Capacitance Calibration Method for Supercapacitor Energy Supply System in Vehicles

Abstract

The present invention belongs to the field of transportation technology and specifically provides a capacitance calibration method for a supercapacitor system in a vehicle powered by a single supercapacitor system or a hybrid energy storage system including supercapacitors. The capacitance calibration method includes: obtaining charging and discharging data of the supercapacitor system in the energy supply device under actual operating conditions; screening out required target segments from the charging and discharging data based on a preset standard time interval and a charge and discharge parameter change threshold corresponding to the standard time interval; utilizing a first-order equivalent circuit model and training the model as an equivalent model of the supercapacitor system using the actual values of the charge and discharge parameters in the target segment to obtain a preliminary calibration result of the supercapacitor system capacitance corresponding to the target segment; and calculating a final calibration result of the supercapacitor system capacitance based on the preliminary calibration results of multiple supercapacitor system capacitances.

Description

Technical Field

[0001] The present invention belongs to the field of transportation technology, and in particular relates to a capacitance calibration method applicable to a supercapacitor energy supply system in a vehicle. Background Art

[0002] As a new energy storage technology, supercapacitors (including double-layer capacitors, lithium-ion capacitors, sodium-ion capacitors, etc.) have the advantages of high power density, fast charging speed, long service life, and a wide operating temperature range. Therefore, various transportation vehicles (such as buses, trams, trains, electric ships, etc.) that use only supercapacitors as their energy supply system, or use hybrid energy storage systems composed of supercapacitors and other energy storage technologies (such as lithium-ion batteries, hydrogen fuel cells, sodium-ion batteries, etc.) as their energy supply system, have developed rapidly.

[0003] In order to ensure the normal operation of such vehicles, it is necessary to evaluate the state of health (SOH) of the supercapacitor system and predict its service life. Among them, the supercapacitor energy supply system involves cells, modules and systems from bottom to top. At present, the health status estimation of supercapacitors usually only involves cells, and is usually only tested under specific working conditions in the laboratory. However, actual vehicles usually pay more attention to the health status of the supercapacitor system, and the cell health status assessment method based on laboratory test data cannot generally be directly applied to supercapacitor systems under actual working conditions. In addition, considering the high time and economic cost of removing the supercapacitor system from the vehicle for offline health status assessment, how to use the operating data of the supercapacitor system under actual working conditions to evaluate its health status online has become an urgent problem to be solved. Summary of the Invention

[0004] In view of the shortcomings of the prior art mentioned above, the purpose of the present invention is to provide a method for online calibration of the capacitance of a supercapacitor system in a vehicle powered by a single supercapacitor system or a hybrid energy storage system including supercapacitors, so as to facilitate the subsequent evaluation of the SOH of the supercapacitor system and improve the safety and economic benefits of the operation of vehicles in the transportation system.

[0005] To achieve the above-mentioned purpose and other related purposes, the present invention provides a capacitance calibration method for a supercapacitor energy supply system in a vehicle, including: obtaining charging and discharging data of the supercapacitor system in the energy supply device under actual operating conditions, wherein the charging and discharging data is formed by charging and discharging parameters of continuously sampled time nodes; screening out the required target segment from the charging and discharging data according to a preset standard time interval and a charging and discharging parameter change threshold corresponding to the standard time interval; utilizing a first-order equivalent circuit model, and training the first-order equivalent circuit model into an equivalent model of the supercapacitor system through the actual values of the charging and discharging parameters in the target segment, so as to obtain a preliminary calibration result of the capacitance of the supercapacitor system corresponding to the target segment; and calculating the final calibration result of the capacitance of the supercapacitor system based on multiple preliminary calibration results of the capacitance of the supercapacitor system.

[0006] According to a specific embodiment of the present invention, the step of screening out the desired target segment from the charge and discharge data according to a preset standard time interval and a charge and discharge parameter change threshold value corresponding to the standard time interval includes: preprocessing the charge and discharge data; and for each time node in the processed charge and discharge data, screening out the discharge segment of the supercapacitor system, i.e., the target segment, according to the standard time interval and the corresponding charge and discharge parameter change threshold value.

[0007] According to a specific embodiment of the present invention, the step of preprocessing the charge and discharge data includes: identifying each time node in the charge and discharge data; if the time node samples multiple charge and discharge parameters, selecting the charge and discharge parameter closest to the next time node as the charge and discharge parameter corresponding to the time node, and using the time node as an available time node; if the time node lacks the corresponding charge and discharge parameter, using the time node as an unavailable time node; integrating several available time nodes and their corresponding charge and discharge parameters.

