural network takes the fuel stack from the locally available maximum power point place to the globally available peak power point position³⁹. After reaching the peak power position of the fuel module, the PSO is utilized to limit the distortions of the interleaved converter power. The merits of this technique are less value of iterations are needed and fast-tracking of the operating point of the fuel module. However, this type of hybridization needs more cost for development and utilizes a greater number of sensors for sensing the fuel stack hydrogen and its associated oxidization reaction⁴⁰. The recent available hybridization-based MPPT technologies outcomes, challenges involved in functioning, and their merits are discussed in Table 1 and the associated types of fuel modules analysis are given in Table 2.

Table 1.

Recently available hybridization involved MPPT methodologies and their outcomes and challenges.

Authors of paper	Year published	Collected variables	Methodology	Monitored parameter	Integrated converter	Challenges, outcomes, merits, and demerits
Bouguerra et al.30	2024	H2, H2O, and O2	FSO-CSA	Duty Singal, fuel current, and voltage	Buck, quadratic converter	The scholars applied the flying squirrel search dependent cuckoo search concept for the z-source quadratic converter to observe the switching pulse generation for the fuel stack system under different hydrogen decomposition conditions. The challenges involved in this methodology are sensor handling and cost optimization. The merits of this controller are the ability to adapt to various water membrane conditions and better dynamic response.
Elymany et al.31	2024	Water membrane, H2, H2O	ZOA-ANFIS Network	Duty signal, fuel voltage, and power	Buck, and boost	The Zebra optimization method is included in this ANFIS block for precision identification of the functioning point of the polymer fuel module. Here, the zebra optimization makes it run the overall system towards the global power-point position. As a result, the system’s working efficiency is improved extended level. The ANFIS structure helps the zebra optimization for accurate MPP finding without distortions. The challenges involved in this method are data interpretation of ANFIS and its associated weight adjustments.
Reddy et al.32	2024	Polymer electrolyte, H2, CO2	ANFIS-GSO	Fuel production power and Duty signal	High gain converter	This paper explains the Glowwarm swarm with ANFIS controller which provides a fast convergence speed for the fuel module at the quick change of system operating temperature conditions. Here, the challenge facing the GSO is high iteration count is required. So, the entire fuel module system faces the load current distortions thereby heat is generated in the system. The features of this algorithm are a few sensor utilization, optimal size, and battery robustness.
Bhavani et al.33	2024	Water membrane, H2, H2O	CS-GWO	Duty signal, fuel voltage, and power	Boost Z-source converter	In this work, the scholars worked out the cuckoo search adaption-based grey wolf controller for optimizing the nonlinear performance of the polymer membrane electrolyte system. The goodness of this algorithm is easily understandable and more efficient for different oxygen decomposition and chemical rates of the fuel module. The major challenge of this algorithm is more operational complexity and heavy size.
Kiran et al.34	2024	Fuel cell H2, H2O and Voltage	ANFIS-PSO hybrid methodology for MPP identification	Duty signal, MPP finding	Quasi Z-source DC-DC converter	In this work, adaptive neuro membership values are utilized for training the ANFIS weights and ANFIS collects the PSO-identified local fuel module maximum power points for extracting the peak voltage from the fuel module. The drawbacks of this concept are development complexity, and the need for high current rating sensors for reading the fuel stack current values.

Table 2.

Comparison of types of fuel cells under different electrolyte utilization conditions.

Type of cells	Temperature (°C)	Electrolyte	Oxidant	Fuel	Efficiency (%)
Alkaline	60–90 °C	Potassium hydroxide	Oxygen	Pure hydrogen	50–55
Molten carbonate	600–700 °C	Molten salt	Oxygen	CO/Hydrogen	45–55
Solid oxide	500–1000 °C	Zirconia -based Ceramic	Oxygen	Natural gas	45–60
Direct methanol	60–200 °C	Polymer	Oxygen	Methyl alcohol	40–55
H3PO4	150–200 °C	Phosphoric acid	Oxygen	Hydrocarbons	40–55
Sulfuric acid	80–90 °C	Sulfuric acid	Oxygen	Impure hydrogen	40–56
Proton-exchange membrane	50–80 °C	Polymer with Proton-exchange membrane	Oxygen	Hydrogen, methanol	40–55
Protonic ceramic	600–700 °C	Thin membrane of Barium Cerium Oxide	Oxygen	Hydrocarbons	45–60

Performance analysis of PEM fuel cell module

Fuel cells are electrochemical cells that use the chemical energy generated by fuel to produce electricity. Electricity is produced when fuel and oxygen are supplied. Fuel cells are the same as batteries to some extent but do not require recharging. They provide high efficiency, low pollution, and low noise. This cell consists of two plates: a negative plate, called the anode, and a positive plate, called the cathode, slotted in by an electrolyte. They are classified into different types based on the type of electrolyte used. Solid oxide fuel cells (SOFC) oxidize fuel to produce electricity, and the electrolyte used in these cells is solid oxide or ceramic⁴¹. The electrolyte should be capable of conducting oxides from the cathode to anode plate. In this case, the fuel is hydrogen or carbon monoxide, and the oxidant is oxygen, where the operating temperature is 500–1000 °C. The key components of SOFC are the anode, cathode, and electrolyte. The anode is typically nickel-based and is used for fuel oxidation, where hydrogen reacts with oxygen ions to produce water and electrons to provide electricity. The solid oxide cathode is made of ceramic materials, such as lanthanum strontium manganite⁴². Oxygen ions are formed when oxygen is released from the air and then move from the electrolyte to the anode. The electrolyte in SOFCs is yttria-stabilized zirconia, which allows oxygen ions to flow from the cathode to the anode while preventing the flow of electrons through it. It plays a major role in the separation of oxidation and reduction reactions⁴³.

