Skip to main content
. 2020 Jan 28;20(3):709. doi: 10.3390/s20030709

Figure 2.

Figure 2

SIMRR multi-mode sensing based on artificial neuron network. (a) Signal processing model using BP-ANN (back-propagation ANN) for the designed SIMRR multi-mode sensor in the Figure 1a. Influence of the network parameters on its measurement results, (b) MSE (mean squared error) versus number of neurons in the hidden layer, (c) MSE versus learning rate, (d) MSE versus number of training datasets, (e) MSE versus training goal error, (f) Comparison between the estimated values of the length change and their corresponding theoretical values.