Secure traceability mechanism of green electricity based on smart contracts and provenance model

Abstract

Green and low-carbon development is an important part of global sustainable development. Green power trading provides strong support and assurance for promoting green and low-carbon development. Due to the long cycle of green power data chains and their susceptibility to malicious tampering, the integrity and traceability of data are difficult to guarantee. Therefore, this paper first proposes a security provenance model with enhanced relations based on the core structure of PROV and blockchain technology, which can securely capture provenance records, use the transfer time and number of transactions between various links in the traceability network as reasoning clues, realize the correlation tracing of the green electricity transfer process. Under the model, a traceability mechanism of green electricity is designed based on smart contracts. Trustworthy green electricity data collection is achieved through data filling and data verification techniques. Traceability query technique is adopted to achieve trustworthy traceability of green electricity. And the effectiveness of the proposed solution is demonstrated through simulation experiments.

1. Introduction

Since the Industrial Revolution, the reliance on fossil fuels has brought significant changes to the way societies operate, resulting in a series of issues including air pollution and climate change. The impact of burning fossil fuels on the environment and human health underscores the necessity of transitioning towards sustainable energy sources. In 2015, the United Nations General Assembly adopted the Sustainable Development Goals (SDGs), providing a robust framework for international cooperation towards achieving a sustainable future for the planet [1]. Global energy goal SDG 7 includes three key objectives: ensuring affordable, reliable, and universal access to modern energy services; substantially increasing the share of renewable energy in the global energy mix; and doubling the global rate of improvement in energy efficiency [2]. In recent decades, the international community has increasingly focused on addressing climate change and continuously sought to establish a carbon-neutral and environmentally friendly model [3].

To address climate change, promoting the development of green energy is particularly important. The green electricity market in the United States includes both mandatory markets based on renewable energy quotas and voluntary trading markets. Thirty-one states in the US have implemented quota systems, with seven states setting renewable energy generation ratio targets. Quota systems have become the most commonly used policy tool for controlling greenhouse gas emissions in various states in the US [4]. The Brazilian government encourages the development of sustainable hydrogen industries using solar energy, biomass energy, and biogas, and constructed the first green hydrogen plant on July 26, 2022. Europe has established a voluntary green electricity Guarantees of Origin (GO) market, covering wind, solar, hydroelectric, nuclear, and biomass power generation. One megawatt-hour of green electricity receives one electronic GO certificate. The new President of the European Union unveiled the outline of the “European Green Deal” in Brussels, aiming for Europe to lead globally in achieving carbon neutrality by 2050, meaning net carbon dioxide emissions will be reduced to zero, thus promoting stable and sustainable economic development in Europe, improving public health and quality of life, and protecting the natural environment [5], [6].

Currently, in the process of green electricity traceability, there exists a phenomenon known as “greenwashing,” where some energy suppliers package electricity from non-renewable sources as “green electricity” and sell it to consumers, posing challenges to the transparency and authenticity of green electricity. The green electricity trading market also faces difficulties in establishing a fully closed-loop trading process and in certifying and tracing the origin of green electricity from both physical and financial dimensions. Additionally, the chain of green electricity data has a long cycle, and for economic reasons, some users may maliciously tamper with green electricity data. It is urgent to achieve transparency and controllability of green electricity throughout the entire process of generation, transmission, distribution, trading, and consumption to ensure the tamper resistance, traceability, confidentiality, and auditability of data.

This paper proposes a green electricity traceability mechanism based on smart contracts and provenance model with enhanced relations and refined entities (ERRE-PM) model. To ensure the authenticity and integrity of traceability information, ERRE-PM model is proposed to guarantee the authenticity of data sources and credibility in the traceability process. To tackle the issue of difficulties in tracing the sources of green electricity, ERRE-PM model provides standardized method for recording data operations, capturing and recording data sources, and using the flow time and transaction frequency between various links in the network as inference clues to achieve association tracing of the green electricity circulation process. Furthermore, to address uncertainties, data quality, and data lineage issues in the circulation process of green electricity data, approximate traceability queries are employed to ensure the accuracy and reliability of green electricity data.

