SOH (state of health) estimation method and system for valve-regulated lead-acid storage battery of transformer substation and storage medium
By measuring the electrochemical impedance spectrum and building an equivalent circuit model, and optimizing the support vector regression model with simulated annealing algorithm, the problem of long-term detection of lead-acid battery in the substation is solved, and the rapid and resource-saving health status estimation is achieved, and the system stability is improved sex.
Patent Information
- Application Number
- CN202411341679.0
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-09-25
- Publication Date
- 2024-11-08
AI Technical Summary
In the prior art, the health status detection of the lead-acid battery in the substation takes a long time and consumes a lot of resources, making it difficult to effectively improve the utilization efficiency and maintenance stability of the lead-acid battery.
By measuring the electrochemical impedance spectrum data, an equivalent circuit model of lead-acid battery based on electrochemical impedance spectrum is constructed, the parameters are fitted using the least squares method, and the support vector regression model is optimized by using a simulated annealing algorithm to estimate the health status of lead-acid battery. .
It realizes rapid and convenient estimation of the health status of lead-acid batteries, reduces energy and time consumption, and improves the utilization efficiency and system stability of lead-acid batteries.
Smart Images

Figure CN118914905A_ABST
Abstract
Description
Technical Field
[0001] The invention belongs to the technical field of battery detection, and relates to a method, system and storage medium for estimating the SOH of a valve-controlled lead-acid battery in a transformer substation. Background Art
[0002] Lead-acid batteries are widely used in electric vehicles, emergency power supplies, communication base stations, uninterruptible power supply systems, solar and wind power generation systems, etc., due to their outstanding cost-effectiveness, excellent reliability and superior performance advantages. In the power system, in order to prevent sudden power outages, valve-regulated lead-acid batteries are generally used as the main component of the DC power supply system of the substation. The batteries will provide emergency power for important equipment in the station during power outages until the power supply of the grid is restored to normal. The lead-acid batteries in the substation system are often used in series and are in a floating charge state for a long time. If there are degraded batteries in the series-connected lead-acid battery group, it will not only reduce the life of the entire lead-acid battery group, but may even cause serious accidents such as explosions.
[0003] At present, the maximum energy storage capacity of lead-acid batteries in substations can only be measured by charging and discharging every two months, which is time-consuming and requires a lot of manpower and material resources, and the entire batch of batteries needs to be replaced every two years. Therefore, it is necessary to find a more convenient and fast method to evaluate the health status (State of Health, SOH) of lead-acid batteries, effectively improve the utilization efficiency of lead-acid batteries, and maintain the stability of substation systems. Summary of the invention
[0004] The technical solution of the present invention is used to solve the problems of excessively long time and high resource consumption when detecting the health status of lead-acid batteries in substations in the prior art.
[0005] The present invention solves the above technical problems through the following technical solutions:
[0006] The present invention provides a method for estimating SOH of a valve-controlled lead-acid battery in a substation, comprising the following steps:
[0007] S1. charging or discharging the battery to a specified state of charge and measuring electrochemical impedance spectroscopy data;
[0008] S2. Construct an equivalent circuit model of a lead-acid battery based on electrochemical impedance spectroscopy;
[0009] S3, using the least squares method to fit each parameter in the equivalent circuit model, and selecting the constant phase element Q with the strongest correlation with the battery SOH as the characteristic parameter;
[0010] S4. A support vector regression model optimized by a simulated annealing algorithm is established, with the parameter value of the constant phase element Q in the equivalent circuit model as the model input and the estimated value of the health state of the battery as the model output, thereby obtaining the estimated value of the SOH of the lead-acid battery.
[0011] Furthermore, the lead-acid battery equivalent circuit model based on electrochemical impedance spectroscopy described in step S2 includes: an inductor L, an ohmic resistor R1, a charge transfer internal resistor R2, a charge diffusion resistor R3, a constant phase element Q, and a capacitive reactance element C; the inductor L, the ohmic resistor R1, the charge transfer internal resistor R2, and the charge diffusion resistor R3 are connected in series in sequence, the constant phase element Q is connected in parallel at both ends of the charge transfer internal resistor R2, and the capacitive reactance element C is connected in parallel at both ends of the charge diffusion resistor R3.
[0012] Furthermore, the inductor L represents the straight line portion of the high-frequency region of the electrochemical impedance spectrum curve, the ohmic resistor R1 is at the point where the high-frequency region and the medium-frequency region of the electrochemical impedance spectrum curve are connected, the charge transfer internal resistor R2 is connected in parallel with the constant phase element Q to represent the irregular semicircular portion of the impedance spectrum in the medium-frequency region of the electrochemical impedance spectrum curve, and the capacitive reactance element C is connected in parallel with the charge diffusion resistor R3 to characterize the diffusion portion in the low-frequency region.