SOFCs produce low emissions, offer high efficiency, and can operate on various fuel types. The disadvantage is that as they operate at high temperatures, materials are degraded, which increases the maintenance cost and decreases the lifespan of the cell. SOFCs are utilized for a wide area of applications, such as providing electricity to industrial and residential utilities, and they are used as auxiliary power units in vehicles and aircraft. However, the drawbacks of solid oxide cells are more temperature limits, and large⁴⁴. So, the PEMFC is selected in this work for giving the electrical supply to the automotive systems. Here, the Proton Exchange membrane fuel cells (PEMFCs) transform a mixture of H and oxygen into electrical energy, releasing water and heat as byproducts. The core components are the anode, cathode, and solid polymer electrolyte, which conduct protons and act as electron insulators. Hydrogen gas is supplied to the anode side, where the platinum catalyst splits into protons and electrons (anode reaction). Protons from the anode side pass to the cathode side through a proton exchange membrane, called proton movement. Electrons were forced to flow through an external circuit to reach the cathode, resulting in an electric current being generated. Cathode reactions occur when oxygen gas is supplied to and at the cathode, all of which react to form water, and the reaction is catalyzed by platinum. The byproducts of the reaction are water and heat. PEMFCs are used in portable power systems, fuel cell vehicles, and stationary power generation systems. Table 3 provides the development of a polymer membrane fuel module under a quick variation of the operational temperature conditions. The utilized polymer membrane fuel cell parameters are collected from the MATLAB Simulink library. The structure of the fuel module is mentioned in Fig. 3a and its associated equivalent circuit is given in Fig. 3b.

Table 3.

Utilized fuel module values for the investigation of the converter.

Variable	Values
Fuel module functioning cells (N)	65
Fuel module functioning H2 pressure	1.8976 bar
Fuel module functioning O2 pressure	1.2032 bar
Inside fuel module air operating value (Ipm)	505.321
Inside fuel module operating gas value (R)	86.2453 [J mol−1 K−1]
Applied Faraday value for the working polymer fuel module	94,3245.108 [C mol−1]
Obtained polymer fuel module voltage at peak demand condition	65.12 V
Obtained polymer fuel module power at peak demand condition	5.9999 kW
Obtained polymer fuel module current at peak demand condition	134 A
Evaluated openly available fuel module voltage	65.213 V
Utilized oxygen content for the polymer fuel module operation	62.8769%
Utilized hydrogen content for the polymer fuel module operation	99.2185%
The entire oxidation value for the polymer fuel system operation	99.1156%
Overall oxygen decomposition happened in the fuel system	22.89712%

Fig. 3.

(a) Utilized polymer fuel module, and (b) electrical circuit of the fuel module.

1

2

3

4

5

6

7

8

From the above Eqs. (1) and (2), the terms H₂, F, O₂, , H₂O, and is represented as hydrogen at the anode block, faraday constant of the water membrane, oxygen at the anode block, reversible open-circuit voltage, water at the cathode block, and variation in the Gibbs free energy of formation (KJ/mol). From Eq. (3), the terms R, _,, ,and is represented as universal gas constant (8.314 KJ/Kmol k), stack temperature, partial pressure of hydrogen, partial pressure of oxygen, partial pressure of water, and entropy variation (KJ/Kmol k). Here, from Eq. (4), the terms ,and total fuel cell current, external current, and current loss are represented. Finally, from Eq. (5), the terms is identified as activation fuel stack voltage loss, and ,,, and are the constant parameters. Finally, the variable is the dissolved oxygen concentration in mol/cm³. The partially available pressure of oxygen content is identified as from Eq. (6). The terms_,,,, and corresponds to the equivalent activation resistance, number of cells connected in series, ohmic voltage losses, resistance to the flow of ions in the membrane (Ω), electronic resistance (Ω), and contact resistance of the electrodes (Ω) from Eqs. (7) and (8).

9

10

11

12

13

14

15

16

17

18

19

20

From Eq. (9) the terms Tm, rm, A, and represents the thickness of the membrane (cm²), the membrane’s resistance, the cell’s active area (cm²), and the average water content of the membrane. From Eqs. (10) and (11) the terms a, , α, R, T, n, F, I_L, and I are represented as a function of water activity, concentration voltage losses, transfer coefficient, universal gas constant, temperature, number of electrons involved in the reaction, faraday constant, limiting current, and operating current. From Eqs. (13) and (14), andis represented as fuel cell stack voltage as a function of the current and output power of the stack and based on the Eqs. (15), (16), and (17), the terms ,,are represented as the partial pressure of oxygen, the molar function of oxygen, atmospheric pressure, pressure of hydrogen, pressure of hydrogen at the inlet, pressure difference, and squared hydrogen flow rate. The fuel cell generated voltage curve, and current curves are mentioned in Fig. 4a, b.

Fig. 4.

(a) Generation of the polymer membrane voltage curve. (b) Generation of the polymer membrane current curve.

Implementation and analysis of various MPPT controllers

For all renewable energy systems, the major issue is nonlinearity in energy generation. Similarly, in the fuel module, the energy generation happened in a nonlinear way. So, more energy extracted from the fuel module is not possible⁴⁵. From the literature study, the researchers developed various types of power point tracking controllers which are soft computing, neural computing, and swarm intelligence controllers, and these techniques help the fuel cell to catch the peak voltage point on the power curve of the source fuel system thereby producing the peak energy to the industrial applications. In this work, the grey wolf algorithm-ANFIS technology is developed for the continuous change of the fuel module’s working temperature. The features of this introduced algorithm are fast identification of the source MPP, less settling time value of MPP, more power transmission efficiency, better robustness, and more adaptability for quick variation of the fuel system functioning temperature. This introduced algorithm is investigated along with a few more techniques which are Modified P&O, Improved IC, Radial Basis Functional neural controller, and swarm intelligence optimized P&O controller.