The rest of the paper is organized as follows: Section 2 introduces data provenance, the PROV model, and the application of blockchain technology in the field of green energy. Section 3 elaborates on the blockchain-based ERRE-PM model. Section 4 introduces the green electricity traceability mechanism based on smart contracts and the improved PROV model. Section 5 evaluates the proposed solutions through experiments. Finally, Section 6 concludes the paper.

2. Related works

2.1. Data provenance and PROV model

Simmhan et al. define data provenance as the process information from source data to derived data products [7]. Lanter defines data provenance as the description of original data before the derivation of target data and its evolutionary process [8]. Greenwood et al., expanded the definition to include metadata for recording workflow evolution processes [9], annotation information, and experimental processes [10]. Dai Chaofan defines data provenance as the information that records changes and processing undergone by data throughout its entire lifecycle from generation to extinction or transformation [11].

An effective model is crucial for data provenance technology. The main data provenance models include the four-dimensional provenance model, flow-based provenance information model, data provenance security model, time-value-centric provenance model, open data provenance model, provenir data provenance model, PrInt data provenance model, etc. The primary methods for data provenance tracking include annotation method and reverse query method. In addition, there are general data tracking methods such as bidirectional pointer tracking method, bit vector storage positioning method, and tracking methods utilizing graph theory principles and specialized query languages [12], [13].

The World Wide Web Consortium (W3C) released the PROV provenance standard [14], It has strong extension standardization, which can accommodate data sources generated from different sources, and traces provenance information at different stages of each entity [15]. The core of the model is PROV-DM [16], which includes instances of three types of PROV elements: entity, activity, and agent [17]. An entity refers to physical, digital, conceptual, or other types of things. An activity refers to actions that use or create entities, while an agent refers to individuals, organizations, or software that perform activities and can have various attributes such as names, contact information, etc. The PROV model represents data provenance through defined classes and relationships. It is illustrated in Fig. 1.

Figure 1.

Core concepts of PROV.

At present, scholars have proposed different PROV model applications in different fields. Yang et al. add entity and relationship types for open and secure access to Earth observation data [18]. In the context of the Internet of Things, Jaigirdar et al. proposed the provenance model Prov-IoT, which combines security metadata with provenance data to lay the foundation for data trust [19]. Gao et al. extended PROV-DM according to the characteristics of big data, enriched the source analysis function, and provided data security supervision methods [20]. Yao et al. integrated the PROV model into the blockchain and designed a multi-bucket index and a skip table index to query provenance information [21].

2.2. Blockchain applications in green power

Blockchain is a publicly transparent, decentralized database [22], characterized by immutability, multi-party consensus, and traceability [23]. In a blockchain network, member nodes do not rely on third parties to arbitrate transactions; instead, they negotiate ledger contents through consensus protocols, using hash encryption algorithms and digital signatures to ensure transaction integrity. Depending on the composition of member nodes and accessibility, blockchains can be categorized as public chains, private chains, and consortium chains [24], [25]. Smart contracts are code stored and executed on blockchain networks [26], [27]. They automatically run when certain conditions or standards are met, without the need for third-party involvement. Smart contracts can express triggers, conditions, and complex logic. The deployment of smart contracts on a blockchain is illustrated in Fig. 2.

Figure 2.

Deployment of smart contracts on blockchain.

Currently, some studies have been conducted on the integration of renewable energy with blockchain and data traceability. For instance, the researchers of the reference [28] analyze the green electricity trading process for the Winter Olympics, designing a green electricity traceability mechanism covering the “generation-transmission-distribution-trading-consumption” five stages through blockchain technology. The authors of the reference [29] introduce a blockchain-based electricity trading and energy management platform, enabling multi-level power trading at electric vehicle charging stations. In the reference [30], the authors combine blockchain technology and green certificate trading mechanism, proposing a comprehensive energy system optimization model with cross-chain trading involving green certificates. Reference [31] proposes a blockchain-based green certificate trading matching and circulation model to enhance the credibility and traceability of renewable energy consumption processes. Reference [32] presents a green certificate and carbon joint trading market model based on smart contracts [33], connecting green energy companies with traditional fossil energy generation companies to promote energy transition. In the reference [34], the authors propose a blockchain-based design for renewable energy microgrids. Reference [35] introduces the concept of electricity supply management systems and the application of blockchain technology. Reference [36] proposes a trustworthy scheduling method for high renewable energy penetration power systems in distribution networks based on blockchain. The researchers of the reference [37] identify installation methods for renewable energy systems and introduce a new blockchain-based framework for distributed renewable energy management systems.