[0013] Furthermore, the fitting formula of the constant phase element Q in step S3 is: Among them, Y0 is the constant phase element parameter, the dimension is Ω -1 cm -2 ·s -n , always positive; j represents the sign of the imaginary part; ω represents the angular frequency; n represents the constant phase element index, which is 0 <n<1,Z Q It represents the impedance of the constant phase element Q.
[0014] Furthermore, the method for establishing the support vector regression model optimized by the simulated annealing algorithm described in step S4 is as follows:
[0015] S4.1. Define the objective function as follows:
[0016]
[0017] Among them, y i is the i-th sample value, is the simulated value of the i-th sample, C represents the penalty factor, g represents the kernel function parameter, and ε represents the insensitivity coefficient;
[0018] S4.2. Use the simulated annealing algorithm to optimize the parameters (C, g, ε). The specific steps are as follows:
[0019] S4.2.1. Initialize parameters; select support vector regression model training samples and prediction samples, perform sample normalization, set the penalty factor C, kernel function parameter g and search space of insensitivity coefficient ε; set the simulated annealing algorithm initial temperature T0, temperature cooling coefficient q, termination temperature T end and the length of the Metropolis chain L; determine the function in step S4.1 as the optimization target fitness function;
[0020] S4.2.2, perform global extreme value optimization;
[0021] S4.2.3. Obtain the optimal solution, that is, the parameter (C, g, ε) value, and use the parameter (C, g, ε) as the optimal learning parameter of the support vector regression model to predict the prediction sample.
[0022] The present invention also provides a substation valve-controlled lead-acid battery SOH estimation system, comprising the following steps: an electrochemical impedance spectrum data measurement module, an equivalent circuit model construction module, a characteristic parameter determination module, and a model prediction module;
[0023] The electrochemical impedance spectroscopy data measurement module is used to charge or discharge the battery to a specified state of charge and measure electrochemical impedance spectroscopy data;
[0024] The equivalent circuit model building module is used to build an equivalent circuit model of a lead-acid battery based on electrochemical impedance spectroscopy;
[0025] The characteristic parameter determination module adopts the least square method to fit each parameter in the equivalent circuit model, and selects the constant phase element Q with the strongest correlation with the battery SOH as the characteristic parameter;
[0026] The model prediction module establishes a support vector regression model optimized by a simulated annealing algorithm, takes the parameter value of the constant phase element Q in the equivalent circuit model as the model input, and the estimated value of the health state of the battery as the model output, thereby obtaining the estimated value of the SOH of the lead-acid battery.
[0027] Furthermore, the lead-acid battery equivalent circuit model based on electrochemical impedance spectroscopy described in the equivalent circuit model construction module includes: an inductor L, an ohmic resistor R1, a charge transfer internal resistor R2, a charge diffusion resistor R3, a constant phase element Q, and a capacitive reactance element C; the inductor L, the ohmic resistor R1, the charge transfer internal resistor R2, and the charge diffusion resistor R3 are connected in series in sequence, the constant phase element Q is connected in parallel at both ends of the charge transfer internal resistor R2, and the capacitive reactance element C is connected in parallel at both ends of the charge diffusion resistor R3; the inductor L represents the straight line portion of the high-frequency region of the electrochemical impedance spectrum curve, the ohmic resistor R1 is at the point where the high-frequency region and the medium-frequency region of the electrochemical impedance spectrum curve are connected, the charge transfer internal resistor R2 and the constant phase element Q are connected in parallel to represent the irregular semicircular portion of the impedance spectrum in the medium-frequency region of the electrochemical impedance spectrum curve, and the capacitive reactance element C and the charge diffusion resistor R3 are connected in parallel to characterize the diffusion portion in the low-frequency region.
[0028] Furthermore, the fitting formula of the constant phase element Q described in the characteristic parameter determination module is: Among them, Y0 is the constant phase element parameter, the dimension is Ω -1 cm -2 ·s -n , always positive; j represents the sign of the imaginary part; ω represents the angular frequency; n represents the constant phase element index, which is 0 <n<1,Z Q It represents the impedance of the constant phase element Q.