[0008] According to a specific embodiment of the present invention, the discharge segment of the supercapacitor system is screened out according to the standard time interval and its corresponding charge and discharge parameter change threshold, that is, the step of the target segment includes: calculating the time interval between each time node and its previous time node, and taking the time node whose time interval does not exceed the standard time interval as the time node after the first screening; integrating several time nodes after the first screening and their corresponding charge and discharge parameters to obtain the charge / discharge segment of the supercapacitor system; for each time node in the charge / discharge segment, screening out the discharge segment according to the charge and discharge parameter change threshold.

[0009] According to a specific embodiment of the present invention, the step of filtering out the discharge segment according to the charge and discharge parameter change threshold includes: calculating the current difference between each time node and its previous time node, and taking the time node where the current difference does not exceed the first threshold in the charge and discharge parameter change threshold as the time node after the second screening; integrating several time nodes after the second screening and their corresponding charge and discharge parameters to obtain the discharge segment; wherein the charge and discharge parameters include at least current.

[0010] According to a specific embodiment of the present invention, the step of filtering out the discharge segment according to the charge and discharge parameter change threshold also includes: for each time node in the discharge segment, calculating the voltage difference between each time node and its previous time node, and taking the time node where the voltage difference does not exceed the second threshold in the charge and discharge parameter change threshold as the time node after the third screening; integrating several time nodes after the third screening and their corresponding charge and discharge parameters to obtain an optimized discharge segment; wherein the charge and discharge parameters also include voltage.

[0011] According to a specific embodiment of the present invention, a first-order equivalent circuit model is used, and the first-order equivalent circuit model is trained to become an equivalent model of the supercapacitor system through the true values of the charge and discharge parameters in the target segment to obtain a preliminary calibration result of the capacitance of the supercapacitor system corresponding to the target segment. The steps include: according to the true values of the charge and discharge parameters corresponding to different time nodes in the target segment, using the first-order equivalent circuit model to calculate the voltage prediction value of the corresponding time node in the target segment; constructing a loss function according to the true value and the voltage prediction value of the charge and discharge parameters corresponding to the target segment, and optimizing the first-order equivalent circuit model by minimizing the loss function; and using the equivalent capacitance parameters in the optimized first-order equivalent circuit model as the preliminary calibration result of the capacitance of the supercapacitor system.

[0012] According to a specific embodiment of the present invention, a loss function is constructed based on the voltage prediction value and the true value of the voltage in the charging and discharging parameters corresponding to the target segment, and the step of optimizing the first-order equivalent circuit model by minimizing the loss function includes: using the gradient descent method to calculate the gradient of the loss function, and optimizing the first-order equivalent circuit model along the descending direction of the gradient until the convergence condition is met.

[0013] According to a specific embodiment of the present invention, the step of calculating the gradient of the loss function using the gradient descent method and optimizing the first-order equivalent circuit model along the descending direction of the gradient until the convergence condition is met includes: optimizing the first-order equivalent circuit model according to the following formula:

[0014]

[0015] Where L represents the loss function, G represents the inverse of the equivalent capacitance parameter of the first-order equivalent circuit model, R represents the inverse of the equivalent resistance parameter of the first-order equivalent circuit model, η represents the learning rate in the gradient descent direction, k represents the number of iterations, n represents the number of time nodes of the predicted voltage in the target segment, i represents the time node of the i-th predicted voltage in the target segment, Represents the voltage prediction value of the i-th time node, V out (t i ) represents the true value of the voltage at the i-th time node; and convergence is completed when the loss function is less than the preset first threshold or the number of iterations is greater than the preset second threshold.

[0016] According to a specific embodiment of the present invention, the step of calculating the final calibration result of the supercapacitor system capacitance based on the preliminary calibration results of the multiple supercapacitor system capacitances includes: taking the average of the preliminary calibration results of all the supercapacitor system capacitances in a day as the final calibration result of the supercapacitor system capacitance for that day; calculating the average of the final calibration results of the supercapacitor system capacitances for each day in a month as the final calibration result of the supercapacitor system capacitance for that month; and calculating the average of the final calibration results of the supercapacitor system capacitances for each month in a year as the final calibration result of the supercapacitor system capacitance for that year.

[0017] The present invention provides a method for online capacitance calibration of a supercapacitor system in a vehicle. Available charge and discharge data can be screened online based on the actual operating data of the vehicle, thereby avoiding the influence of the discontinuity, large fluctuation, large granularity, and fragmentation of the actual operating data. The accuracy and reliability of subsequent capacitance calibration are also correspondingly improved. The equivalent model of the supercapacitor system is optimized with a gradient descent strategy through the screened charge and discharge data, thereby realizing online capacitance calibration of the supercapacitor system, thereby facilitating online evaluation of the SOH of the supercapacitor system and improving the safety and economic benefits of the vehicle in actual working conditions.