3. The proposed ERRE-PM model

3.1. PROV model for green power traceability

The PROV model plays an important role in the context of green power, providing a standardized methodology for documenting data manipulation. In the context of green power traceability, the PROV model can be used to track information on renewable energy sources, power transmission, conversion, trading, etc. The model captures the entities involved in the relevant processes, the timing and duration of the activities, and detailed information on the inputs and outputs they generate, among other things. For example, in the case of a solar power plant that produces green electricity, the PROV model can be used to track the source of the solar panels, track the efficiency of the power plant, and monitor the energy usage of the end user. This information can be used to verify the authenticity and quality of the green power produced and ensure that it meets regulatory requirements.

The three instances of the PROV model, Entity, Activity, and Agent, make it possible to track entities representing other components such as transmission lines, substations, transformers, etc., activities such as power data collection, conversion of the form of electricity (e.g., voltage regulation or power factor correction), and subjects involved in the processes of power generation, trading, and so on. The PROV model allows to obtain new information from existing source data, e.g., based on the relationship lineAdjusted(), as shown in equation (1):

$$\text{ prov: }\mathrm{Activity}(a)\wedge \text{ prov: Entity }(e)\wedge \text{ prov: used }(a,e)\wedge \mathrm{prov}:\mathrm{Agent}(g)\wedge \text{ prov: wasAssocaitedWith }(a,g)\Rightarrow \text{ prov: lineAdjusted }(e,g)$$ (1)

For Activity a and Entity e, if Activity a uses Entity e and there is a connection between Activity a and Agent g, then it can be inferred that the line Entity e is adapted by Agent g.

3.2. The proposed ERRE-PM model

While the PROV model outlines fundamental relationships among entities, activities, and agents, it may not adequately address the finer details required for green electricity traceability. Thus, this paper extends PROV-DM and designs the ERRE-PM model to enhance relationships and refined entities. It further specifies instances of Entity, Activity, and Agent, expanding the source relationships beyond the definitions in PROV-DM to achieve more granular traceability information. This enhancement bolsters the traceability capability, elevates the credibility and reliability of traceability information, and facilitates traceability analysis.

3.2.1. Example refinements

(1) Entity:

The formal definition of an Entity is articulated as Entity: <ID, [attr1 = value1, attr2 = value2, ...]>, where <ID> denotes the unique identifier of the entity, uniquely identifying the entity in the traceability process. The list [attr1 = value1, attr2 = value2, ...] represents the attributes of the entity, encompassing the properties of the entity and their corresponding values. This allows for the addition of various attributes according to specific needs, with appropriate values assigned.

Entity includes entities such as energy production, energy trading, and energy consumption. Energy production entities can be refined into different types of energy production facilities, such as wind energy, solar energy, hydro energy, etc., and corresponding attributes are defined for each energy production facility.

(2) Activity:

The formal definition of an Activity is given as: Activity: <ID, Type, status, [attr1 = value1, attr2 = value2, ..., energyType = energyTypeValue, quantity = quantityValue, timestamp = timestampValue, ...]>. This paper extends additional attributes such as status, which denotes the execution state of the Activity, facilitating activity tracking, exception handling, and traceability result analysis. “energyType” specifies the type of energy, “quantity” represents the amount of electricity produced/traded/consumed, and “timestamp” denotes the time stamp of production/trade/consumption. Thereby enabling a more detailed recording and representation of activities within the green electricity traceability process.

Activity includes activities such as energy trading, green power certification, data processing and analysis, and power generation. Data processing and analysis can be refined into data cleaning, conversion, analysis, and mining to support data analysis for green power traceability. Another example is that green power certification can be refined into green power certification audits, green power certification certificate issuance, etc. to ensure the credibility of green power certification.

(3) Agent:

The formal definition of an Agent is articulated as Agent: <ID, name, [attr1 = value1, ...]>, where ID is the unique identifier of the Agent, used to identify the Agent's identity within the green electricity traceability system. The name is used to describe the Agent's name or identifier. [attr1 = value1, ...] represents the set of attributes for the Agent, which can include zero or more attribute-value pairs, used to describe the Agent's characteristics, properties, or other information. The attributes and corresponding values can be defined and expanded according to actual application needs.