[0029] Furthermore, the method for establishing the support vector regression model optimized by the simulated annealing algorithm described in the model prediction module is as follows:
[0030] (1) The objective function is defined as follows:
[0031]
[0032] Among them, y i is the i-th sample value, is the simulated value of the i-th sample, C represents the penalty factor, g represents the kernel function parameter, and ε represents the insensitivity coefficient;
[0033] (2) Use the simulated annealing algorithm to optimize the parameters (C, g, ε). The specific steps are as follows:
[0034] 1) Initialize parameters; select support vector regression model training samples and prediction samples, perform sample normalization, set the penalty factor C, kernel function parameter g and search space of insensitivity coefficient ε; set the simulated annealing algorithm initial temperature T0, temperature cooling coefficient q, termination temperature T end and the length of the Metropolis chain L; determine the function in step (1) as the optimization target fitness function;
[0035] 2) Perform global extreme value optimization;
[0036] 3) Obtain the optimal solution, that is, the parameter (C, g, ε) value, and use the parameter (C, g, ε) as the optimal learning parameter of the support vector regression model to predict the prediction sample.
[0037] A storage medium stores a computer program, which, when executed by a processor, executes the steps of the above-mentioned method for estimating SOH of a valve-regulated lead-acid battery in a substation.
[0038] The advantages of the present invention are:
[0039] The present invention establishes an equivalent circuit model, fits the equivalent circuit model with the electrochemical impedance spectrum data of the battery, uses the obtained characteristic parameters to train a support vector regression model optimized by a simulated annealing algorithm, and establishes a connection between the impedance spectrum fitting data and the battery health state. Thus, the SOH of the lead-acid battery can be estimated directly by measuring the electrochemical impedance spectrum, thereby avoiding multiple charge and discharge cycles of the battery, reducing energy consumption, and saving energy and time. BRIEF DESCRIPTION OF THE DRAWINGS
[0040] Figure 1 This is a flow chart of a method for estimating SOH of a valve-controlled lead-acid battery in a substation according to a first embodiment of the present invention;
[0041] Figure 2 Schematic diagram of an equivalent circuit model of a method for estimating SOH of a valve-regulated lead-acid battery in a substation according to Embodiment 1 of the present invention;
[0042] Figure 3 This is an algorithm flow chart of the SA-SVR model used in the SOH estimation method for valve-controlled lead-acid batteries in substations according to the first embodiment of the present invention;
[0043] Figure 4 This is a training and testing flow chart of the SA-SVR model used in the substation valve-regulated lead-acid battery SOH estimation method of the first embodiment of the present invention;
[0044] Figure 5 This is a schematic diagram of an electrochemical impedance spectrum obtained by discharging the first battery after constant capacity at 10% SOC and measuring the electrochemical impedance spectrum in the SOH estimation method of the valve-regulated lead-acid battery in the substation according to the first embodiment of the present invention;
[0045] Figure 6 A schematic diagram of an electrochemical impedance spectrum obtained by discharging the second battery after constant capacity at 10% SOC and measuring the electrochemical impedance spectrum in the SOH estimation method for valve-regulated lead-acid batteries in substations according to the first embodiment of the present invention;
[0046] Figure 7This is a lead-acid battery SOH estimation effect diagram of a training set of a substation valve-regulated lead-acid battery SOH estimation method according to the first embodiment of the present invention;
[0047] Figure 8 This is a lead-acid battery SOH estimation effect diagram of a test set of the substation valve-regulated lead-acid battery SOH estimation method according to the first embodiment of the present invention. DETAILED DESCRIPTION
[0048] In order to make the purpose, technical solution and advantages of the embodiments of the present invention clearer, the technical solution in the embodiments of the present invention will be clearly and completely described below in combination with the embodiments of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.
[0049] The technical solution of the present invention is further described below in conjunction with the accompanying drawings and specific embodiments:
[0050] Embodiment 1
[0051] like Figure 1 As shown, the method for estimating SOH of a valve-controlled lead-acid battery in a substation according to an embodiment of the present invention comprises the following steps:
[0052] Step 1: Charge or discharge the battery to a specified SOC and measure the electrochemical impedance spectroscopy
[0053] According to the battery historical operation data, the terminal voltage data of the battery is measured, the battery is charged or discharged to a specified state of charge, left to stand for 10 minutes, and an electrochemical impedance spectroscopy measurement experiment is performed to obtain electrochemical impedance spectroscopy data; the specified state of charge refers to a state of charge with the best impedance spectrum discrimination determined based on the impedance spectrum curve of the equally spaced SOC of multiple lead-acid batteries measured in advance.