[0018] The online capacitance calibration method provided by the present invention is not only applicable to the capacitance calibration of a single supercapacitor system, but also to the capacitance calibration of a hybrid energy storage system including supercapacitors, so that it can also be applied in the face of the development of diverse energy supply devices in future vehicles. BRIEF DESCRIPTION OF THE DRAWINGS

[0019] Figure 1 This is a flow chart of a specific embodiment of a capacitance calibration method for a supercapacitor energy supply system in a vehicle provided by the present invention;

[0020] Figure 2 A curve diagram showing the voltage and current changes of the supercapacitor system in the energy supply device provided by the present invention, with the charging behavior interspersed with the discharging behavior;

[0021] Figure 3 This is a schematic diagram of the voltage and current actually sampled at different time points by the supercapacitor system in the energy supply device provided by the present invention;

[0022] Figure 4 A curve diagram showing the voltage change during the discharge phase of the supercapacitor system in the energy supply device provided by the present invention;

[0023] Figure 5 This is a circuit diagram of a first-order equivalent circuit model of a supercapacitor system in the energy supply device provided by the present invention;

[0024] Figure 6 This is a schematic diagram of the preliminary calibration results of the discharge segment of the supercapacitor system in the energy supply device provided by the present invention. DETAILED DESCRIPTION

[0025] The following describes the embodiments of the present invention with reference to the accompanying drawings and preferred embodiments. Those skilled in the art will readily appreciate the other advantages and benefits of the present invention from the disclosure herein. The present invention may also be implemented or applied through various other specific embodiments, and the various details in this specification may be modified or altered based on different viewpoints and applications without departing from the spirit of the present invention. It should be understood that the preferred embodiments are intended only to illustrate the present invention and are not intended to limit the scope of protection of the present invention.

[0026] It should be noted that the illustrations provided in the following embodiments are merely schematic illustrations of the basic concept of the present invention. Therefore, the illustrations only show components related to the present invention and are not drawn according to the number, shape, and size of components in actual implementation. In actual implementation, the type, quantity, and proportion of each component may be changed arbitrarily, and the component layout may also be more complex.

[0027] In the following description, numerous details are discussed to provide a more thorough explanation of the embodiments of the present invention. However, it will be apparent to those skilled in the art that the embodiments of the present invention may be practiced without these specific details. In other embodiments, well-known structures and devices are shown in block diagram form rather than in detail to avoid obscuring the embodiments of the present invention.

[0028] First of all, it should be noted that in order to enable people in this technical field to better understand the solution of this application, the technical background of this application is explained accordingly.

[0029] Normally, external means are required to evaluate the health status of a vehicle that uses a single supercapacitor system or a hybrid energy storage system containing supercapacitors as the energy supply device. Specifically, several supercapacitor modules are selected from the vehicle's energy supply device for offline factory monitoring to obtain the electrostatic capacitance retention rate, thereby indirectly reflecting the health status of the entire vehicle.

[0030] However, the supercapacitor modules selected for offline return monitoring are randomly sampled from different locations in the standard box of the vehicle. The sample size accounts for a relatively low proportion of the total number of modules and lacks a certain degree of universality. In addition, during the offline return monitoring process, the vehicle is out of service, involving immeasurable time and economic costs. Therefore, offline return monitoring is not often used to evaluate the health of the supercapacitor system of a vehicle in operation.

[0031] In addition, since the actual operating data of the supercapacitor system of the carrier vehicle is discontinuous, highly volatile, granular, and scattered, the existing methods, such as equivalent circuit models based on algorithm optimization strategies or mathematical optimization strategies, are used to calibrate the capacitance of the supercapacitor system, which performs poorly and cannot identify abnormal fragments in the actual operating data well, resulting in differences between the calibration results and the actual results.

[0032] Therefore, from the system level, in order to reduce time and economic costs, this application proposes an online capacitance calibration method for a vehicle powered by a single supercapacitor system or a hybrid energy storage system containing supercapacitors, so as to facilitate the subsequent evaluation of the SOH of the energy supply device in the vehicle.

[0033] See Figure 1 As shown, a capacitance calibration method for a supercapacitor energy supply system in a vehicle includes:

[0034] Step S100 , obtaining charging and discharging data of a supercapacitor system in an energy supply device under actual operating conditions, wherein the charging and discharging data is formed by charging and discharging parameters at continuously sampled time nodes.

[0035] It is understandable that in actual applications, when a vehicle that uses a single supercapacitor system or a hybrid energy storage system containing supercapacitors as its energy supply device is in operation, it is mainly powered by the discharge of the supercapacitor system in the energy supply device; and when arriving at a station, the supercapacitor system can be short-term recharged by fast charging, that is, "recharge at the station and discharge between stations." In addition, the vehicle is also accompanied by irregular energy recovery during operation. Therefore, the supercapacitor system in its energy supply device will both discharge and charge during the operation of the vehicle. The actual operating data of the corresponding vehicle includes the charging and discharging data of the supercapacitor system, for example, see Figure 2The supercapacitor system shown in the figure has charging behaviors interspersed between discharging behaviors during actual operation.