Modified P&O MPP tracking for a source fuel module

From the renewable literature analysis, it has been noted that the P&O is studied for all non-conventional energy systems because of its merits are very simple to understand, easy to develop the algorithm, more efficient for identifying the local power point positions, more suitable for polymer fuel module-based traffic monitoring system, and more adaption towards under various fuel module water membrane conditions⁴⁶. However, the disadvantages of this controller are more disturbances in source energy production and heavy heat generation losses. So, the modified step P&O technology is implemented in the article⁴⁷ for optimizing the oscillations of the utilized renewable power system. The basic concept of using this technique is about varying the primary fuel stack input voltage (that means either to increase or decrease) or altering the duty cycle of the DC–DC Converter and measuring the amount of variation of the output power until. This is named the P&O method since its effectiveness drastically depends upon the power rise curve over the voltage under the most potent point. Here the newly measured power was then compared with the pre-perturbation measured power. With the increase in voltage, the power at the left side of the MPP was raised while the voltage at the right side of the MPP was reduced⁴⁸. The location of the MPP is identified by using the Eqs. (21) to (23). The step variation of this selected P&O technology is mentioned in Eqs. (24) and (25).

21

22

23

24

25

where dP is the change in power and dV is the change in voltage. This process is repeated continuously to maintain the system operating at or near the MPP. From Eq. (25), the variable ʎ denotes the duty signal and the detailed working structure is mentioned in Fig. 5. From Fig. 5, the term ϒ represents the step length on the P-I curve of the fuel module which is between 0.4 and 0.6.

Fig. 5.

Step change based on the P&O power point-catching technique.

Improved IC MPP tracking for a source fuel module

In the article⁴⁹, the authors recommended the incremental conductance controller because its merits are optimal disturbances across the functioning point of the selected fuel module system. Also, it provides fewer disturbances of interleaved DC–DC converter load voltage and helps the fuel stack provide a safe operating region of the battery state of charge delivery. In the article⁵⁰, the Incremental Conductance (IC) was designed as an MPPT control algorithm for the fuel cells to optimize power extraction. The technique compares the increase in the current (dI) to the rise in the system’s voltage (dV). The objective is to track the MPP by finding the slope of the power-voltage (P-V) characteristic and adjusting the operating point in this area of the MPP. At the point of maximum power point (MPP) of the power-voltage curve, its slope makes zero, which means dP/dV = 0. Using the relationship dP/dV = I+(V*dI/dV), which simplifies to, dI/dV=− I/V.

26

27

Before the MPP, where power increases with increasing voltage (dP/dV > 0), and the incremental conductance is greater than the instantaneous conductance (dI/dV>− I/V). After the MPP, where power decreases with increasing voltage (dP/dV < 0) and the incremental conductance is less than the instantaneous conductance (dI/dV<− I/V). The advantages of IC provide better accuracy for tracking the MPP under varying fuel supply and load conditions than the basic MPPT algorithms such as P&O and it keeps on fluctuating just around the MPP to provide steady power flow. Also, it is commonly used in fuel cell vehicles, portable fuel cell systems, and grid-connected fuel cell systems. The operational structure of the modified IC is mentioned in Fig. 6. From Fig. 6, the duty of the converter is varied based on the MPP location on the P-V curve of the fuel module which is given in Eqs. (26) to (28).

Fig. 6.

Improved IC for the proposed polymer membrane fuel stack.

Radial basis functional neural model for a source fuel module

ANNs are computer models, which replicate the particular structure and function of the human brain in computing processes. They are composed of multiple successive layers of artificial neurons that facilitate the computation and learning of patterns in data. Neurons receive inputs, collect them through the weighted sum, and apply an activation function to give results. Some of the significant layers of ANNs include the input layer, hidden layer, and output layer⁵¹. The input layer accepts several different data types as input, often given by the programmer. The hidden layer is the layer that exists somewhere in between the input and the output layers. It does all the computations for detecting hidden features or even patterns in a certain data set. Some information transformations are achieved through this layer that is hidden and the information that is produced is transmitted through this layer of the neural network. The artificial neural network receives input, adds a weighted sum of the inputs, and possesses a bias factor as well.

This computation can be expressed in the terminology of a so-called transfer function. Parallel processing, the inability to store data in the entire network, the ability to process with less or incomplete knowledge, the ability to have memory distribution, and the capability of fault tolerance are strengths of this network⁵². However, the disadvantages of the neural network include that it is not easy to identify how the network behaves, the network heavily relies on the hardware, it is not easy to explain an issue to the network, and there is no guarantee of proper formation of the network. ANNs are popular intelligent technologies such as image recognition, natural language processing and robotics, and self-driving cars. In the article⁵³, the scholars developed the RBFN model for catching the peak power point of the fuel stack module as shown in Fig. 7.

28

29

30

31

32

33

34

Fig. 7.

RBFN neural MPPT controller for the fuel module system.

where P, L, and Y are the neuron layers. From Eq. (33) and Eq. (34), g, W, and k are defined as node number, and its weights.