[0054] Step 2: Construct an equivalent circuit model based on electrochemical impedance spectroscopy
[0055] This embodiment adopts a first-order equivalent circuit LR (QR) (CR) structure model as a lead-acid battery equivalent circuit model based on electrochemical impedance spectroscopy, such as Figure 2 As shown, the equivalent circuit model includes: an inductor L, an ohmic resistor R1, a charge transfer internal resistor R2, a charge diffusion resistor R3, a constant phase element Q, and a capacitive reactance element C; the inductor L, the ohmic resistor R1, the charge transfer internal resistor R2, and the charge diffusion resistor R3 are connected in series in sequence, the constant phase element Q is connected in parallel at both ends of the charge transfer internal resistor R2, and the capacitive reactance element C is connected in parallel at both ends of the charge diffusion resistor R3.
[0056] The inductor L represents the straight line portion of the high frequency region of the electrochemical impedance spectrum curve, and the ohmic resistor R1 is at the point where the high frequency region and the medium frequency region of the electrochemical impedance spectrum curve are connected, that is, the composite element LR is used to describe the curve of the entire high frequency band equivalently, so as to more accurately reflect the dynamic response characteristics of the battery in the high frequency region. The charge transfer internal resistance R2 is connected in parallel with the constant phase element Q to effectively characterize the irregular semicircular portion of the impedance spectrum in the medium frequency region of the electrochemical impedance spectrum curve. The low frequency region is generally generated by the lack of the external electric field, which causes the charge inside the battery to migrate to the electrode in the form of diffusion. The charge diffusion portion of the low frequency region of the electrochemical impedance spectrum of the lead-acid battery measured by the present invention is not an ideal straight line with a tangent angle of 45°, so the capacitive reactance element C is connected in parallel with the resistor R3 to characterize the diffusion portion of the low frequency region. The equivalent circuit model thus established can effectively reflect the internal structure of the battery and obtain a more accurate fitting result.
[0057] Step 3: Impedance spectrum data fitting and determination of characteristic parameters
[0058] In step 1, the battery is charged or discharged to a specified SOC and the electrochemical impedance spectrum is measured. In this embodiment, the impedance spectrum data is measured by discharging the battery to 30% SOC and 60% SOC, and the obtained impedance spectrum data is fitted by the least squares method to obtain each parameter value in the equivalent circuit model as shown in Tables 1 and 2. The correlation between each parameter and the SOH value is judged respectively, and finally the n value in the fitting formula of the constant phase element Q with the strongest correlation with the battery SOH is selected, that is, Qn shown in the table as the characteristic parameter.
[0059] Table 1 Equivalent circuit model parameters and SOH comparison table (30% SOC)
[0060]
[0061]
[0062] Table 2 Equivalent circuit model parameters and SOH comparison table (60% SOC)
[0063] SOH L <![CDATA[R1]]> <![CDATA[Q-Y0]]> Q <![CDATA[R2]]> C <![CDATA[R3 <!-- 5 -->]]> 0.1770 8.44E-08 1.41E-03 1.07E+03 9.76E-02 2.04E+10 3.76E+04 1.78E-03 0.7542 1.12E-07 9.21E-04 2.04E+03 1.27E-01 4.88E+05 5.03E+04 8.04E-04 0.6667 7.93E-08 9.69E-04 1.89E+03 1.14E-01 1.04E+06 4.16E+04 1.06E-03 0.8055 6.60E-08 8.95E-04 2.53E+03 1.34E-01 3.44E+05 4.31E+04 8.01E-04 0.9475 8.44E-08 7.79E-04 4.62E+03 2.22E-01 9.62E-03 5.78E+04 6.01E-04 0.7405 9.51E-08 9.31E-04 2.02E+03 1.22E-01 6.44E+05 6.61E+04 4.89E-02 0.9722 8.83E-08 7.22E-04 5.62E+03 2.23E-01 6.14E-03 6.27E+04 4.89E-04 0.8372 7.16E-08 8.67E-04 2.87E+03 1.44E-01 9.15E+04 4.31E+04 7.47E-04 0.9896 8.51E-08 7.09E-04 5.69E+03 2.28E-01 4.37E-03 6.03E+04 4.94E-04 0.8466 7.54E-08 8.52E-04 3.54E+03 1.99E-01 9.78E+03 4.95E+04 7.34E-04 0.7182 8.73E-08 9.35E-04 2.00E+03 1.20E-01 7.08E+05 4.01E+04 9.10E-04 0.8258 7.17E-08 8.75E-04 2.61E+03 1.39E-01 2.35E+05 5.26E+04 7.26E-04 0.9020 8.12E-08 8.07E-04 3.93E+03 2.16E-01 1.49E+03 4.63E+04 6.46E-04 0.3313 8.69E-08 1.09E-03 1.26E+03 1.02E-01 