[0036] Furthermore, the charge and discharge data of the supercapacitor system can be obtained by collecting the charge and discharge parameters of the supercapacitor system at preset time intervals. For example, according to the national standard for remote transmission, the voltage across the supercapacitor system, or the input / output current, is sampled every 10 seconds, and the charge and discharge data is composed of the voltage and current corresponding to several consecutive sampling time nodes. Of course, the sampling time interval is not strictly enforced in actual operation due to the interference of external operating conditions, and the data obtained in the standby state needs to be adjusted accordingly according to actual needs.

[0037] In addition, it should be noted that, since the actual operating data of the carrier is discontinuous, highly volatile, granular, and fragmented, the charge and discharge data mentioned in this embodiment at least includes the collection of partial fragments of the discharge behavior and / or charging behavior of the supercapacitor system. For example, the voltage across the supercapacitor system, or the input / output current, is sampled once every 10 seconds. Due to other factors such as the maintenance status of the carrier, or the data transmission status, or the outage status, the corresponding charge and discharge parameters are not collected within a period of time. In this case, the data portion cannot be used as charge and discharge data to calibrate the capacitance of the supercapacitor system.

[0038] Accordingly, the proportion of time nodes with voltage and current samples in the acquired charging and discharging data can be preliminarily restricted according to actual needs. For example, the sampling diagram can be used to observe whether the sampling of time nodes in a certain section of data is missing a large number of corresponding charging and discharging parameters. If so, it will not be considered as available data for calibrating the capacitance of the supercapacitor system.

[0039] Step S200 , filtering out a desired target segment from the charge and discharge data according to a preset standard time interval and a charge and discharge parameter change threshold value corresponding to the standard time interval.

[0040] It is understandable that, unlike the data obtained from laboratory tests, the actual data of the vehicle in actual operation will affect the data sampling time interval due to uncertain external working conditions, which in turn affects the data quality. In the collected charge and discharge data, there will be missing charge and discharge parameters according to the sampling time nodes, or there will be multiple sets of charge and discharge parameters recorded, for example Figure 3 The different time nodes shown correspond to the voltage and current of the supercapacitor system. To address this phenomenon, it is necessary to first clean the acquired charge and discharge data, that is, pre-process the charge and discharge data.

[0041] Specifically, identify each time node in the charge and discharge data:

[0042] If a time node lacks the corresponding charge and discharge parameters, for example, the voltage and / or current of the supercapacitor system at this time is not recorded at the time node, then the time node is considered to be unusable data and does not contain the charge and discharge parameters for subsequent capacitance calibration, that is, an unusable time node. For example, the original time nodes for charge and discharge data sampling include: time node 10s, time node 20s, and time node 30s. If the corresponding voltage and current are not sampled at time node 20s, then the time node is considered to have no available operating data. The preprocessed charge and discharge data only includes the voltage and current corresponding to time node 10s and the voltage and current corresponding to time node 30s.

[0043] If multiple sets of charge and discharge parameters are sampled at a certain time node, one of them needs to be selected to represent the charge and discharge parameters corresponding to that time node. At the same time, it is preliminarily considered that the time node can be retained as usable operating data, i.e., a usable time node. Therefore, in this embodiment, the charge and discharge parameters closest to the next moment are selected from the multiple sampled charge and discharge parameters, i.e., the most recent data with higher real-time performance, to represent the charge and discharge parameters corresponding to that time node.

[0044] After the above processing steps, several available time nodes and their corresponding charge and discharge parameters are integrated to obtain the charge and discharge data after preliminary processing.

[0045] It should be noted that when multiple charge and discharge parameters are sampled at a certain time node, the above is only a preferred embodiment, including but not limited to the above method. For example, the average value can be obtained from multiple charge and discharge parameters as the charge and discharge parameter of the time node, or other methods can be used to select one that meets the requirements from multiple charge and discharge parameters. There are no excessive restrictions on this. Those skilled in the art can make modifications and improvements to the embodiments of the present invention without departing from the spirit of the present invention. They still fall within the scope of the invention application of the present invention.

[0046] Furthermore, considering the fast charging characteristics of supercapacitors, the time intervals between sampling time points in actual applications are relatively large. Therefore, the actual operation data contains fewer sampling time points for the supercapacitor system charging behavior, and the corresponding recorded charging and discharging parameters are also relatively few. However, at the same time, the current value during supercapacitor charging behavior is relatively large, so the supercapacitor charging behavior is manifested as large current fluctuations within a limited time point. Directly using the data fragments corresponding to the charging behavior will introduce large errors.