PSO-P&O hybrid technique for a source fuel module

Particle Swarm Optimization (PSO) is a flexible approach to deal with the non-linear characteristics of fuel modules. Let employ no of particles that form a swarm these particles navigate in the search space to find the best solution, and each particle keeps a record of the best point observed so far equivalent to maximum fitness value, this point is referred to as personal best (P_best). Another best value recorded by PSO is global best (G_best), which is obtained from the neighborhood of the particle. Every element in the PSO Algorithm increases its velocity & position based on three aspects the best experience it has had (P_best), based on neighborhood experience (G_best), and depending on the exploration of a particle for solving the solution Space⁵⁴.

In this algorithm, the first step is the initialization of agents which includes the maximum number of iterations, the size of the swarm, the dimension of the search space, PSO constants weight, cognition, and conspiracy coefficients (w, c₁, c₂), and two random numbers r₁ and r₂. Next, the current fitness value is evaluated and after this individual and global best values will be updated, then the velocity and position of each particle are updated and at last convergence is determined, when the convergence criterion is reached the iterative algorithm is terminated⁵⁵. Advantages of this algorithm include that it is easy to understand and implemented quickly. The convergence rate in each iteration is moderate while it is easily trapped in high dimensional space at local optimums is the demerit of this algorithm. However, the drawbacks of this algorithm are that low-quality solution faces difficulties in the starting state of convergence, and need more memory to update the weights.

In the article⁵⁶, the authors developed the swarm optimization-P&O block for finding the MPP position of the electric vehicle-based fuel stack. This algorithm swarm velocity and its associated weights are adjusted by applying Eqs. (35), and (36). The fuel module MPP tracking process is defined in Fig. 8. Here, the K_i, K_P, K_d values are selected for adjusting the step constant of this PSO hybrid technology and its utilized values are 0.02, 0.0262, and 0.074 respectively. The starting local MPPs are captured by considering swarm technology. Finally, conventional technology is considered for finding the accurate less distorted quasi-source DC-DC circuit power.

35

36

Fig. 8.

Systematic diagram of PSO-P&O MPPT controller.

Here, each particle is itself arbitrarily located in the s-dimensional space as a candidate solution. The velocity of ith particle V_s = (V_sl, V_s12…., v_id) is defined as the change of its position more explicitly. The flying direction of each particle is the dynamical interaction of individual and social flying experience which can also be defined as the metric of interpersonal differential affect. However, the challenges involved in this algorithm are moderate accuracy, more installation space needed, and less reliability.

Proposed ANFIS-GWA hybrid technique for a source fuel module

From the literature study, fuzzy can work for non-linearity problems, and to track the maximum power more accurately⁵⁷. Here, fuzzy consists of four forms of blocks namely, fuzzification, rule base, interference, and defuzzification. Fuzzification analyzes the input signals and provides them with fuzzy values. These input signals are clear inputs such as the change in the voltage reading. The rules block, like the linguistic rules, defines what control action should take place for a given certain set of input values. The inference form essentially holds the responsibility of drawing an inference about the collected data based on the applied control action. The defuzzification form is one of the forms in fuzzy logic calculation that is used to convert fuzzy information into non-fuzzy information that is suitable for the process to be controlled. The drawbacks of this fuzzy technology are more time-consuming for tunning, and improper membership function selection creates the oscillations issue across the MPP point. In this article, grey wolf algorithm-based ANFIS technology is developed for the fuel module power production system. Here, the GWO technique is more often applied to solving multi-objective optimization problems across different disciplines, such as engineering, artificial intelligence, and machine learning.

The key concepts of GWO are hierarchy and the hunting process. The hierarchy of grey wolves contains α, β, δ, and here, alpha (α) is the first rank in a hunt, it determines where the pack will hunt and participates in the decision-making process. For GWO, α is the best in the candidate solutions and β is the second in command, helping the alpha and it is the second-best solution possible. Delta (δ) is the third level, after α and β conventions. It represents the third-best solution, and omega (ω) is the rest of the pack. The hunting process encircles prey, hunting as α, β, and δ, and updates position according to ω wolves and their best solutions. When solutions are close to the optimal solutions the wolves move closer to the prey, hence minimizing the exploration factor as shown in Fig. 9(a). From Fig. 9(b), the hunting behavior is mathematically modeled using the movement of each wolf in the search space for α, β, and δ. The position of a wolf is updated according to the following equation,

37

38

39

Fig. 9.

(a) Developed PSEUDO code of ANFIS-based GWO controller for a polymer fuel module. (b) Utilized flowchart for proposed ANFIS-based grey wolf controller for fuel stack.

where is used to represent one position of the wolf at a particular iteration and refers to the positions that are occupied by the wolves. Here, at the beginning of the technique, the wolfs are randomly initiated and after searching the local peak power points the grey wolf shifts from its working condition to fuzzy block by selecting Eq. (38). From Eq. (39), the parameter γ is selected as a step adjustment value which is equal to 0.027. The major features of this controller are fast tracing speed of the functioning point of the MPP, less load voltage settling time, better reliability, fewer iterations for finding the better membership values, and higher operational efficiency when associated with the other controllers.