Agents include various types of roles in green power traceability scenarios, such as green power certification agencies, green power producers and consumers, etc. Each Agent has different attributes and attribute values. Through the formal definition of Agent, the Agent is identified, described and associated to support data management, source tracing and audit verification in the green power traceability process.

3.2.2. The expansion of dependency relationships

The PROV-DM model defines seven types of relationships to describe interactions among entities, activities and agents. However, in the context of green electricity, these relationships do not suffice to depict complex associations and fail to deliver accurate and reliable traceability information. Therefore, this paper introduces the ERRE-PM model, which expands upon the relationships among entities, activities, and agents in the PROV-DM model to better accommodate the needs of green electricity traceability.

(1) Transformation:

It refers to the conversion relationship between entities and activities, used to describe how an entity changes through the transformation of an activity. It is formally defined as Transformation: <Entity, Activity, Time, [Quality, Constraints, attr1 = value1, ...]>, where Quality indicates the quality characteristics of the transformation outcome, such as accuracy, integrity, assessing the effect and reliability of the transformation process. Constraints represent the conditions or limitations during the transformation process. The Transformation relationship records the conversion process of energy or other entities in the green electricity traceability process, spanning the production, consumption, and storage phases of energy.

(2) Verification:

It denotes the relationship between entities and agents, aimed at describing the process by which an agent verifies or confirms an entity. It is formally defined as Verification: <Entity, Agent, Time, [Method, Result, Confidence, attr1 = value1, ...]>, where Entity represents the entity being verified, such as energy or materials. Agent signifies the verifying agent, which could be regulatory bodies or third-party certification organizations. Method indicates the specific methods, algorithms, or procedures used for verification. Result documents the verification outcome, recorded in binary or quantitative evaluation formats. Confidence expresses the confidence level of the verification results. In green electricity traceability scenarios, the Verification relationship can be utilized to describe the processes of verification, review, and inspection of energy or other entities to ensure their authenticity and reliability.

(3) Distribution:

It refers to the relationship between entities and agents, aimed at describing the process through which an agent distributes an entity. It is formally defined as Distribution: <Agent, Entity, Time, [Location, Medium, Quantity, Proof, attr1 = value1, ...]>. Here, Agent denotes the agent responsible for distribution, typically producers, suppliers, or distributors. Entity represents the distributed entity, such as energy, products, or materials. Location indicates the specific location of the distribution. Medium refers to the medium or communication channel used for distribution, like network transmission or logistic transport. Proof is the method of proving the distribution operation, ensuring its credibility, which could be a hash value, for example. In the context of green electricity traceability, the Distribution relationship can be used to describe the distribution, transmission, and delivery processes of energy or other entities, including the transfer of energy from producers to consumers and the distribution of energy within the power grid.

(4) Involvement:

It denotes the participatory relationship between agents and activities, describing how an agent engages in an activity. It is formally defined as Involvement: <Agent, Activity, Time, [Role, Contribution, attr1 = value1, ...]>, where Role indicates the role played by the Agent within the activity, Contribution denotes the degree of the Agent's contribution or influence within the activity, quantifiable as an indicator, and [attr1 = value1, ...] represents optional attributes providing additional information about the Involvement relationship, such as the manner and extent of participation. The Involvement relationship outlines the participation of various roles in the green electricity traceability process, including data collection, verification, and publication.

3.3. ERRE-PM model based on blockchain

ERRE-PM model can capture data sources, record data transformations and operations performed. So, it contributes to the quality of the data. However, it still has the following shortcomings.

• •Access to and distribution of traceability information is restricted.
• •It is possible to describe how data is created, transformed, and utilized, but data verification and the prevention of data tampering cannot be ensured.
• •Ensuring the authenticity and integrity of traceability information is challenging, and security cannot be guaranteed.

The combination of blockchain technology with the ERRE-PM model can effectively address these issues. For the problem of restricting access to and distribution of traceability information, blockchain technology offers a decentralized architecture that allows anyone to access information stored on the blockchain, and also enables specific users to be authorized to update or modify the information.

