Vehicle fuel cell service life intelligent prediction method based on multi-feature index fusion

Abstract

The invention relates to the technical field of fuel cells, in particular to an intelligent prediction method for the service life of a vehicle fuel cell based on multi-feature index fusion, which can effectively improve the prediction accuracy of the service life of the fuel cell and further facilitate the improvement of the operation reliability and safety of the fuel cell. According to the multi-feature health index fusion, the output voltage of the fuel cell reconstructs a novel health index representing a fluctuation item, a periodic item and a long-term attenuation component of the fuel cell, and the provided intelligent prediction method for the service life of the vehicle fuel cell integrates various prediction algorithms and optimization algorithms. The method can select a proper prediction method according to the type of input data to improve the accuracy and robustness of life prediction.

Description

Technical Field

[0001] The present invention relates to the field of fuel cell technology, and more specifically to an intelligent prediction method for the life of a vehicle fuel cell based on the fusion of multiple feature indicators, which can effectively improve the accuracy of prediction of the service life of a fuel cell, thereby facilitating improving the reliability and safety of fuel cell operation. Background Art

[0002] In recent years, energy issues have become increasingly important in environmental protection and economic development. On the one hand, fossil energy reserves are limited and cannot be regenerated; on the other hand, the demand for energy in industry and daily life is increasing. Proton Exchange Membrane Fuel Cell (PEMFC), as a highly regarded clean energy technology, has shown broad application prospects due to its advantages such as high energy conversion efficiency, no pollution, simple structural design, stable operation and low noise.

[0003] The PEMFC stack is composed of multiple single cells connected in series. Any degradation in the performance of a single cell will lead to a decrease in the efficiency of the entire stack and cause voltage decay, thus affecting the normal operation of the system. Directly testing the performance of the fuel cell may cause the stack to be damaged or even scrapped. During the long-term use of the fuel cell, the stress differences caused by humidity and temperature changes and insufficient gas supply often cause performance degradation and a decrease in output voltage. Once the proton exchange membrane is damaged, its performance cannot be restored by adjusting the operating conditions, which may cause irreversible damage and ultimately lead to significant economic losses. Therefore, it is crucial to ensure that the PEMFC system is in good condition for its continued operation. By monitoring the system operating parameters in real time and predicting the remaining useful life (RUL) of the fuel cell, preventive maintenance can be carried out before the stack fails completely. This can not only shorten downtime, but also extend the safe use cycle of the system and improve the economic benefits and reliability of operation.

[0004] The existing PEMFC life intelligent prediction algorithm based on voltage as input index has large cumulative errors in long-term prediction, and the deviation between the predicted value and the true value is large; the predicted value cannot reflect the actual attenuation behavior of PEMFC, making it difficult to achieve accurate PEMFC RUL evaluation. The operating data of fuel cells is nonlinear and multidimensional, containing important degradation information of fuel cells. Voltage as a health indicator contains the nonlinear attenuation characteristics and test noise of fuel cells, making it difficult to improve the accuracy of long-term life prediction. Summary of the invention

[0005] In view of the shortcomings and deficiencies in the prior art, the present invention proposes a method for intelligently predicting the life of a vehicle fuel cell based on the fusion of multiple feature indicators, which can have high prediction accuracy and strong robustness under multiple working conditions.

[0006] The present invention is achieved by the following measures: A method for intelligent prediction of vehicle fuel cell life based on fusion of multiple feature indicators, characterized by comprising the following steps: Step 1: Collecting continuous voltage sampling data of the proton exchange membrane fuel cell, and preprocessing the collected data to obtain the stack voltage of the proton exchange membrane fuel cell; Step 2: Use the EMD empirical mode decomposition method to decompose the stack voltage that characterizes the attenuation of the proton exchange membrane fuel cell PEMFC into multiple intrinsic mode functions IMFs of different characteristic scales; Step 3: Establish a weighted attention model to fuse and reorganize the IMF. The weighted attention model reorganizes the IMF into an attenuation component 1 for characterizing the volatility and noise of the fuel cell, an attenuation component 2 for characterizing the periodic change of the fuel cell, and an attenuation component 3 for characterizing the long-term aging trend of the fuel cell; Step 4: Through the intelligent prediction method of vehicle fuel cell life based on the fusion of multi-feature health indicators, the attenuation component one, attenuation component two, and attenuation component three are predicted respectively. Among them, the gray wolf optimization algorithm GWO combined with the gated recurrent unit GRU is used to predict the attenuation component one, the SSA-LSTM prediction method is used to predict the attenuation component two, and the sequence model Transformer based on the attention mechanism is used to predict the attenuation component three.