Development of a new single switch converter circuit

Renewable energy sources produce very low-level load voltages which are improved by selecting the high voltage transformation DC-DC converter. In the article⁵⁸, the scholar utilized the non-isolated technology to enhance the power transmission efficiency of the polymer fuel system. This one diode and one switch conversion circuit provides low development costs, low-level implementation complexity, and ease of design. However, the voltage utilization factor of this converter is low. To mitigate this problem, a new converter topology is developed in this work to improve the voltage rating of the fuel stack module. The structure of the converter circuit is mentioned in Fig. 10(a), and the 1st switching working condition is illustrated in Fig. 10b. Finally, the second switching strategy is shown in Fig. 10c. From Fig. 10a, the converter utilized the Metal Oxide Semiconductor Field Effect Transistor (MOSFET) Switch (T), five electrostatic elements which are C_q, C_j, C_g, C_s, C_m, and three electromagnetic elements L_c, L_b, and L_n. Finally, the selected diode switches and load-resistive elements are D_c, D_b, D_n, and R_load respectively. Here, the basic boost setup is formed by considering the elements L_c, T, C_j, and D_c, and the components L_b, C_q, L_c, C_j, C_s, D_n, and D_b (L²D²C³) help the entire converter circuit to enhance the load power profile of the system.

Fig. 10.

(a) Switching strategy-I, (b) switching strategy-II, and (c) switching strategy-III for polymer fuel cell system.

Switching strategy analysis-I for PEMFC

The proposed switching converter circuit operation is illustrated in Table 4. From Fig. 10a, the electromagnetic and electrostatic components currents, obtained voltages, and their charging and energy-delivering states are indicated as I_Lc, I_Lb, I_Ln, I_Lc−crig, I_Lb−crig, I_Ln−crig, I_Lc−drig, I_Lb−drig, I_Ln−drig, V_Lc, V_Lb, V_Ln, V_Lc−crig, V_Lb−crig, V_Ln−crig, V_Lc−drig, V_Lb−drig, V_Ln−drig, I_Cq, I_Cj, I_Cg, I_Cs, I_Cm, I_Cq−crig, I_Cj−crig, I_Cg−crig, I_Cs−crig, I_Cm−crig, I_Cq−drig, I_Cj−drig, I_Cg−drig, I_Cs−drig, I_Cm−drig, V_Cq−crig, V_Cj−crig, V_Cg−crig, V_Cs−crig, V_Cm−crig, V_Cq−drig, V_Cj−drig, V_Cg−drig, V_Cs−drig, and V_Cm−drig respectively. For convenient investigation, there few assumptions are made here for the investigation of the overall power converter which is the internal resistive property of capacitive and inductive components is zero, and the utilized power switching state in this converter is ideal. Finally, the large value electrostatic and electromagnetic elements are considered in this work for obtaining the uniform load voltage of the system. In this switching state, the DC-DC circuit moves into a Continuous Conduction State (CCS) when the gate signal of the MOSFET has high source voltage. Also, it works in a Discontinuous Conduction State (DCS) at variable gate signal amplitude. From Fig. 10a, the inductive currents, and capacitive voltages are derived by selecting the Kirchoofs voltage and current law.

40

41

Table 4.

Overall operational modes of the introduced direct current DC–DC circuit.

Devices	Strategy-1	Strategy-2	Strategy-3
T	On mode	Block state	Block state
Dc	Block state	On mode	Block state
Db	Block state	On mode	Block state
Dn	Block state	On mode	Block state

Switching strategy analysis-II for PEMFC

In this working state of the proposed DC-DC circuit, the supplied gate signal amplitude is moderate for working under CCS & DCS as mentioned in Fig. 10b. As for the first switching strategy, the switch works for uniform output voltage generation. In this stage, the switch moves into blocking mode, and the electrostatic elements’ stored energies try to switch on all three didoes. Here, the C_q, C_j, C_m deliver the complete energy to the consumer through the diodes D_c, D_b, and D_n. The obtained currents and voltages for the electrostatic elements are mentioned in Eqs. (42) and (43).

42

43

Switching strategy analysis-III for PEMFC

Here, the supplied voltage amplitude to the proposed converter is very low-level. As a result, the circuit moves completely in a blocking state. In this funning state, the addition of all electromagnetic inductive components currents is zero and its associated voltage appears across each inductor is zero.

44

45

Steady-state strategy of the proposed converter

The steady-state analysis of the introduced single switch converter has been done by focusing on the converter conduction state one and conduction state two. The voltage conversion value of the converter is evaluated by utilizing the Fig. 11a, b. Figure 11a, and Eq. (43) are considered for the derivation of the converter load voltage. From Eq. (46), the source electrostatic capacitive elements C_q, C_g, C_s voltages are equal and it is derived in Eq. (26). Similarly, the voltage stress appeared at each passive component is mentioned in Eq. (50), and the voltage balanced equation is applied to the all-electromagnetic elements for obtaining the charging currents.

46

47

48

49

50

51

52

53

54

55

56

57

58

59

60

61

62

63

64

65

66

67

68

Fig. 11.

(a) Introduced single switch converter-I and II, (b) strategy-III.

Discontinuous analysis of single switch proposed converter

For the investigation of the utilized converter circuit at a discontinuous state, there are three working strategies of the introduced converter selected for the overall polymer fuel module system. Here, the entire applied to the gate terminal is negligible. As a result, the consumer load observed power was also reduced extensively thereby zero heating effects on the MOSFET device. Based on the voltage-second concept, the capacitive elements’ current is exactly equal to one-third of the fuel stack voltage, and from Fig. 11b, the discontinuous state operational time value in switching strategies I, II, and III are defined as DT and D_xT, and (1-(D + D_x))T respectively. The peak existing currents of diodes are illustrated in Fig. 12. From Fig. 12, the diode peak summation currents are equal to the switching current which is mentioned in Eq. (74). The available inductive ripple currents under the converter discontinuous state are mentioned in Eq. (73). Finally, the operational time constant and gain at discontinuous stage are obtained as,

69

70

71

72

73

74

75

76

77

78

Fig. 12.

Available peak diode currents at the discontinuous state of the converter.