Blockchain technology offers a tamper-proof solution for ERRE-PM model data. When data is stored on the blockchain, it becomes an immutable part of a distributed ledger, unable to be changed or deleted. Moreover, blockchain's cryptographic hashing and consensus mechanisms provide a secure and verifiable method to record the origin of data. Recording ERRE-PM model data on the blockchain enables the detection of any changes or modifications using the data's cryptographic hash. Additionally, the immutability and transparency of blockchain allow for the verification of data authenticity, making it easier to identify any tampering or modifications, thus enhancing the credibility of the data.

Blockchain technology enhances the authenticity of data through the use of digital signatures to sign and verify each transaction. It provides a timestamp to verify the timing of each transaction, clearly documenting the history of data modifications and updates, thus improving the granularity and accuracy of ERRE-PM model data. Timestamps enable the establishment of a verifiable audit trail through event causality chains. Data replication can ensure data integrity. By integrating consensus mechanisms into the ERRE-PM model, consensus mechanisms help eliminate errors and fraudulent activities, maintain data immutability, and ensure consensus on data interpretation. Fig. 3 illustrates the blockchain-based ERRE-PM model, which supports regulatory bodies in the full traceability and audit of data. Through smart contracts, green electricity traceability information is recorded in a transparent and tamper-proof manner. The distributed nature and consensus mechanism of the blockchain ensure the credibility of the traceability process. Digital signatures guarantee the authenticity of the data, while cryptographic theories such as public key infrastructure and identity-based cryptography protect the privacy and security of sensitive green electricity transaction data.

Figure 3.

ERRE-PM model based on blockchain.

4. Traceability mechanism of green electricity

4.1. Data collection of electric power

By collecting active power, the consumption of electricity can be calculated; by collecting harmonics, power factor, voltage phase imbalance, current phase imbalance, etc., the quality of electrical energy can be analyzed; by collecting current (phase A/B/C), voltage, active power, reactive power, apparent power, etc., the operating condition of equipment can be understood. However, during the process of power acquisition, power acquisition equipment may be affected by weather, signals, human factors, resulting in problems such as data missing and low integrity of the collected data, which affect the reliability and accuracy of data analysis. Therefore, it is necessary to perform operations such as jitter removal, data validation, and data imputation on the collected power data to ensure the availability and integrity of the collected data. The power source data acquisition process is shown in Fig. 4.

Figure 4.

Traceability data acquisition process.

Bump removal is used to resist the interference of unstable but not yet abnormal data. The specific principle is as follows: assuming the input data is, select the two data points before and after this data point, and compare the aggregation of the two same length windows before and after [38]. If the ratio exceeds the ratio, it is considered as abnormal data. The equation for bump data detection is shown in equation (2) (where n is the length of the detection window and $2n<m$):

$${\mathrm{y}}_{n+1}+\dots +{\mathrm{y}}_{2n}>a({\mathrm{y}}_1+\dots +{\mathrm{y}}_n)$$ (2)

In order to ensure the accuracy and reliability of green power data collection, the Kalman filter method is used to predict and estimate the dynamic system state. By dynamically adjusting filter parameters to respond to data noise, it effectively filters out environmental and measurement noise, and integrates data from different sensors to improve overall measurement accuracy. The Kalman filter formula consists of two parts: the state equation and the observation equation. The state equation represents the dynamic relationship between state variables and is typically described using a first-order linear difference equation [39]. Its general form is shown in equation (3):

$$x_k=F_k{x}_{k-1}+w_k$$ (3)

where $x_k$ represents the state variables at time k, $F_k$ denotes the system state transition matrix, and $w_k$ represents the noise term, typically assumed to follow a Gaussian white noise distribution with mean 0 and variance Q.

The observation equation represents the relationship between state variables and measurement values, usually described using a linear combination relationship. Its general form is shown in equation (4):

$$z_k=H_k{x}_k+v_k$$ (4)

where $z_k$ represents the measurement values at time k, $H_k$ denotes the transition matrix between state variables and measurement values, and $v_k$ represents the observation error term, typically assumed to follow a Gaussian white noise distribution with mean 0 and variance R. In the Kalman filter algorithm, the variances R and Q of Gaussian white noise represent the uncertainty of system noise and observation noise, respectively. The Kalman filter calculates the optimal estimates and variances of state variables through recursive iteration and adjusts and corrects the estimates and variances of state variables based on the latest observation values, thereby maintaining an accurate estimation of the system state.