[0007] The pretreatment in step 1 of the present invention comprises: Step 1-1: Perform Gaussian filtering on the sampled data to filter out noise and abnormal peaks. The Gaussian filtering is expressed by the following formula: (1), (2), (3), In the formula, K(t) represents the standard normal distribution of parameter data at time t, f(t j ) represents the filtered data, u(t j ) represents parameter data, n represents the number of parameter data, and H represents bandwidth; Step 1-2: Use the cubic spline method to perform interpolation to obtain sampling data with uniform time intervals.

[0008] In step 2 of the present invention, voltage is selected as the health indicator. The EMD empirical mode decomposition method decomposes the voltage data into several IMFs and one residual according to the time scale of the voltage itself, wherein the IMF must satisfy two constraints: constraint 1, in the entire data segment, the number of extreme points and the number of zero crossing points need to be equal or differ by at most 1, constraint 2, at any given moment, the upper envelope formed by the local maximum value and the lower envelope formed by the local minimum value are averaged at 0, assuming that the PEMFC voltage is ,in is the voltage in the rth time period, the EMD empirical mode decomposition method includes the following steps: Step 2-1: Find the voltage All extreme points of , and are drawn using cubic spline interpolation The upper and lower envelopes of and ; Step 2-2: The average envelope is extracted from the average of the upper envelope and the lower envelope, expressed as , (4); Step 2-3: Using the original data Subtract the mean Get the data ; Step 2-4: Judgment Does it meet the two constraints of the IMF? If so, is the decomposed IMF, denoted as w(r). If it does not meet the requirement, then according to Re-execute steps 2-1 to 2-4, where the residual term R(r) is as follows: (5); Step 2-5: Repeat steps 2-1 to 2-4 until the residual becomes a monotonic function, the EEMD empirical mode decomposition method is completed, and the voltage data is finally decomposed into k An IMF and a residual, as shown below: (6).

[0009] In step 3 of the present invention, the core of the attention mechanism of the weighted attention model is to calculate the correlation between each pair of data, where the weights are calculated in the form of dot products: (7), In the formula is a learnable weight matrix, is the query vector, is the key vector, Represents the query vector The transpose of the above Normalization is performed through the function, as shown below: (8), The specific steps include: Step 3-1: Based on the attention weight, perform weighted summation on the second set of data to obtain the fused data of each sample in the first set of data, as expressed in the following formula: (9), where each z i All based on The result obtained by correlation fusion with the second set of data; Step 3-2: Concatenate or weight the first set of data and its corresponding fusion representation to generate new data : (10), where ReLu is a nonlinear activation function; Step 3-3: The final output fusion data is expressed as follows: (11), H is the new data set after fusing the two sets of data through the attention mechanism.

[0010] In step 4 of the present invention, when predicting the attenuation component 1, the Grey Wolf Optimization Algorithm GWO combined with the gated recurrent unit GRU is used for prediction, wherein the gating mechanism of GRU is good at modeling medium- and long-term dependencies, and the global search of GWO avoids local optimality and enhances adaptability to periodic data. The GRU model described in step 4 is composed of an update gate and a reset gate, and the calculation formula of the GRU model is: (12), (13), middle z t is the output of the update gate, h t-1 is the hidden state at the previous time step, is the current input data, W z and b z They are the weight matrix and bias respectively. The reset gate controls the influence of the previous hidden state on the current calculation. r t To reset the gate output, W r and b r are the weight matrix and bias respectively, the current candidate hidden state It is calculated based on the hidden state adjusted by the reset gate; GWO achieves global optimization by simulating the hunting behavior of a group of gray wolves. The core of GWO lies in the balance between global search and local utilization of the solution space. The current position of each gray wolf in GWO represents a solution (representing a set of hyperparameters in GRU optimization), and the position update formula is as follows: (14), In the formula X i t+1 For the t+1 The first iteration i The wolf's position vector; X leader t is the position vector of the leader wolf at the tth iteration, i.e., the current optimal solution, A is the control factor for adjusting the step size, and C is the coefficient for adjusting the position of the target prey. A and C are defined as follows: A = 2·a·r1-a (15), C = 2·r2 (16), where a is a factor that decreases with each iteration over time, r1 and r2 are random numbers between 0 and 1, and a, β, and δ are X leader t The optimal solution, suboptimal solution and third optimal solution in . The wolf pack updates its position according to the positions of the three leaders a, β and δ. By constantly approaching the optimal value, each wolf gradually finds the global optimal value. In the process of GWO optimizing the hyperparameters of the GRU model, the position of each wolf represents a set of GRU hyperparameters. For each position, the corresponding GRU model is trained and the fitness function (MSE) is calculated.