1.75E+09 4.38E+04 2.45E-03 0.7764 7.55E-08 9.18E-04 2.20E+03 1.34E-01 4.19E+05 4.42E+04 6.36E-04 0.8464 7.78E-08 8.57E-04 3.20E+03 1.78E-01 1.62E+04 4.66E+04 6.31E-04 0.8466 7.75E-08 8.56E-04 3.37E+03 1.83E-01 1.31E+04 4.35E+04 7.23E-04 0.8817 7.93E-08 8.32E-04 3.90E+03 2.02E-01 5.08E+03 6.46E+04 7.53E-04 0.5822 8.84E-08 1.02E-03 1.81E+03 1.11E-01 2.09E+07 4.23E+04 1.60E-03 0.7137 7.93E-08 9.56E-04 1.98E+03 1.15E-01 9.38E+05 3.32E+04 9.37E-04 0.9644 7.83E-08 7.47E-04 4.80E+03 2.22E-01 6.20E-03 4.00E+04 7.08E-04 0.4737 1.18E-07 1.06E-03 1.38E+03 1.02E-01 8.13E+08 4.30E+04 1.36E-03 0.5752 7.65E-08 1.04E-03 1.45E+03 1.06E-01 3.68E+07 3.35E+04 1.18E-03 0.7161 7.66E-08 9.50E-04 1.98E+03 1.17E-01 8.15E+05 3.45E+04 9.06E-04 0.9027 8.62E-08 7.92E-04 3.97E+03 2.20E-01 8.02E-02 4.20E+04 1.29E-03 0.8227 9.01E-08 8.82E-04 2.61E+03 1.38E-01 3.39E+05 2.93E+04 1.38E-03 0.8354 9.04E-08 8.70E-04 2.73E+03 1.43E-01 1.02E+05 3.81E+04 1.12E-03
[0064] Among them, the fitting formula of the constant phase element Q is Among them, Y0 is the constant phase element parameter, the dimension is Ω -1 cm -2 ·s -n , always positive; j represents the sign of the imaginary part; ω represents the angular frequency; n represents the constant phase element index, which is 0 <n<1。
[0065] Step 4: Use the SA-SVR model to process the characteristic parameters to obtain the battery SOH estimate
[0066] The SA-SVR model (Support Vector Regression Model Optimized by Simulated Annealing Algorithm) is a fusion of the Simulated Annealing algorithm (SA) and the Support Vector Regression model (SVR). The Support Vector Regression model can obtain the optimal parameters under the optimization of the simulated annealing algorithm. The parameter value of the constant phase element Q in the equivalent circuit model obtained by the SA-SVR model is input, and the estimated value of the health status of the battery is output.
[0067] Specifically, the SA-SVR model is established as follows:
[0068] Step 4.1, define the objective function as follows:
[0069]
[0070] Among them, y i is the i-th sample value, is the simulated value of the i-th sample. The idea of optimizing the parameters of the SVR model is to search for a set of parameters (C, g, ε) through an iterative algorithm and minimize the fitness function through SVR learning.
[0071] Step 4.2: Use the SA algorithm to optimize the parameters (C, g, ε). The specific algorithm steps include:
[0072] Step 4.2.1, Initialize parameters. Select the SVM model training samples and prediction samples, perform sample normalization, set the penalty factor C, kernel function parameter g and search space of insensitivity coefficient ε; set the SA algorithm initial temperature T0, temperature cooling coefficient q, termination temperature T end and the length of the Metropolis chain L. Determine the function in step 4.1 as the optimization objective fitness function.
[0073] Step 4.2.2: Annealing operation: Perform global extreme value optimization according to the above SA algorithm.
[0074] Step 4.2.3: Prediction. Obtain the optimal solution, i.e., the parameter (C, g, ε) value, and use the parameter (C, g, ε) as the best learning parameter of SVR to predict the prediction sample.
[0075] like Figure 3As shown in Figure 1, the specific operation steps of the SA-SVR model are as follows: (1) Replace the main function of SVR with the objective function of simulated annealing so that the optimization process can directly act on the SVR model to achieve parameter optimization; (2) Set the input variables of SA as the key parameters of SVR: set the penalty factor C, kernel function parameter g and search space of the insensitivity coefficient ε; (3) In each SA iteration, calculate the prediction error as the return value of the objective function to evaluate the performance of the current parameter combination; (4) Initialize the SA parameters, that is, set the Markov chain length ML, temperature attenuation factor D, step size factor S of the Metropolis sampling criterion, initial temperature T0, and final temperature T end , the number of inner iterations G and the number of outer iterations M. The setting of these parameters is crucial to the convergence speed and optimization effect of the algorithm; (5) After repeated iterations and optimization of SA, the system can accurately predict the objective function value with the minimum error and the corresponding optimal parameter combination.