4.2. Traceability mechanism based on smart contracts and ERRE-PM model

The uncertainty of green electricity data pervades various stages such as generation, transmission, distribution, trading, and consumption, with a potential risk of data tampering. Thus, establishing a system of mutual trust and privacy protection is essential to eliminate information silos in green electricity production and transmission, enhancing data authenticity and integrity.

This paper proposes a traceability mechanism of green electricity based on smart contracts and the ERRE-PM model. Data supplementation and data verification technology are adopted to achieve accurate and trustworthy data collection. Then using the ERRE-PM model to register related information such as generation, transmission and distribution, trading and consumption on the blockchain, realizing consensus and storage of data, providing a data basis for green electricity traceability. The smart contracts associated with the ERRE-PM model verify the electricity source and generate digital certificates, stored on the blockchain. The digital certificates contain information such as the amount of electricity generated, the source of electricity, and production time. The traceability of green electricity can be completed through smart contracts and traceability query technology. It is shown in Fig. 5.

Figure 5.

Green power traceability machine mechanism.

At first, the electric power data is collected. Then, it undergoes business data processing and hash calculation, which is then signed before calling an interface on the blockchain network to complete the data uploading. After the data is uploaded, the blockchain network verifies the consistency of the data through a consensus mechanism, then generates a hash value using a hash algorithm. This hash value is stored on the blockchain network via a Merkle tree, and the location of the data block is also stored on the blockchain, allowing the blockchain network to control and manage access to the data. Moreover, for sensitive data in the transaction process, symmetric encryption algorithms, asymmetric encryption algorithms, hash algorithms are used to ensure the integrity and privacy of transaction data, and digital signatures are employed to secure the safety of green electricity transactions.

By recording timestamps of green electricity transactions or energy production activities, along with the participants and the number of transactions using the ERRE-PM model, it is possible to extract the flow time and transaction count between traceability links in the data provenance network. This provides a basis for inferring the credibility of key nodes, constructing an efficient and reliable green electricity transaction traceability model in conjunction with smart contract technology. Techniques such as timestamps and digital signatures ensure the authenticity and integrity of green electricity transaction data.

When renewable energy generators supply energy to the grid, smart contracts can instantly record the data's origin and verify compliance with standards. The ERRE-PM model facilitates the tracking and verification process, ensuring data accuracy based on captured traceability information. Through smart contracts, traceability queries enable trustworthy traceability of green electricity.

Traceability queries through smart contracts can complete the credible traceability of green electricity. Algorithm 1 implements the green electricity traceability query, which includes parameters such as producer $(pro)$, date, certificate information $(cert)$, electricity transmission records $(transfer)$, power record number $(recordID)$, electricity receiver address $(receiver)$, and transaction information $(tran)$. Functions $recordGP()$ and $recordPT()$ log green electricity information and green electricity transmission and transaction information, respectively. The function $TraceabilityQuery$ conducts green electricity traceability queries, returning relevant information about the query results.

Algorithm 1.

Green power traceability query

Smart contracts perform green electricity verification on the green electricity data and source records provided by the ERRE-PM model. The verification algorithm of green electricity is as follows in Algorithm 2, in which Tid, Type, and Generation represent the transaction id, type of renewable energy, and quantity of renewable energy generated, respectively. Total, gs, and rc represent total generated power, green energy standards, and renewable energy certificates, respectively. Initially, the ratio of Generation to total is compared with the green certification level to see if it meets the standard. The functions vcIssuer(rc) and vcInfo(rc) verify whether the certificate comes from a trusted third party and whether the certificate information matches the actual situation, respectively. If all processes are verified, green electricity verification succeeds. Otherwise, it fails.

Algorithm 2.

Green power verification

4.3. Data traceability query of green electricity

4.3.1. Uncertain complex events in power measurement

In the process of green energy data collection, transmission, and utilization, data enters processors in the form of streams. Stream data arrives rapidly and has characteristics such as multiple sources, large data volumes, and complex formats. The data elements of stream data appear in chronological order and exhibit time series correlation. Therefore, it is challenging to manage and analyze the traceability of stream data.