[0011] In step 4 of the present invention, when predicting the attenuation component 2, the SSA-LSTM prediction method is adopted, wherein the LSTM model is composed of an input gate, a forget gate and an output gate, and the calculation formula of the LSTM model is expressed by the following formula: (17), (18), (19), In the formula, Represents input data, h t-1 represents the state of the hidden layer at time t-1, , and Represent the calculation results of the input gate, the weight matrix and the bias term respectively, , and Represent the calculation results of the output gate, the weight matrix and the bias term respectively, , and Respectively represent the calculation results of the forget gate, the weight matrix and the bias term, σ represents the activation function; SSA is a swarm intelligence optimization algorithm based on the foraging behavior of sparrows. It optimizes the objective function by simulating the foraging behavior of sparrows. The optimization goal of the LSTM model is to minimize the prediction error. The mean square error MSE is used as the objective function. SSA constructs the initial solution space population by randomly generating the initial sparrow position. Each position Table 1 shows the potential combination of hyperparameters of LSTM in dimension d, expressed as formula (20): (20), In the formula X t Randomly distributed in the solution space X In order to cover a wide range of hyperparameters, for each sparrow position X t Train the corresponding LSTM model and calculate its fitness value X i , which is related to the value of the objective function. The lower the fitness value, the better the hyperparameter combination is in improving the performance of the LSTM model. SSA simulates the behavior of sparrows when foraging, including foragers (exploratory behavior) and predators (exploitative behavior). The position update rule is based on the following formula: (twenty one), In the formula X i t Indicates t In the iteration i The position of a sparrow, X j t and X k t are the positions of the best solution and the worst solution in the current population, respectively. α and β are adjustment factors. SSA optimizes the current solution by performing a local search near the current solution. The method is to update the parameters by adjusting the position of the parameters in the formula. The local search is expressed by the following formula: (22), where X best t is the current optimal solution, γ is the local search factor, rand is a random number. SSA explores other areas of the solution space through random perturbations and global update strategies to avoid falling into local optimality. The specific formula is as follows: (twenty three), In the formula X global tis the global optimal solution, δ is the global search factor, SSA gradually converges during multiple iterations, and by continuously updating the position of the solution, the group gradually approaches the global optimal solution. The convergence condition is usually based on the stability of the fitness value or the preset number of iterations. Finally, the optimized LSTM model is better than the unoptimized model in terms of the objective function. Based on the above mathematical principles, SSA can efficiently optimize the hyperparameters of the LSTM model and improve its prediction performance in PEMFC fluctuation terms.

[0012] In step 4 of the present invention, when predicting the attenuation component three, the sequence model Transformer based on the attention mechanism is used for prediction. The position encoding vector added to the input of the Transformer encoding and decoding can represent the position of the current data and the distance between different data. The position encoding formula is expressed by CPE: (twenty four), Where C pos Represents the position of the current data in the sample, and d represents the dimension of the position encoding vector. i Represents the position index of each value, sine encoding is used for even positions and cosine encoding is used for odd positions. The attention mechanism receives the input or output of the previous encoder and multiplies the received data by different weights to obtain the Q, K, and V matrices. The similarity between data is shown in the following formula: (25), Where Q is the query matrix, K is the key matrix, QK T Calculate the attention weight of Q on V. d k is the dimension of the K matrix, which is used to normalize the attention mechanism. Softmax is a normalized exponential function; set up P(X) is the label sequence, Q(X) is the output sequence predicted by Transformer, and KL divergence is an indicator to measure the similarity of two probability distributions. Its calculation formula is as follows: (26), Where q(x i ) is the i-th predicted value in the Transformer predicted output sequence, p(x i ) is the label sequence q(x i ) is the predicted value at the corresponding moment, l is the sequence length, is the KL divergence loss function of two sequences P(X) and Q(X), The smaller it is, the closer the two data distributions are. By repeatedly training the neural network, the distribution of Q(X) can be approximated to P(X).

[0013] Compared with the prior art, the present invention integrates multiple feature health indicators to reconstruct the output voltage of the fuel cell into a new health indicator that characterizes the fuel cell's fluctuation term, periodic term and long-term attenuation component. The proposed intelligent prediction method for the life of automotive fuel cells integrates a variety of prediction algorithms and optimization algorithms, which can select a suitable prediction method according to the input data type to improve the accuracy and robustness of life prediction. Through verification based on real fuel cell aging data, the results show that the method proposed in the present invention has high accuracy and strong robustness in predicting the remaining service life under multiple working conditions, and its predicted value can reflect the actual attenuation behavior of the fuel cell. BRIEF DESCRIPTION OF THE DRAWINGS

[0014] Attached Figure 1 It is a flow chart of the present invention.