[0076] like Figure 4 As shown, the SA-SVR model is trained using the impedance spectrum fitting data and SOH data of multiple lead-acid batteries measured in advance, so that it can achieve the optimal solution for parameter estimation. The trained SA-SVR model is then used to estimate the SOH of the lead-acid battery. By inputting the fitting data, the SOH estimation value of the lead-acid battery can be directly obtained.
[0077] Test verification
[0078] like Figure 5 and Figure 6 As shown in the figure, two batteries with fixed capacity were selected to discharge at 10% SOC and measure the electrochemical impedance spectrum experiment. According to the impedance spectrum performance, the same SOC point was set to 60% SOC. The batteries were discharged to 60% SOC, and the electrochemical impedance spectrum data was measured. Figure 2 The equivalent circuit shown uses the least square method to fit the characteristic parameters, and selects the constant phase element Q as the characteristic parameter. The battery to be tested is divided into a 70% training set and a 30% test set. It should be noted that the training set is the part of the data sample used to train the SA-SVR model, and the independent test set is a part of the data sample without the test set.
[0079] Input the characteristic parameters of the training set into the SA-SVR model and train the SA-SVR model. For the training set effect, refer to Figure 7 ; From the figure, we can see that the root mean square error between the predicted value and the true value is RMSE = 0.0072789. Then, the characteristic parameters of the test set are input into the SA-SVR model. The model's effect on the battery SOH estimation of the test set is shown in Figure 8,It can be seen from the figure that the root mean square error RMSE between the predicted value and the true value is 0.0098176, which fully proves the accuracy of the SA-SVR model prediction of the present invention.
[0080] Embodiment 2
[0081] A substation valve-regulated lead-acid battery SOH estimation system comprises the following steps: an electrochemical impedance spectrum data measurement module, an equivalent circuit model construction module, a characteristic parameter determination module, and a model prediction module;
[0082] The electrochemical impedance spectroscopy data measurement module is used to charge or discharge the battery to a specified state of charge and measure electrochemical impedance spectroscopy data;
[0083] The equivalent circuit model building module is used to build an equivalent circuit model of a lead-acid battery based on electrochemical impedance spectroscopy;
[0084] The lead-acid battery equivalent circuit model based on electrochemical impedance spectroscopy includes: an inductor L, an ohmic resistor R1, a charge transfer internal resistor R2, a charge diffusion resistor R3, a constant phase element Q, and a capacitive reactance element C; the inductor L, the ohmic resistor R1, the charge transfer internal resistor R2, and the charge diffusion resistor R3 are connected in series in sequence, the constant phase element Q is connected in parallel at both ends of the charge transfer internal resistor R2, and the capacitive reactance element C is connected in parallel at both ends of the charge diffusion resistor R3.
[0085] The inductor L represents the straight line portion of the high frequency region of the electrochemical impedance spectrum curve, the ohmic resistor R1 is at the point where the high frequency region and the medium frequency region of the electrochemical impedance spectrum curve are connected, the charge transfer internal resistor R2 is connected in parallel with the constant phase element Q to represent the irregular semicircular portion of the impedance spectrum in the medium frequency region of the electrochemical impedance spectrum curve, and the capacitive reactance element C is connected in parallel with the charge diffusion resistor R3 to characterize the diffusion portion in the low frequency region.
[0086] The characteristic parameter determination module adopts the least square method to fit each parameter in the equivalent circuit model, and selects the constant phase element Q with the strongest correlation with the battery SOH as the characteristic parameter;
[0087] The fitting formula of the constant phase element Q is: Among them, Y0 is the constant phase element parameter, the dimension is Ω -1 cm -2 ·s -n , always positive; j represents the sign of the imaginary part; ω represents the angular frequency; n represents the constant phase element index, which is 0 <n<1,Z Q It represents the impedance of the constant phase element Q.
[0088] The model prediction module establishes a support vector regression model optimized by a simulated annealing algorithm, takes the parameter value of the constant phase element Q in the equivalent circuit model as the model input, and the estimated value of the health state of the battery as the model output, thereby obtaining the estimated value of the SOH of the lead-acid battery.