Comprehensive analysis of wide voltage DC–DC converter

The wide voltage-based single switch introduced converter is analyzed comparatively by considering the various existing power switching converters which are defined in Table 5. From Table 5, the introduced converter topology takes very few electromagnetic components thereby optimizing the overall converter inductive current ripples and power distribution losses of the fuel module power system. Here, the proposed converter electrostatic capacitors are five which are very low when associated with the quasi z-source DC–DC converter circuit and three-phase dual module DC–DC converter. Here, the boundary conditions state that the introduced converter gives very low-level voltage stress on all power switches thereby running the converter most efficiently under multiple temperature values of the fuel system. The compared converters’ voltage transformation variation along with the duty signal and its associated voltage that appears on the switch are represented in Fig. 13a, b.

Table 5.

Comprehensive analysis of universal wide voltage DC–DC converter circuit.

Topology	Gain	Devices used	Ground connected	Capacitors and inductors	Current distortions	Switching stress	Diode stress
TPMWSC59		4 power diodes & 1 DC–DC Switch	No	4, 1	Yes
GHVBPC60		1 power diode & 1 DC–DC Switch	No	1, 1	Uniform	1	1
ZSQPC61		3 power diodes & 1 DC–DC Switch	Yes	2, 2	Uniform	1	1
DPSSPC61		3 power diodes & 1 DC–DC Switch	No	3, 2	Uniform
DFHVPC		2 power diodes & 2 DC–DC Switches	Yes	3, 4	Uniform
TSSPC		4 power diodes & 1 DC–DC Switches	Yes	3, 3	Yes
ZDPHGPC		3 power diodes & 2 DC–DC Switches	No	5, 4	Yes	0.5	0.5
LCQPTC		2 power diodes & 2 DC-DC Switches	No	2, 2	Yes	0.5	0.5
SSWVSC		3 power diodes & 1 DC–DC Switch	No	5, 3	Uniform

Fig. 13.

(a) Variation of duty signal with power converter gain. (b) Converter gain variation for voltage stress on MOSFET switch.

Simulative validation of the SSWVSC-fed PEMFC system

The SSWVSC is implemented along with the polymer membrane-based fuel module system by utilizing the MATLAB Simulink tool. In this proposed system, the polymer cell is selected because of its merits low weight, and volume of the cell when associated with the other fuel cells. Also, the cells have the properties of high-power transformation density, fast system operating response, more working efficiency, safe handling, and easily generate a high amount of fuel module voltage. Finally, it works at all system functioning temperature conditions for different natural sunlight conditions. The selected polymer fuel module power and its associated rated voltages are 5.99 kW, and 65.12 V respectively. Here, the fuel stack produced at the current level is 134 A which is a very high amount of current and it can’t be directly used for the four-wheeler hydrogen vehicle system. Here, the single switch wide voltage source converter is developed for the polymer fuel module system to reduce the current conduction levels of the entire system. The utilized Cq, Cj, Cg, Cs, and Cm values are 244 µF, 244 µF, 244 µF, 244 µF, 244 µF, and Lc, Lb, and Ln values are 12.5 mH, 12.5 mH, and 12.5 mH respectively. Finally, the operated consumer load resistor value is 68 Ω.

Analysis of static polymer fuel module fed SSWVSC power system at T = 305 K

The introduced converter circuit source capacitive element suppresses the disturbances of the fuel module production voltage for diverse environmental fuel stack temperature conditions. This capacitor C_q is developed for storing the electricity under the switching state of the converter. Here, one operational temperature is considered for the investigation of the entire automotive system and the circuit L_b, C_q, L_c, C_j, C_s, D_n, and D_b removes the disturbances of consumer load voltage and it made for enhancing the voltage level from the lower value to higher value under operational fuel stack temperature condition 300 K. At this 300 K value, the fuel module production voltage is nonlinear. To develop the linearity in the fuel module system, the modified grey wolf technology-dependent ANFIS block is integrated along with the polymer fuel module system.

From Fig. 14a–f, the evaluated currents, fuel module voltages, system powers, and efficiencies of MPPT controller under stable operational temperature by applying the MP and O, ICC, RBFN, PSO-P&O, and ANFIS-GWO technologies are 104.27 A, 39.478 V, 4,116.6 W, 8.669 A, 403.634 V, 3499.11 W, 85.02%, 104.096 A, 39.563 V, 4,118.36 W, 8.710 A, 401.907 V, 3500.61 W, 85.89%, 103.48 A, 39.892 V, 4,128.30 W, 8.778 A, 404.458 V, 3550.341 W, 86.00%, 103.17 A, 40.191 V, 4,146.65 W, 8.820 A, 408.178 V, 3600.13 W, 86.62%, 109.43 A, 40.470 V, 4,428.79 W, 8.845 A, 430.611 V, 3808.76 W, and 86.66% respectively. Figure 14e, shows that the introduced converter is effectively enhancing the fuel module voltage from a very low value to the required load demand value and it suppresses the fluctuations of the source voltage efficiently with low cost. The system operational point stabilizing time under 305 K by interfacing the MP and O, ICC, RBFN, PSO-P&O, and ANFIS-GWO algorithms are 0.046s, 0.0483s, 0.0399s, 0.0365s, and 0.0371s respectively.

Fig. 14.

(a) Polymer fuel module current, (b) obtained source voltages, (c) fuel module power, (d) SSWVSC current, (e) obtained load voltage, (f) power at static 305 K.