[0015] Attached Figure 2 It is a flow chart of an embodiment of the present invention.

[0016] Attached Figure 3 It is a static working condition prediction result diagram in an embodiment of the present invention.

[0017] Attached Figure 4 It is a diagram of prediction results under quasi-dynamic conditions in an embodiment of the present invention.

[0018] Attached Figure 5 4 is a flow chart of attenuation component three in an embodiment of the present invention. DETAILED DESCRIPTION

[0019] The present invention will be further described below in conjunction with the accompanying drawings and embodiments.

[0020] Embodiment: This example proposes a method as shown in the attached Figure 2 And attached Figure 5 The intelligent prediction method for the life of a fuel cell for a vehicle based on the fusion of multiple feature indicators shown includes the following steps: Step S1: Normalize the collected fuel cell voltage at continuous time and map it to the interval [0,1]. In order to reduce the interference of clutter data, pre-process the sampled data through Gaussian filtering, including the following steps: Step S11: performing Gaussian filtering on the sampled data to filter out noise and abnormal peaks; Step S12: Using the cubic spline method to perform interpolation, obtain sampling data with uniform time intervals. The Gaussian filter formula is: (1), (2), (3), In the formula, K(t) represents the standard normal distribution of parameter data at time t, f(t j) represents the filtered data, u(t j ) represents parameter data, n represents the number of parameter data, and H represents bandwidth; Step S2: Empirical Mode Decomposition (EMD) is used to decompose the voltage decay characteristics of the fuel cell stack into multiple intrinsic mode functions (IMFs) of different characteristic scales. The fuel cell prediction algorithm selects voltage as the health indicator. The EMD decomposition algorithm decomposes the voltage data into several IMFs and one residual according to the time scale of the voltage itself. The IMF must meet two constraints: (1) In the entire data segment, the number of extreme points and the number of zero crossing points need to be equal or differ by at most 1, and (2) At any given moment, the upper envelope formed by the local maximum value and the lower envelope formed by the local minimum value are averaged at 0. In this example, the PEMFC voltage is assumed to be ,in is the voltage in the rth time period, and the EMD decomposition process is as follows: Step S21: Find the voltage All extreme points of , and are drawn using cubic spline interpolation The upper and lower envelopes of and ; Step S22: The average envelope is extracted from the average of the upper envelope and the lower envelope, expressed as , (4); Step 2-3: Using the original data Subtract the mean Get the data ; Step 2-4: Judgment Does it meet the two constraints of the IMF? If so, is the decomposed IMF, denoted as w(r). If it does not meet the requirement, then according to Re-execute steps 2-1 to 2-4, where the residual term R(r) is as follows: (5), Step 2-5: Repeat steps 2-1 to 2-4 until the residual becomes a monotonic function, the EEMD empirical mode decomposition method is completed, and the voltage data is finally decomposed into k An IMF and a residual, as shown below: (6) In step 3 of the present invention, the core of the attention mechanism of the weighted attention model is to calculate the correlation between each pair of data, where the weights are calculated in the form of dot products: (7), In the formula is a learnable weight matrix, is the query vector, is the key vector, Represents the query vector The transpose of Normalization is performed through the function, as shown below: (8), The specific steps include: Step S31: Based on the attention weight, perform weighted summation on the second set of data to obtain the fused data representation of each sample in the first set of data: (9), In the formula, each All based on The result obtained by correlation fusion with the second set of data; Step S32: concatenate or weight the first set of data and its corresponding fusion representation to generate new data As shown below: (10), Where ReLu is a nonlinear activation function; Step S33: The final output fusion data is represented as follows; (11), H is the new data set after fusing the two sets of data through the attention mechanism; Step S4: The weighted attention model reorganizes the IMF into an attenuation component 1 that characterizes the volatility and noise of the fuel cell, an attenuation component 2 that characterizes the periodic change of the fuel cell, and an attenuation component 3 that characterizes the long-term aging trend of the fuel cell; The intelligent prediction method for the life of automotive fuel cells includes the gated recurrent unit (GRU), long short-term memory network (LSTM), and Transformer. The optimization algorithms are the Sparrow Search Algorithm (SSA) and the Grey Wolf Optimizer (GWO). Step S41: Based on the SSA-LSTM prediction method, since it can effectively capture short-term fluctuations and nonlinear high-frequency information, SSA-LSTM is used to predict the attenuation component 1 of PEMFC. Among them, the LSTM model consists of an input gate, a forget gate, and an output gate, and its calculation formula is: (17), (18), (19), In the formula, Represents input data, represents the state of the hidden layer at time t-1, , and Represent the calculation results of the input gate, the weight matrix and the bias term respectively, , and Represent the calculation results of the output gate, the weight matrix and the bias term respectively, , and They represent the calculation results of the forget gate, the weight matrix and the bias term respectively, and σ represents the activation function; Step S42: SSA is a swarm intelligence optimization algorithm based on the foraging behavior of sparrows. Its mathematical principle is to optimize the objective function by simulating the foraging behavior of sparrows. The optimization goal of the LSTM model is to minimize the prediction error, and the mean square error (MSE) is usually used as the objective function.