[0047] For example, see Figure 2The charging behavior shown is interspersed in the discharge behavior. The charging behavior starts at 48311s and ends at 48372s, with a duration of 61s. There are 7 effective sampling time nodes. Among them, the maximum current is 253A, and the maximum value of the voltage mutation between adjacent time nodes is 17V. It can be seen that the data fragments recording the charging behavior are not sufficient to provide sufficient specific details to explain the charging behavior itself. Therefore, in this embodiment, it is necessary to filter out the discharge fragments from the above-mentioned processed charge and discharge data as usable data, that is, for each time node in the processed charge and discharge data, according to the standard time interval and the corresponding charge and discharge parameter change threshold, the discharge fragment of the supercapacitor system, that is, the target fragment, is filtered out.

[0048] Specifically, considering that transmission errors may occur in actual acquisition, such as delays in acquisition or recording, a delay may occur when sampling should be performed at a certain time node according to the preset time interval. For example, the previous time node is the 10th second, the preset sampling interval is 10 seconds, and the next time node should be sampled at the 20th second according to the preset time interval. However, in actual application, due to transmission errors, sampling may be completed at the 25th second. If the sampling interval of 10 seconds is used as the screening criterion at this time, the time node of 25 seconds is removed, which may segment the complete discharge behavior process.

[0049] Therefore, in this embodiment, a standard time interval is set for screening, which is equivalent to allowing a sampling error range. For example, according to the above, the charge and discharge parameters sampled at the 25th second are allowed as usable data, and the time node of the 25th second is retained accordingly.

[0050] It is understandable that if the error is large and exceeds this error range, the data loses its corresponding timeliness and cannot be used as usable data. Specifically, by identifying each time node in the processed charge and discharge data, calculating the time interval between each time node and its previous time node, and taking the time node whose time interval does not exceed the standard time interval as the time node after the first screening, and integrating several time nodes after the first screening and their corresponding charge and discharge parameters, the charge / discharge segment of the supercapacitor system is obtained, which corresponds to the charge / discharge behavior of the supercapacitor system in the charge and discharge data.

[0051] Furthermore, since the goal is to obtain the discharge behavior process of the supercapacitor system, it is necessary to screen out the discharge segments from the above-mentioned charge / discharge segments. It should be noted here that the time intervals between adjacent time nodes in the charge / discharge segments obtained after the preliminary screening of the above-mentioned standard time intervals meet the standard time intervals. Therefore, the charge and discharge parameter change threshold corresponding to the standard time interval can be directly used as a standard to judge the charge and discharge parameter changes between adjacent time nodes in the charge / discharge segments. Moreover, in this embodiment, the charge and discharge parameters of the supercapacitor are mainly current and voltage, and the current changes and / or voltage changes between adjacent time nodes can be identified for screening.

[0052] Therefore, for each time node in the charge / discharge segment, the discharge segment is screened out according to the charge / discharge parameter change threshold.

[0053] First, we screen from the current aspect. It should be added here that the supercapacitor system in the carrier vehicle exhibits a "constant current" discharge phenomenon accompanied by small fluctuations during the actual discharge behavior. The current value remains constant most of the time, but it is not strictly constant, and basically fluctuates within ±2A.

[0054] In order to obtain a complete discharge segment, the first threshold value in the charge and discharge parameter change threshold value can be set to 2A, that is, a certain degree of current fluctuation is allowed in the discharge behavior. However, when the current difference between adjacent time nodes is greater than the set standard, that is, the first threshold value, it means that the current behavior is not in the same discharge behavior, and the corresponding time node is not considered as available data for subsequent capacitance calibration of the supercapacitor system, that is, the current difference between each time node and its previous time node is calculated, and the time node where the current difference does not exceed the first threshold value in the charge and discharge parameter change threshold value is used as the time node after the second screening, and several time nodes after the second screening and their corresponding charge and discharge parameters are integrated to obtain the discharge segment of the supercapacitor system, which corresponds to the discharge behavior of the supercapacitor system in the charge and discharge data.

[0055] Secondly, taking into account the abnormal voltage situation, in order to ensure the reliability of the obtained discharge fragments, it is also necessary to screen from the voltage level. The second threshold value in the charge and discharge parameter change threshold value can be set to 2V accordingly, that is, a certain degree of voltage fluctuation is allowed in the discharge behavior. However, when the current difference between adjacent time nodes is greater than the set standard, that is, the first threshold value, it means that there is an abnormality in the current behavior, and the corresponding time node is not considered as available data for subsequent capacitance calibration of the supercapacitor system to ensure the accuracy and reliability of the data, that is, for each time node in the discharge fragment, the voltage difference between each time node and its previous time node is calculated, and the time node where the voltage difference does not exceed the second threshold value in the charge and discharge parameter change threshold value is used as the time node after the third screening, and several time nodes after the third screening and their corresponding charge and discharge parameters are integrated to obtain the optimized discharge fragment.