Analysis of dynamic polymer fuel module system at T = 305 K, 355 K, 405 K, 455 K, and 505 K

The introduced converter circuit is integrated with the grey wolf-based ANFIS technology for capturing the peak power from the proposed polymer fuel system to produce the electrical energy for heavy-duty electrical vehicle systems. Here, the introduced system is analyzed at quick variation of the supply operational temperature. From Fig. 15a–f, the fuel supply currents, input voltages to the converter, load delivered powers, introduced controller efficiency and settling times at 355 K, and 405 K temperatures by connecting the MP and O, ICC, RBFN, PSO-P&O, and proposed MPPT controllers are 107.86 A, 46.002 V, 4,961.94 W, 8.861 A, 486.562 V, 4311.43 W, 86.89%, 0.021s, 108.42 A, 46.221 V, 5,011.49 W, 8.88 A, 493.078 V, 4378.54 W, 87.37%, 0.0272s, 108.58 A, 46.366 V, 5,034.85 W, 9.088 A, 484.205 V, 4400.462 W, 87.40%, 0.0285s, 113.70 A, 46.552 V, 5,293.39 W, 9.108 A, 508.649 V, 4632.78 W, 87.52%, 0.0299s, 117.68 A, 46.71 V, 5,497.29 W, 9.121 A, 524.355 V, 4782.65 W, 87.81%, 0.0321s, 103.14 A, 47.99 V, 4,950.00 W, 8.781 A, 492.971 V, 4328.78 W, 87.45%, 0.02s, 116.42 A, 48.771 V, 5,678.19 W, 9.911 A, 503.423 V, 4989.432 W, 87.87%, 0.020s, 117.08 A, 48.91 V, 5,726.56 W, 10.70 A, 507.817 V, 5033.65 W, 87.9%, 0.0226s, 115.85 A, 50.88 V, 5,894.68 W, 10.719 A, 521.253 V, 5187.321 W, 88.106%, 0.0255s, 115.73 A, 51.05 V, 5,908.19 W, 10.783 A, 547.262 V, 5199.21 W, 88.21%, and 0.0271s. The analysis of the polymer fuel cell at 455 K and 505 K is given in Table 6.

Fig. 15.

(a) Polymer fuel module current, (b) obtained source voltages, (c) fuel module power, (d) SSWVSC current, (e) obtained load voltage, (f) power at dynamic 305 K, 355 K, 405 K, 455 K, and 505 K.

Table 6.

Comparative Analysis of grey wolf technology optimized ANFIS membership selection based PEMFC for electric vehicles.

Technology	Source current (A)	Source voltage (V)	Source power (W)	Load current (A)	Load voltage (V)	Load power (W)	Efficiency of MPPT (%)	MPP settling time (s)	Design complexity
Operation of the polymer fuel power production system at 305 K
MP and O	104.27	39.478	4116.6	8.669	403.634	3499.11	85.02	0.046	More
ICC	104.096	39.563	4118.36	8.710	401.907	3500.61	85.89	0.0483	Heavy
RBFN	103.48	39.892	4128.30	8.778	404.458	3550.341	86.00	0.0399	More
PSO-P&O	103.17	40.191	4146.65	8.820	408.178	3600.13	86.62	0.0365	Medium
ANFIS-GWO	109.43	40.470	4428.79	8.845	430.611	3808.76	86.66	0.0371	Less
Operation of the polymer fuel power production system at 355 K
MP and O	107.86	46.002	4961.94	8.861	486.562	4311.43	86.89	0.021	More
ICC	108.42	46.221	5011.49	8.88	493.078	4378.54	87.37	0.0272	Heavy
RBFN	108.58	46.366	5034.85	9.088	484.205	4400.462	87.40	0.0285	More
PSO-P&O	113.70	46.552	5293.39	9.108	508.649	4632.78	87.52	0.0299	Medium
ANFIS-GWO	117.68	46.71	5497.29	9.121	524.355	4782.65	87.81	0.0321	Less
Operation of the polymer fuel power production system at 405 K
MP and O	103.14	47.99	4950.00	8.781	492.971	4328.78	87.45	0.02	More
ICC	116.42	48.771	5678.19	9.911	503.423	4989.432	87.87	0.020	Heavy
RBFN	117.08	48.91	5726.56	10.70	507.817	5033.65	87.9	0.0226	More
PSO-P&O	115.85	50.88	5894.68	10.719	521.253	5187.321	88.106	0.0255	Medium
ANFIS-GWO	115.73	51.05	5908.19	10.783	547.262	5199.21	88.21	0.0271	Less
Operation of the polymer fuel power production system at 455 K
MP and O		49.71	4,950.71	9.427	467.395	4406.14	89.05	0.0179	More
ICC	107.746	49.88	5,374.40	10.214	468.931	4789.67	89.12	0.0183	Heavy
RBFN	102.90	49.989	5,144.21	10.783	425.639	4589.67	89.22	0.018	More
PSO-P&O	106.64	50.381	5,372.88	11.299	424.895	4800.89	89.354	0.0191	Medium
ANFIS-GWO	117.04	50.443	5,904.06	11.363	484.086	5300.67	89.78	0.019	Less
Operation of the polymer fuel power production system at 505 K
MP and O	105.83	47.67	5044.94	8.841	520.415	4600.99	91.298	0.0141	More
ICC	105.516	47.82	5045.80	9.542	482.793	4606.82	91.376	0.01543	Heavy
RBFN	107.36	48.897	5249.69	9.781	490.78	4800.324	91.44	0.014	More
PSO-P&O	109.84	49.71	5460.51	10.8	462.576	4995.823	91.49	0.0176	Medium
ANFIS-GWO	113.59	49.887	5665.99	10.836	479.904	5200.25	91.78	0.012	Less