[0021] SSA constructs the initial solution space population by randomly generating initial sparrow positions. X t represents the potential combination of LSTM hyperparameters in dimension d, (20), In the formula X t Randomly distributed in the solution space X to cover a wide range of hyperparameters. For each sparrow position X t Train the corresponding LSTM model and calculate its fitness value X i , which is related to the value of the objective function. The lower the fitness value, the better the hyperparameter combination is in improving the performance of the LSTM model. SSA simulates the behavior of sparrows when foraging, including foragers (exploratory behavior) and predators (exploitative behavior). The position update rule is based on the following formula: (twenty one), In the formula X i t Indicates t In the iteration i The position of a sparrow, Xj t and X k t are the positions of the best solution and the worst solution in the current population, respectively, and α and β are adjustment factors.

[0022] SSA optimizes the current solution by performing a local search near the current solution by updating the parameters by adjusting the position of the parameters in the formula. The local search can be expressed as follows: (twenty two), in, X best t is the current optimal solution, γ is the local search factor, rand is a random number. SSA explores other areas of the solution space through random perturbations and global update strategies to avoid falling into local optimality. The specific formula is as follows: (twenty three), In the formula X global t is the global optimal solution, and δ is the global search factor. SSA gradually converges during multiple iterations, and by continuously updating the position of the solution, the group gradually approaches the global optimal solution. The convergence condition is usually based on the stability of the fitness value or the preset number of iterations. In the end, the optimized LSTM model outperforms the unoptimized model in terms of the objective function. Based on the above mathematical principles, SSA can efficiently optimize the hyperparameters of the LSTM model and improve its prediction performance in PEMFC fluctuation terms.

[0023] Step S43: The gating mechanism of GRU is good at modeling medium- and long-term dependencies, while the global search of GWO avoids local optimality and enhances adaptability to periodic data. GWO-GRU is very suitable for predicting the decay component of PEMFC periodic terms2.

[0024] Preferably, the LSTM model described in step S4 is composed of an update gate and a reset gate, and its calculation formula is: (12), (13), Where z t is the output of the update gate, h t-1 is the hidden state at the previous time step, is the current input data, W z and b zare the weight matrix and the bias respectively. The reset gate controls the influence of the previous hidden state on the current calculation. t To reset the gate output, W r and b r are weight matrices and biases respectively. Current candidate hidden state It is calculated based on the hidden state adjusted by the reset gate.

[0025] Step S44: GWO achieves global optimization by simulating the hunting behavior of a group of gray wolves. The core of GWO lies in the balance between global search and local utilization of the solution space. The current position of each gray wolf in GWO represents a solution (representing a set of hyperparameters in GRU optimization). The position update formula is as follows: (14), In the formula X t+1 i For the t+1 The first iteration i The wolf's position vector; X leader t is the position vector of the leader wolf at the tth iteration, i.e., the current optimal solution. A is the control factor for adjusting the step length, and C is the coefficient for adjusting the position of the target prey. A and C are defined as follows: (15), (16), Where a is a factor that decreases with each iteration over time, and r1 and r2 are random numbers between 0 and 1. a, β, and δ are X leader t The optimal solution, suboptimal solution, and third optimal solution in . The wolf pack updates its position according to the positions of the three leaders a, β, and δ. By constantly approaching the optimal value, each wolf gradually finds the global optimal value. In the process of GWO optimizing the hyperparameters of the GRU model, the position of each wolf represents a set of GRU hyperparameters. For each position, the corresponding GRU model is trained and the fitness function (MSE) is calculated.

[0026] Step S45: Transformer has good feature extraction capability, which is very suitable for predicting the decay component 3 representing the long-term decay of PEMFC.