[0056] For details, please refer to Figure 4 In the discharge segment shown, the discharge behavior of the supercapacitor system is 7A constant current discharge, and the voltage drops from 792V to 789V. It should be noted that, due to the limitation of the sensor's acquisition accuracy, Figure 4 It appears that the voltage remains unchanged for a period of time, but in fact the voltage is slowly decreasing by decimal places.

[0057] Step S300: Using a first-order equivalent circuit model, the first-order equivalent circuit model is trained into an equivalent model of the supercapacitor system using the actual values of the charge and discharge parameters in the target segment, so as to obtain a preliminary calibration result of the supercapacitor system capacitance corresponding to the target segment.

[0058] Since the charging and discharging parameters of the supercapacitor system are mainly voltage and current, a first-order equivalent circuit model, namely a resistor-capacitor (RC) model, is adopted in this embodiment. The discharge fragments of the supercapacitor system obtained by screening in the above steps are used to train the model into an equivalent model of the supercapacitor system to obtain the capacitance value of the supercapacitor system.

[0059] For details, please refer to Figure 5 The RC model of the supercapacitor system is shown, where V out (t) and I out (t) represents the terminal voltage and terminal current of the supercapacitor system sampled at different time nodes. It can be understood that the charge and discharge parameters of the supercapacitor system mentioned in this embodiment are its terminal voltage and terminal current. C (t) is the voltage of the equivalent capacitor in the RC model, and V C The equation for (t) can be referred to as follows:

[0060]

[0061] Among them, t1 is the current time node, t2 is the next time node, C is the capacitance of the equivalent capacitor in the RC model, and the voltage V on the supercapacitor system out (t) Satisfy:

[0062] V out (t) = V C (t)+I out (t)·R,

[0063] Based on the above two formulas, the voltage prediction value of the corresponding time node in the target segment can be calculated according to the actual values of voltage and current corresponding to different time nodes in the target segment. It can be understood that if there are i time nodes in a target segment, i-1 voltage prediction values can be calculated through the RC model. For details, please refer to the following:

[0064]

[0065] Where R is the resistance of the equivalent resistor in the RC model.

[0066] Furthermore, a loss function is constructed based on the actual value of the voltage and the predicted value of the voltage in the charge and discharge parameters corresponding to the target segment, and the parameters of the equivalent capacitance and equivalent resistance in the RC model are optimized by minimizing the loss function. It can be understood that the loss function is used to measure the difference between the predicted result of the model and the true value. By continuously adjusting the parameters of the model, the loss function is gradually reduced, that is, the predicted result is as close to the true value as possible, the parameters of the model are continuously optimized, and the prediction ability is gradually improved. In this embodiment, the gradient of the loss function is calculated using the gradient descent method, and the RC model is optimized along the descending direction of the gradient until the convergence condition is met to update the model parameters.

[0067] Specifically, first according to the voltage prediction value The difference between the actual voltage value and the actual voltage value is used to construct the loss function L, which can be seen as follows:

[0068]

[0069] Where n represents the number of time nodes of the predicted voltage in a target segment, i represents the time node of the i-th predicted voltage in the target segment, Represents the voltage prediction value of the i-th time node, V out (t i ) represents the true value of the voltage at the i-th time node.

[0070] Based on the above loss function, the calculation formula of the corresponding gradient is established:

[0071]

[0072] Where G is the reciprocal of the capacitance C of the equivalent capacitor.

[0073] The resistance value R of the equivalent resistor and the reciprocal value G of the equivalent capacitor in the next iteration can be adjusted according to the gradient descent direction and the learning rate η until the convergence condition is met and the iteration is completed, so that the RC model can be equivalent to a supercapacitor system. The specific iterative optimization can be seen as follows:

[0074]

[0075] Where k is the number of iterations.

[0076] In a specific embodiment, when the learning rate η is 0.001, when the loss function L is less than 0.002 or the number of iterations k is greater than 500, the resistance R of the equivalent resistor and the capacitance C of the equivalent capacitor obtained at this time are the optimal parameters, and the capacitance C of the equivalent capacitor of the RC model at this time can be regarded as the capacitance of the supercapacitor system, that is, the preliminary calibration result of the capacitance of the supercapacitor system corresponding to the target segment.