Uncertainty exists in various stages of green energy data processing, which may result in inaccurate data and data loss. Taking electric energy measurement as an example, electric energy is affected by factors such as the accuracy of measuring instruments, load characteristics, and environment, leading to errors in electric energy measurement and affecting the accuracy and reliability of measurement results. Uncertainty permeates throughout the process of data evolution, significantly affecting the accuracy and effectiveness of data query results. Therefore, it is necessary to analyze the sources of uncertainty, trace back green energy data, and improve the effectiveness and data quality of green energy data.

The table below takes electric energy measurement as an example and selects the installation location of measuring devices, electric energy meters, transformers, electronic energy meters, electronic card meters, secondary wire voltage drop, and electric load as basic events. The probabilities of occurrence of relevant atomic events $x_i$ are shown in Table 1.

Table 1.

Electrical energy metering probabilistic atomic events.

Event ID	Description	Probability of occurrence
x1	Energy meter error	0.02
x2	The electrical coil is broken	0.05
x3	Replace the battery coil	0.20
x4	The transformer has an error	0.07
x5	Secondary wire voltage drop error	0.02
x6	Aging of equipment	0.10
x7	Electricity meter adjustment	0.10
x8	Electric overload	0.08
x9	Adjusting load	0.15
x10	Optimized line connection	0.08
x11	Regular replacement of electronic metering meters, electronic card tables and other equipment	0.05
x12	Measuring device installation position deviation	0.35
x13	Adjust the installation position of the metering device	0.06

4.3.2. Approximate traceability query

In the process of green energy data flow, a large number of sensors, mobile terminals, smart devices, video surveillance equipment, they will continuously generate probabilistic atomic events. In these links, adopting the strategy of approximate traceability representation to track several most important traces will be more reasonable without affecting the query results [39]. Approximate traceability queries are suitable for the analysis of large datasets and allow for the identification of differences in green tracking data. Approximate traceability does not track every derivation but only the most influential facts to compress data, reducing the amount of computational resources required for analysis. This approach can handle queries and provide explanations while directly returning the most important derivations [40].

Definition 1

When ${\lambda }_t^s$ and ${\lambda }_t$ are consistent in most assignments, ${\tilde{\lambda }}_t^s$ can be defined as an approximation of ${\lambda }_t$. Then, we can obtain the approximate traceability query function ${\tilde{\lambda }}^s(G,W)$, the specific definition is shown in equation (5):

$${\tilde{\lambda }}^s(G,W)\stackrel{\text{def}}{=}\prod _{t:t\in W}{\tilde{\lambda }}_t^s(G)\prod _{t:t\notin W}(1-{\tilde{\lambda }}_t^s(G))$$ (5)

Definition 2

By replacing the traceability ${\lambda }_t$ of each tuple t with ${\lambda }_t^s$, we can obtain the new probability distribution function ${\tilde{\mu }}^{\mathcal{S}}(W)$ in the W-world, the specific definition is shown in equation (6):

$${\tilde{\mu }}^{\mathcal{S}}(W)\stackrel{\text{def}}{=}\sum \tilde{\lambda }(G,W)\prod _{i:g_i\in G}p(g_i)\prod _{j:g_j\in G}(1-p(g_j))$$ (6)

To ensure that the traceability function of each tuple has an -approximation value and the approximate probability is true and reliable, we have the following definition:

Definition 3

As shown in equation (7), for a set of atomic events G, given the probability distribution of G, if

$$E_p{(\widetilde{{\lambda }_{\mathrm{t}}}-{\lambda }_t)}^2\ll \epsilon$$ (7)

then $\widetilde{{\lambda }_{\mathrm{t}}}$ is an approximate value of ${\lambda }_t$, where $E_P$ represents the expected value assigned to atomic events by the probability function P.