[0027] Preferably, the position encoding vector added to the input by encoding and decoding in the Transformer described in step S4 can represent the position of the current data and the distance between different data. The position encoding formula is represented by CPE: (twenty four), In the formula C pos Represents the position of the current data in the sample, and d represents the dimension of the position encoding vector. i Represents the position index of each value, sine encoding is used for even positions and cosine encoding is used for odd positions. The attention mechanism receives the input or output of the previous encoder and multiplies the received data by different weights to obtain the Q, K, and V matrices. The similarity between data is shown in the following formula: (25), Where Q is the query matrix, K is the key matrix, QK T Calculate the attention weight of Q on V. d k is the dimension of the K matrix, which is used to normalize the attention mechanism. Softmax is a normalized exponential function.

[0028] set up P(X) is the label sequence, Q(X) is the output sequence predicted by Transformer. KL divergence is an indicator to measure the similarity of two probability distributions, and its calculation formula is as follows: (26), In the formula For Transformer, predict the i-th predicted value in the output sequence. For the label sequence The predicted value at the corresponding moment. l is the sequence length. is the KL divergence loss function of the two sequences P(X) and Q(X). The smaller it is, the closer the two data distributions are. By repeatedly training the neural network, the distribution of Q(X) can be approximated to P(X).

[0029] In order to further demonstrate the accuracy of the prediction of PEMFC attenuation trend under different test conditions in this example, the calculation is performed using a real PEMFC aging data set. EMD adaptively decomposes the voltage of the PEMFC into 5 IMFs and a residual value according to the frequency components of the signal. These 5 IMFs represent different frequency bands, which usually decrease from high frequency to low frequency. The first IMF usually represents the high-frequency component in the signal, that is, the most rapidly changing part. It usually contains fast oscillations or noise in the signal. The second IMF represents the lower-frequency component in the signal, that is, the slower-changing part, but still contains some higher-frequency dynamic features. The third IMF further represents the lower-frequency oscillating component, which is usually the slower-changing part of the signal and may be related to some periodic changes or trends. The fourth IMF represents the lower-frequency component, which may be a slow-changing trend, periodic change, or amplitude change in the signal. The fifth IMF: usually represents the low-frequency component of the signal, which may be related to the overall trend or baseline drift of the signal. If the signal contains a large periodic trend or low-frequency change, the final IMF will contain these components. The residual term is usually a low-frequency trend component or the average value of the signal. This residual term has no obvious high-frequency component and usually represents the long-term trend or slow-changing part of the signal.

[0030] SSA-LST, GWO-GRU and Echo State Network (ESN) were selected as comparison methods. The four methods were compared using the first 20% of the real fuel cell data for training data. The root mean square error (RMSE) and mean absolute percentage error (MAPE) of the prediction within the entire life cycle of the fuel cell were selected for analysis. The calculation formula is: (27), (28), In the formula, is the predicted value, y i Represents the real data value, n is the number of tests, and N is a natural number.

[0031] The comparative experimental results under static and dynamic conditions of PEMFC are shown in Table 1: Table 1

[0032] It can be seen from the experimental results in Table 1 that the RMSE and MAE of this example are smaller than those of the comparison method under the same data set and prediction step size. The experimental results in Table 1 prove that the intelligent prediction method for vehicle fuel cell life based on multi-feature health index fusion proposed by the present invention has high prediction accuracy and robustness under different working conditions.

[0033] In order to further demonstrate the accuracy of the prediction of the remaining useful life of PEMFC under different test conditions, the performance is verified by using a real PEMFC aging data set. The output of the prediction method proposed in the present invention is the time T and the PEMFC voltage value at that moment. The prediction error calculation formula for defining the remaining useful life (RUL) of PEMFC is as follows: (29), In the formula is the real RUL, is the predicted RUL.

[0034] The prediction results are shown in Table 2: Table 2

[0035] As shown in Table 2 and Appendix Figure 3 , Figure 4 As shown in the figure, the prediction values ​​of ESN, LSTM and BiLSTM in static conditions cannot decay to the fault voltage, and the prediction values ​​of the three prediction methods are obviously distorted. Therefore, the three methods cannot decay the PEMFC in static conditions with limited data. The method proposed in the present invention can accurately predict the RUL of PEMFC in both static and quasi-dynamic conditions, proving the effectiveness of the method proposed in the present invention.