[0077] For details, please refer to Figure 6 The preliminary calibration results of multiple target segments are shown, among which the maximum capacitance is 357.80F, the minimum capacitance is 353.14F, the range of all capacitance values is 4.66F, and the standard deviation is 1.53F, which meets the allowable error range. It can be seen that the above preliminary calibration of the capacitance value of each target segment can obtain consistent calibration results.

[0078] Step S400 , calculating a final calibration result of the supercapacitor system capacitance based on a plurality of preliminary calibration results of the supercapacitor system capacitance.

[0079] It is understandable that the charging and discharging data may only be a section of actual data during the operating hours of the carrier on that day. For example, if the working hours of a carrier are 8:00-10:00 and 16:00-18:00, its charging and discharging data from 8:00-10:00 and 16:00-18:00 can be obtained. Corresponding discharge segments are obtained by screening according to different charging and discharging data, and the corresponding preliminary calibration results are calculated based on the discharge segments, so that the preliminary calibration results of the capacitance of multiple supercapacitor systems can be obtained.

[0080] However, the preliminary calibration results corresponding to a certain discharge segment are not sufficient to accurately describe the actual capacitance of the supercapacitor system. In order to improve the reliability of the capacitance calibration of the supercapacitor system, several preliminary calibration results calculated can be classified and analyzed according to the time of day, month, and year, and then the final calibration results of the supercapacitor system on that day, month, and year can be obtained.

[0081] Specifically, the average of all preliminary calibration results within a day can be used as the final calibration result of the supercapacitor system capacitance for that day. The average of the final calibration results of the supercapacitor system capacitance for each day within a month can be used as the final calibration result of the supercapacitor system capacitance for that month. Similarly, the average of the final calibration results of the supercapacitor system capacitance for each month within a year can be used as the final calibration result of the supercapacitor system capacitance for that year.

[0082] It should be noted that the above is only a preferred solution, which is implemented when the actual operating data of the carrier is complete enough. If the sampling setting factors of the carrier or other working conditions result in the failure to perform sampling every day within a year or a month, the final calibration result of the supercapacitor system capacitance value selected on a certain day within a month can be used as a representative of the current month. Similarly, the final calibration result of the supercapacitor system capacitance value selected on a certain month within a year can also be used as a representative of the current year. There are no excessive restrictions on this. Without departing from the spirit of the present invention, the modifications and embellishments made to the embodiments of the present invention by those skilled in the art still fall within the scope of the invention patent application of the present invention.

[0083] It should be noted that the step division of the various methods above is only for the purpose of clear description. During implementation, they can be combined into one step or some steps can be split and decomposed into multiple steps. As long as they contain the same logical relationship, they are all within the scope of protection of this patent; adding insignificant modifications to the algorithm or process or introducing insignificant designs without changing the core design of the algorithm and process are all within the scope of protection of this patent.

[0084] In summary, the present invention provides a method for online calibration of the capacitance of a supercapacitor system in a carrier. Available charging and discharging data can be screened out online based on the actual operating data of the carrier, avoiding the influence of the discontinuous, large fluctuation, large granularity, and fragmentation of the actual operating data. The accuracy and reliability of subsequent capacitance calibration are also improved accordingly. The equivalent model of the supercapacitor system is optimized with a gradient descent strategy through the screened charging and discharging data to realize online calibration of the capacitance of the supercapacitor system, so as to realize online evaluation of the SOH of the supercapacitor system and improve the safety and economic benefits of the carrier in actual working conditions.

[0085] The online capacitance calibration method provided by the present invention is not only applicable to the capacitance calibration of a single supercapacitor system, but also to the capacitance calibration of a hybrid energy storage system including supercapacitors, so that it can also be applied in the face of the development of diverse energy supply devices in future vehicles.

[0086] The present invention provides a capacitance calibration method for a vehicle powered by a supercapacitor system or a hybrid energy storage system containing a supercapacitor. The method can filter out available data based on the actual operating data of the vehicle powered by the supercapacitor system or the hybrid energy storage system containing a supercapacitor, thereby avoiding the influence of the discontinuity, large fluctuation, large granularity, and fragmentation of the actual operating data, thereby correspondingly improving the accuracy and reliability of subsequent capacitance calibration. In addition, by establishing a supercapacitor equivalent model and optimizing the parameters of the model using a gradient descent strategy, the capacitance of the supercapacitor system in the vehicle powered by the supercapacitor system during actual operation can be calibrated online, thereby increasing the safety and economic benefits of the vehicle powered by the supercapacitor system in actual working conditions, and facilitating the online evaluation of the SOH of the supercapacitor system.

[0087] The above embodiments are merely illustrative of the principles and effects of the present invention and are not intended to limit the present invention. Anyone skilled in the art may modify or alter the above embodiments without departing from the spirit and scope of the present invention. Therefore, any equivalent modifications or alterations made by one of ordinary skill in the art without departing from the spirit and technical principles disclosed herein are intended to be covered by the claims of the present invention.