Definition 4

The function $\mathrm{IN}{f}_{x_i}({\lambda }_t)$ represents the influence of $x_i$ on ${\lambda }_t$, as shown in equation (8):

$$\text{IN}{f}_{x_i}({\lambda }_t)\stackrel{\text{def}}{=}P({\lambda }_t(G)\ne {\lambda }_t(G\oplus \{i\}))$$ (8)

By the function $\mathrm{IN}{f}_{x_i}({\lambda }_t)$ defined in Definition 4, we can determine the k most influential variables. Taking event $t_5$ as an example, considering the impact of $x_9$ on $t_5$, when $x_8$ is true and $x_{10}$ is false, only $x_9$ can change the value of ${\lambda }_{t_5}$, with a probability of occurrence of 0.08×(1-0.08)=0.0736. It can be seen that the degree of influence of $x_9$ on ${\lambda }_{t_5}$ can be calculated through this method.

5. Experimental verification

To evaluate the feasibility and performance of the proposed green energy traceability mechanism, we developed a system using the Hyperledger Fabric blockchain experimental platform and ran various components of the network using Docker containers. The experimental environment is as follows: CPU: Intel(R) Core(TM) i7-8750H CPU @ 2.20GHz, Memory: 16GB, Hard Disk: 1TB, Virtual Machine Operating System: Ubuntu 16.04.

To test the performance of data on-chain and traceability queries, it is necessary to assess the system's performance under load and stress, measure the latency and throughput of the system when processing large amounts of data, and ensure that the system can handle high-concurrency verification and query requests while maintaining stable and reliable performance. Therefore, in this paper, we added a test plan using the stress testing tool JMeter, set the number of threads in the thread group to 2000 to simulate 2000 users sending requests to test the performance and stability of the target system.

By measuring the latency of data on-chain and traceability queries, we can understand the response speed of the system when processing requests. Latency testing helps determine the response time of the system under different loads, thereby evaluating the real-time performance of the system. By simulating concurrent requests and measuring the number of requests the system can handle per unit time, we can evaluate the throughput of the system. Throughput testing can help determine the system's processing capacity when facing a large number of requests, thereby evaluating the scalability and performance limits of the system.

As shown in Fig. 6, the number of successful requests per second for data on-chain is relatively high, and the number of failed requests is low for most of the time, almost zero. This indicates that the system can efficiently process data on-chain requests for most of the time and rarely encounter errors, demonstrating high stability and reliability. However, there is an increase in the number of failed requests between runtime intervals of 700 ms to 1100 ms and 1500 ms to 1700 ms. This issue may be related to other factors such as resource constraints and network latency, which require further analysis and investigation.

Figure 6.

Throughput of data link.

According to Fig. 7, the average response time of traceability queries is relatively short, with the highest not exceeding 2000 ms. The results indicate that the processing speed of traceability queries is relatively fast, and the system can efficiently and quickly complete traceability query operations. The system is effective in tracing the source of green energy through traceability queries and providing accurate and reliable query results. As shown in Fig. 8, the number of successful traceability query requests per second is relatively high, with very low failed request numbers, almost zero. This indicates that the system exhibits good performance and stability, efficiently processing traceability query requests within the runtime and rarely encountering query failures. It demonstrates high efficiency and reliability in processing query requests.

Figure 7.

Traceability query delay.

Figure 8.

Throughput of traceability queries.

6. Conclusion

This paper proposed ERRE-PM model based on of PROV-DM and block-chain technology. We also designed a green energy traceability mechanism based on smart contracts and the ERRE-PM model. Through data filling and verification techniques, trustworthy green energy data collection is achieved. In the process of green power generation, transmission and distribution, trading, and electricity consumption, the blockchain-based ERRE-PM security model can securely capture traceability records. By extracting the flow time and transaction frequency between various links in the data lineage traceability network, it provides a rational basis for the credibility of key nodes. Through experimental tests on data on-chain and traceability query performance, it is demonstrated the effectiveness of the proposed solution.

Funding statement

This work was supported by State Grid Corporation Technology Project “Research and application of key technologies for coordinated operation of green power trading in the national unified electricity market” (SGSH0000DJJS2310175).

CRediT authorship contribution statement

Shaoyuan Yu: Writing – original draft. Jing Yang: Writing – original draft. Jia Ni: Writing – original draft. Chunyi Chen: Writing – original draft. Tao Yu: Writing – original draft. Ailin Chen: Writing – original draft. Jian Geng: Writing – original draft. Hua Zhong: Writing – original draft. Weihua Weng: Writing – original draft. Tao Wei: Writing – original draft.

Declaration of Competing Interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Data availability

The data that has been used is confidential.