Claims

1. An intelligent prediction method for vehicle fuel cell life based on multi-feature index fusion, characterized in that: The following steps are involved: Step 1: Collecting continuous voltage sampling data of the proton exchange membrane fuel cell, and preprocessing the collected data to obtain the stack voltage of the proton exchange membrane fuel cell; Step 2: Use the EMD empirical mode decomposition method to decompose the stack voltage that characterizes the attenuation of the proton exchange membrane fuel cell PEMFC into multiple intrinsic mode functions IMFs of different characteristic scales; Step 3: Establish a weighted attention model to fuse and reorganize the IMF. The weighted attention model reorganizes the IMF into an attenuation component 1 for characterizing the volatility and noise of the fuel cell, an attenuation component 2 for characterizing the periodic change of the fuel cell, and an attenuation component 3 for characterizing the long-term aging trend of the fuel cell; Step 4: Through the intelligent prediction method of vehicle fuel cell life based on the fusion of multi-feature health indicators, the attenuation component one, attenuation component two, and attenuation component three are predicted respectively. Among them, the gray wolf optimization algorithm GWO combined with the gated recurrent unit GRU is used to predict the attenuation component one, the SSA-LSTM prediction method is used to predict the attenuation component two, and the sequence model Transformer based on the attention mechanism is used to predict the attenuation component three.

2. The intelligent prediction method for vehicle fuel cell life based on multi-feature index fusion according to claim 1 is characterized in that: The pretreatment in step 1 includes: Step 1-1: Perform Gaussian filtering on the sampled data to filter out noise and abnormal peaks. The Gaussian filtering is expressed by the following formula: （1）， （2）， (3), where K(t) represents the standard normal distribution of parameter data at time t, and f(t j ) represents the filtered data, u(t j ) represents parameter data, n represents the number of parameter data, and H represents bandwidth; Step 1-2: Use the cubic spline method to perform interpolation to obtain sampling data with uniform time intervals.

3. The intelligent prediction method for vehicle fuel cell life based on multi-feature index fusion according to claim 1 is characterized in that: In step 2, voltage is selected as the health indicator. The EMD empirical mode decomposition method decomposes the voltage data into several IMFs and 1 residual according to the time scale of the voltage itself. The IMF must meet two constraints: Constraint 1: In the entire data segment, the number of extreme points and the number of zero crossings need to be equal or differ by at most 1. Constraint 2: At any given moment, the upper envelope formed by the local maximum and the lower envelope formed by the local minimum are averaged at 0. Let the PEMFC voltage be ,in is the voltage in the rth time period, the EMD empirical mode decomposition method includes the following steps: Step 2-1: Find the voltage All extreme points of , and are drawn using cubic spline interpolation The upper and lower envelopes of and ; Step 2-2: The average envelope is extracted from the average of the upper envelope and the lower envelope, expressed as , （4）； Step 2-3: Using the original data Subtract the mean Get the data ; Step 2-4: Judgment Does it meet the two constraints of the IMF? If so, is the decomposed IMF, denoted as w(r). If it does not meet the requirement, then according to Re-execute steps 2-1 to 2-4, where the residual term R(r) is as follows: （5）； Step 2-5: Repeat steps 2-1 to 2-4 until the residual becomes a monotonic function, the EEMD empirical mode decomposition method ends, and the voltage data is finally decomposed into k An IMF and a residual, as shown below: （6）。 4. The intelligent prediction method for vehicle fuel cell life based on multi-feature index fusion according to claim 1 is characterized in that: In step 3, the core of the attention mechanism of the weighted attention model is to calculate the correlation between each pair of data, where the weights are calculated in the form of dot products: (7), where is a learnable weight matrix, is the query vector, is the key vector, Represents the query vector The transpose of Normalization is performed through the function, as shown below: (8) specifically comprises the following steps: Step 3-1: Based on the attention weight, perform weighted summation on the second set of data to obtain the fused data of each sample in the first set of data, as expressed in the following formula: (9), where each All based on The result obtained by correlation fusion with the second set of data; Step 3-2: Concatenate or weight the first set of data and its corresponding fusion representation to generate new data : (10), where ReLu is a nonlinear activation function; Step 3-3: The final output fusion data is expressed as follows: (11), H is the new data set after fusing the two sets of data through the attention mechanism.

5. The intelligent prediction method for vehicle fuel cell life based on multi-feature index fusion according to claim 1 is characterized in that: In step 4, when predicting the attenuation component 1, the Grey Wolf Optimization Algorithm GWO combined with the gated recurrent unit GRU is used for prediction, wherein the gating mechanism of the gated recurrent unit GRU is good at modeling medium- and long-term dependencies, and the global search of GWO avoids local optimality and enhances adaptability to periodic data. The gated recurrent unit GRU described in step 4 is composed of an update gate and a reset gate. The calculation formula of the gated recurrent unit GRU is: (12), (13), where is the output of the update gate, is the hidden state at the previous time step, is the current input data, and They are the weight matrix and bias respectively. The reset gate controls the influence of the previous hidden state on the current calculation. To reset the gate output, and are the weight matrix and bias respectively, the current candidate hidden state It is calculated based on the hidden state adjusted by the reset gate; GWO achieves global optimization by simulating the hunting behavior of a group of gray wolves. The core of GWO lies in the balance between global search and local utilization of the solution space. The current position of each gray wolf in GWO represents a solution, and the position update formula is as follows: (14), where For the t+1 The first iteration i The wolf's position vector; is the position vector of the leading wolf at the tth iteration, that is, the current optimal solution, A is the control factor for adjusting the step size, and C is the coefficient for adjusting the position of the target prey. A and C are defined as follows: A = 2·a·r1-a (15), C = 2·r2 (16), where a is a factor that decreases with each iteration over time, r1 and r2 are random numbers between 0 and 1, and α, β, and δ are The wolves update their positions according to the positions of the three leaders α, β and δ. By continuously approaching the optimal value, each wolf gradually finds the global optimal value. In the process of GWO optimizing the hyperparameters of the GRU model, the position of each wolf represents a set of GRU hyperparameters. For each position, the corresponding GRU model is trained and the fitness function MSE is calculated.