[0089] The method for establishing the support vector regression model optimized by the simulated annealing algorithm is as follows:
[0090] (1) The objective function is defined as follows:
[0091]
[0092] Among them, y i is the i-th sample value, is the simulated value of the i-th sample, C represents the penalty factor, g represents the kernel function parameter, and ε represents the insensitivity coefficient;
[0093] (2) Use the simulated annealing algorithm to optimize the parameters (C, g, ε). The specific steps are as follows:
[0094] 1) Initialize parameters; select support vector regression model training samples and prediction samples, perform sample normalization, set the penalty factor C, kernel function parameter g and search space of insensitivity coefficient ε; set the simulated annealing algorithm initial temperature T0, temperature cooling coefficient q, termination temperature T end and the length of the Metropolis chain L; determine the function in step (1) as the optimization target fitness function;
[0095] 2) Perform global extreme value optimization;
[0096] 3) Obtain the optimal solution, that is, the parameter (C, g, ε) value, and use the parameter (C, g, ε) as the optimal learning parameter of the support vector regression model to predict the prediction sample.
[0097] Embodiment 3
[0098] An electronic device includes a memory and a processor, wherein the memory is used to store a program that supports the processor to execute the distributed node time synchronization method in a local area network in embodiment 1, and the processor is configured to execute the program stored in the memory.
[0099] Embodiment 4
[0100] A storage medium stores a computer program, and when the computer program is executed by a processor, the steps of the distributed node time synchronization method in a local area network in embodiment 1 are executed.
[0101] The above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit the same. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that the technical solutions described in the aforementioned embodiments may still be modified, or some of the technical features may be replaced by equivalents. However, these modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the embodiments of the present invention.
Claims
1. A method for estimating SOH of a valve-controlled lead-acid battery in a substation, characterized in that: The following steps are involved: S1. charging or discharging the battery to a specified state of charge and measuring electrochemical impedance spectroscopy data; S2. Construct an equivalent circuit model of a lead-acid battery based on electrochemical impedance spectroscopy; S3, using the least squares method to fit each parameter in the equivalent circuit model, and selecting the constant phase element Q with the strongest correlation with the battery SOH as the characteristic parameter; S4. A support vector regression model optimized by a simulated annealing algorithm is established, with the parameter value of the constant phase element Q in the equivalent circuit model as the model input and the estimated value of the health state of the battery as the model output, thereby obtaining the estimated value of the SOH of the lead-acid battery.
2. The method for estimating SOH of a valve-controlled lead-acid battery in a substation according to claim 1, characterized in that: The lead-acid battery equivalent circuit model based on electrochemical impedance spectroscopy described in step S2 includes: an inductor L, an ohmic resistor R1, a charge transfer internal resistor R2, a charge diffusion resistor R3, a constant phase element Q, and a capacitive reactance element C; the inductor L, the ohmic resistor R1, the charge transfer internal resistor R2, and the charge diffusion resistor R3 are connected in series in sequence, the constant phase element Q is connected in parallel at both ends of the charge transfer internal resistor R2, and the capacitive reactance element C is connected in parallel at both ends of the charge diffusion resistor R3.
3. The method for estimating SOH of a valve-controlled lead-acid battery in a substation according to claim 2, characterized in that: The inductor L represents the straight line portion of the high frequency region of the electrochemical impedance spectrum curve, the ohmic resistor R1 is at the point where the high frequency region and the medium frequency region of the electrochemical impedance spectrum curve are connected, the charge transfer internal resistor R2 is connected in parallel with the constant phase element Q to represent the irregular semicircular portion of the impedance spectrum in the medium frequency region of the electrochemical impedance spectrum curve, and the capacitive reactance element C is connected in parallel with the charge diffusion resistor R3 to characterize the diffusion portion in the low frequency region.
4. The method for estimating SOH of a valve-controlled lead-acid battery in a substation according to claim 2, characterized in that: The fitting formula of the constant phase element Q described in step S3 is: Among them, Y0 is the constant phase element parameter, the dimension is Ω -1 cm -2 ·s -n , always positive; j represents the sign of the imaginary part; ω represents the angular frequency; n represents the constant phase element index, which is 0 <n<1,Z Q It represents the impedance of the constant phase element Q.
5. The method for estimating SOH of a valve-controlled lead-acid battery in a substation according to claim 1, characterized in that: The method for establishing the support vector regression model optimized by the simulated annealing algorithm described in step S4 is as follows: S4.
1. Define the objective function as follows: Among them, y i is the i-th sample value, is the simulated value of the i-th sample, C represents the penalty factor, g represents the kernel function parameter, and ε represents the insensitivity coefficient; S4.