Experimental evaluation of SSWVSC

The introduced converter circuit is tested here by using the programmable power DC source (M62020P8058) instead of the fuel module. Here, the programmable supply is assumed as a source for testing the wide voltage gain DC-DC circuit. From Fig. 16, a 0 to 12 V transformer is used to give energy to the TLP-440. The TLP-440 provides complete isolation between the MOSFET and high-power rating source voltages. It is an 8-pin static device that provides 7 mA, and 5 V energy to the power switch. Here, the Gallium Arsenide (GaAlAs) power semiconductor is selected to obtain the switching pulses to the power transmission circuit with the help of a photodetector circuit. In this converter design circuit, the MOSFET device is used because of its features are high operational temperature withstand ability, less conduction resistance, high amount of power transmission efficiency, negligible power distribution losses, and moderate size.

Fig. 16.

Tested prototype model of the proposed new DC-DC circuit.

From Fig. 17, the operational switching frequency is equal to f_s=20 kHz. In this proposed circuit, an analog discovery device is selected for supplying the gate-source voltage to the MOSFET. The properties of analog discovery are up to six pulses generation, able to provide 0.01 duty cycle to 0.99 duty signal. In this setup, the tested duty signal strength is 0.1 for observing the improved load voltage profile. The source voltage applied in this experimental setup is 61.7 V, and the identified internal resistive properties of the inductive and electro-static capacitive elements are rC_q, rC_j, rC_g, rC_s, rC_m, rL_c, rL_b, and rL_n are 12 mΩ, and 28mΩ respectively. The evaluated switch voltage drop and its conduction resistance are 1 V, and 18 mΩ. Finally, the diode voltage drops and its equivalent on state resistance are 70 mΩ, and 0.9 V respectively. Here, the obtained inductive ripples, and capacitive ripples are 25% and 10%.

Fig. 17.

Applied switching signal to power MOSFET device and its gate-source voltage.

From Fig. 18, the input voltage of 61.7 V is improved to 123.2 V by utilizing the 0.1 duty signal strength and 20 kHz switching frequency. Here, the evaluated source inductor current is 3.77 A, and the 100 W load consumed current is 1.45 A which indicates that the proposed converter voltage transmission is improved by reducing the conduction losses of the entire system. The overall experimental prototype model design constraints are mentioned in Table 7. Based on the overall experimental investigation, the developed prototype model gives 22.6 W overall power distribution loss. The operated power diodes loss is 3.616 W and it is a 16% loss from the entire power loss. Also, the conduction of switch and switching loss values are 5.424 W, and 3.842 W respectively. These losses are 24%, and 17% of the entire circuit losses. Finally, the capacitive and inductive losses are 5.2 W, 4.9 W which are 23%, and 21.68% of the overall loss.

Fig. 18.

Utilized programmable source voltage and its operational load voltage.

Table 7.

Utilized converter-developed parameters for testing the load voltage profile.

Variables	Values
Identified MOSFET IC number	IRF640N
Applied programmable source voltage	75.12 V
Rated voltage measurement meters	100:1, 5 kV
Optocoupler tested rated voltage level	6 to 9 V
Testing rated load value	100 W bulb
Entire converter operated switching frequency	20 kHz
Obtained voltage level of the converter circuit	122.9 Volts
Cq, Cj, Cg, Cs, and Cm	244 µF
Lc, Lb, and Ln values	12.5 mH
Selected DSO model for testing	TPS-2024B

Conclusion

The overall polymer fuel module fed wide power transmission DC-DC circuit is investigated by applying the MATLAB tool. Here, the PEMFC is utilized in the first objective because its advantages are more power density productivity, quick operational response, ease of handling, capability to operate at lower source temperature values, and low volume. The fuel module produces a highly nonlinear voltage which is linearized in the second objective by proposing the grey wolf-fed ANFIS technology. The features of ANFIS hybrid MPPT technology are less development cost, fewer sensing devices needed, more fuel module power extraction efficiency, and high robustness. However, the fuel module voltage production is low which is improved in objective three by using the single switch power transmission circuit. The introduced converter features are less voltage appearance across the diodes, more voltage transmission ratio, and are widely used for all renewable energy power production systems. Finally, due to the lack of facilities, the introduced converter is investigated by applying the programmable power source.

Abbreviations

MPPT

Maximum power point tracking

FSO

Flying squirrel search optimization

CSO

Cuckoo search algorithm

ANFIS

Adaptive neuro-fuzzy inference system

P&O

Perturb and observe

PEMFC

Proton exchange membrane fuel cell

SSWVSC

Single switch wide voltage supply converter

FLC

Fuzzy logic controller

GSO

Glowwarm swarm optimization

ZDPHGPC

Zsource dual phase global power circuit

MOSFET

Metal oxide semiconductor field effect transistor

MPO

Modified perturb and observe

PSO

Particle swarm optimization

TPMWSC

Triple phase mode wide source converter

GHVBPC

Global high voltage boost power circuit

ZSQPC

Zsource quadratic power circuit

DPSSPC

Dual power single switch power converter

DFHVPC

Dual flow high voltage power converter

TSSPC

Triple source supply power circuit

LCQPTC

Linear continuous quadratic power transmission circuit

Author contributions

All authors contributed to the study, conception, and design. all authors commented on the manuscript. All authors read and approved the final manuscript.

Data availability

The datasets used and/or analyzed during the current study are available from the corresponding author upon reasonable request.

Declarations

Competing interests

The authors declare no competing interests.

Ethical approval

This paper does not contain any studies with human participants or animals performed by any of the authors.