6. The intelligent prediction method for vehicle fuel cell life based on multi-feature index fusion according to claim 1 is characterized in that: In step 4, when predicting the attenuation component 2, the SSA-LSTM prediction method is used, where the LSTM model consists of an input gate, a forget gate, and an output gate. The calculation formula of the LSTM model is expressed as follows: （17）， （18）， （19）， In the formula, Represents input data, represents the state of the hidden layer at time t-1, , and Represent the calculation results of the input gate, the weight matrix and the bias term respectively, , and Represent the calculation results of the output gate, the weight matrix and the bias term respectively, , and Respectively represent the calculation results of the forget gate, the weight matrix and the bias term, represents the activation function; SSA is a swarm intelligence optimization algorithm based on the foraging behavior of sparrows. It optimizes the objective function by simulating the foraging behavior of sparrows. The optimization goal of the LSTM model is to minimize the prediction error. The mean square error MSE is used as the objective function. SSA constructs the initial solution space population by randomly generating the initial sparrow position. Each position represents the potential combination of hyperparameters of LSTM in dimension d, expressed as formula (20): （20）， In the formula Randomly distributed in the solution space X to cover a wide range of hyperparameters, for each sparrow position Train the corresponding LSTM model and calculate its fitness value , which is related to the value of the objective function. The lower the fitness value, the better the hyperparameter combination is in improving the performance of the LSTM model. SSA simulates the foraging behavior of sparrows, including foragers, i.e., exploration behavior, and predators, i.e., exploitation behavior. The position update rule is based on the following formula: （21）， In the formula Indicates t In the iteration i The position of a sparrow, and are the positions of the best solution and the worst solution in the current population, respectively. α and β are adjustment factors. SSA optimizes the current solution by performing a local search near the current solution. The method is to update the parameters by adjusting the position of the parameters in the formula. The local search is expressed by the following formula: （22）， in, X best t is the current optimal solution, γ is the local search factor, rand is a random number. SSA explores other areas of the solution space through random perturbations and global update strategies to avoid falling into local optimality. The specific formula is as follows: （23）， In the formula X global t is the global optimal solution, δ As the global search factor, SSA gradually converges during multiple iterations, and by continuously updating the position of the solution, the group gradually approaches the global optimal solution.

7. The intelligent prediction method for vehicle fuel cell life based on multi-feature index fusion according to claim 1 is characterized in that: In step 4, when predicting the attenuation component three, the sequence model Transformer based on the attention mechanism is used for prediction. The position encoding vector added to the input of encoding and decoding in the Transformer indicates the position of the current data and the distance between different data. The position encoding formula is expressed by CPE: （24）， In the formula C pos represents the position of the current data in the sample, d represents the dimension of the position encoding vector, i Represents the position index of each value. Even positions are encoded by sine and odd positions are encoded by cosine. The attention mechanism receives the input or output of the previous encoder and multiplies the received data by different weights to obtain the Q, K, and V matrices. The similarity between the data is shown in the following formula: （25）， Where Q is the query matrix, K is the key matrix, QK T Calculate the attention weight of Q on V, d k is the dimension of the K matrix, which is used to normalize the attention mechanism, and softmax is the normalized exponential function; set up P(X) is the label sequence, Q(X) is the output sequence predicted by Transformer, and KL divergence is an indicator to measure the similarity of two probability distributions. Its calculation formula is as follows: （26）， In the formula For Transformer, predict the i-th predicted value in the output sequence. For the label sequence The predicted value at the corresponding moment, l is the sequence length, is the KL divergence loss function of two sequences P(X) and Q(X), The smaller it is, the closer the two data distributions are. By repeatedly training the neural network, the distribution of Q(X) is approximated to P(X).

Images