2. Use the simulated annealing algorithm to optimize the parameters (C, g, ε). The specific steps are as follows: S4.2.
1. Initialize parameters; select support vector regression model training samples and prediction samples, perform sample normalization, set the penalty factor C, kernel function parameter g and search space of insensitivity coefficient ε; set the simulated annealing algorithm initial temperature T0, temperature cooling coefficient q, termination temperature T end and the length of the Metropolis chain L; determine the function in step S4.1 as the optimization target fitness function; S4.2.2, perform global extreme value optimization; S4.2.
3. Obtain the optimal solution, that is, the parameter (C, g, ε) value, and use the parameter (C, g, ε) as the optimal learning parameter of the support vector regression model to predict the prediction sample.
6. A substation valve-controlled lead-acid battery SOH estimation system, characterized in that: The following steps are involved: Electrochemical impedance spectroscopy data measurement module, equivalent circuit model construction module, characteristic parameter determination module, model prediction module; The electrochemical impedance spectroscopy data measurement module is used to charge or discharge the battery to a specified state of charge and measure electrochemical impedance spectroscopy data; The equivalent circuit model building module is used to build an equivalent circuit model of a lead-acid battery based on electrochemical impedance spectroscopy; The characteristic parameter determination module adopts the least square method to fit each parameter in the equivalent circuit model, and selects the constant phase element Q with the strongest correlation with the battery SOH as the characteristic parameter; The model prediction module establishes a support vector regression model optimized by a simulated annealing algorithm, takes the parameter value of the constant phase element Q in the equivalent circuit model as the model input, and the estimated value of the health state of the battery as the model output, thereby obtaining the estimated value of the SOH of the lead-acid battery.
7. The substation valve-regulated lead-acid battery SOH estimation system according to claim 6, characterized in that: The lead-acid battery equivalent circuit model based on electrochemical impedance spectroscopy described in the equivalent circuit model construction module includes: an inductor L, an ohmic resistor R1, a charge transfer internal resistor R2, a charge diffusion resistor R3, a constant phase element Q, and a capacitive reactance element C; the inductor L, the ohmic resistor R1, the charge transfer internal resistor R2, and the charge diffusion resistor R3 are connected in series in sequence, the constant phase element Q is connected in parallel at both ends of the charge transfer internal resistor R2, and the capacitive reactance element C is connected in parallel at both ends of the charge diffusion resistor R3; the inductor L represents the straight line part of the high-frequency region of the electrochemical impedance spectrum curve, the ohmic resistor R1 is at the point where the high-frequency region and the medium-frequency region of the electrochemical impedance spectrum curve are connected, the charge transfer internal resistor R2 and the constant phase element Q are connected in parallel to represent the irregular semicircular part of the impedance spectrum in the medium-frequency region of the electrochemical impedance spectrum curve, and the capacitive reactance element C and the charge diffusion resistor R3 are connected in parallel to characterize the diffusion part in the low-frequency region.
8. The substation valve-regulated lead-acid battery SOH estimation system according to claim 6, characterized in that: The fitting formula of the constant phase element Q described in the characteristic parameter determination module is: Among them, Y0 is the constant phase element parameter, the dimension is Ω -1 cm -2 ·s -n , always positive; j represents the sign of the imaginary part; ω represents the angular frequency; n represents the constant phase element index, which is 0 <n<1,Z Q It represents the impedance of the constant phase element Q.
9. The substation valve-regulated lead-acid battery SOH estimation system according to claim 6, characterized in that: The method for establishing the support vector regression model optimized by the simulated annealing algorithm described in the model prediction module is as follows: (1) The objective function is defined as follows: Among them, y i is the i-th sample value, is the simulated value of the i-th sample, C represents the penalty factor, g represents the kernel function parameter, and ε represents the insensitivity coefficient; (2) Use the simulated annealing algorithm to optimize the parameters (C, g, ε). The specific steps are as follows: 1) Initialize parameters; select support vector regression model training samples and prediction samples, perform sample normalization, set the penalty factor C, kernel function parameter g and search space of insensitivity coefficient ε; set the simulated annealing algorithm initial temperature T0, temperature cooling coefficient q, termination temperature T end and the length of the Metropolis chain L; determine the function in step (1) as the optimization target fitness function; 2) Perform global extreme value optimization; 3) Obtain the optimal solution, that is, the parameter (C, g, ε) value, and use the parameter (C, g, ε) as the optimal learning parameter of the support vector regression model to predict the prediction sample.
10. A storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the steps of the method for estimating SOH of a valve-regulated lead-acid battery in a substation according to any one of claims 1 to 5 are executed.