Claims

1. A capacitance calibration method for a supercapacitor energy supply system in a vehicle, characterized in that: include: Obtaining charge and discharge data of the supercapacitor system in the energy supply device under actual operating conditions, wherein the charge and discharge data is formed by charge and discharge parameters at continuously sampled time nodes; wherein the continuously sampled time nodes are obtained by continuous sampling at preset sampling time intervals; The method comprises the following steps: screening out a desired target segment from the charge and discharge data according to a preset standard time interval and a charge and discharge parameter change threshold value corresponding to the standard time interval; and screening out a discharge segment of the supercapacitor system, i.e., the target segment, from each time node in the processed charge and discharge data according to the standard time interval and the charge and discharge parameter change threshold value corresponding thereto; wherein the standard time interval is at least greater than the sampling time interval; A first-order equivalent circuit model is used, and the first-order equivalent circuit model is trained into an equivalent model of the supercapacitor system through the true values of the charge and discharge parameters in the target segment to obtain a preliminary calibration result of the capacitance of the supercapacitor system corresponding to the target segment, and the steps include: according to the true values of the charge and discharge parameters corresponding to different time nodes in the target segment, the voltage prediction value of the corresponding time node in the target segment is calculated by using the first-order equivalent circuit model; according to the true value and the voltage prediction value of the charge and discharge parameters corresponding to the target segment, a loss function is constructed, and the first-order equivalent circuit model is optimized by minimizing the loss function; the equivalent capacitance parameters in the optimized first-order equivalent circuit model are used as the preliminary calibration result of the capacitance of the supercapacitor system; wherein, the gradient of the loss function is calculated by the gradient descent method, and the equivalent capacitance parameters in the first-order equivalent circuit model are optimized along the descending direction of the gradient until the convergence condition is met; Calculating a final calibration result of the supercapacitor system capacitance based on the multiple preliminary calibration results of the supercapacitor system capacitance; The step of selecting the discharge segment of the supercapacitor system, i.e., the target segment, according to the standard time interval and the corresponding charge and discharge parameter change threshold comprises: Calculating the time interval between each time node and its previous time node, and taking the time node whose time interval does not exceed the standard time interval as the time node after the first screening; integrating several time nodes after the first screening and their corresponding charge and discharge parameters to obtain the charge / discharge segment of the supercapacitor system; for each time node in the charge / discharge segment, screening the discharge segment according to the charge and discharge parameter change threshold, and the steps include: Calculating the current difference between each time node and the previous time node, and taking the time node where the current difference does not exceed the first threshold value of the charge and discharge parameter change threshold value as the time node after the second screening; integrating several time nodes after the second screening and their corresponding charge and discharge parameters to obtain the discharge segment; The charge and discharge parameters include at least current.

2. The capacitance calibration method for a supercapacitor energy supply system in a vehicle according to claim 1, characterized in that: The step of preprocessing the charge and discharge data includes: Identify each time node in the charge and discharge data: If there are multiple charge and discharge parameters sampled at the time node, the charge and discharge parameter closest to the next time node is selected as the charge and discharge parameter corresponding to the time node, and the time node is used as an available time node; If the time node lacks the corresponding charge and discharge parameters, the time node is regarded as an unavailable time node; Integrate several available time nodes and their corresponding charging and discharging parameters.

3. The capacitance calibration method for a supercapacitor energy supply system in a vehicle according to claim 1, characterized in that: The step of selecting the discharge segment according to the charge and discharge parameter change threshold further includes: For each time node in the discharge segment, Calculating the voltage difference between each time node and the previous time node, and taking the time node at which the voltage difference does not exceed the second threshold value in the charge and discharge parameter change threshold value as the time node after the third screening; Integrate several time nodes after the third screening and their corresponding charge and discharge parameters to obtain the optimized discharge segment; Wherein, the charging and discharging parameters also include voltage.

4. The capacitance calibration method for a supercapacitor energy supply system in a vehicle according to claim 1, characterized in that: The steps of calculating the gradient of the loss function using a gradient descent method and optimizing the first-order equivalent circuit model along the descending direction of the gradient until a convergence condition is met include: The first-order equivalent circuit model is optimized according to the following formula: Where L represents the loss function, G represents the inverse of the equivalent capacitance parameter of the first-order equivalent circuit model, R represents the equivalent resistance parameter of the first-order equivalent circuit model, η represents the learning rate in the gradient descent direction, k represents the number of iterations, n represents the number of time nodes of the predicted voltage in the target segment, and i represents the time node of the i-th predicted voltage in the target segment. Represents the voltage prediction value of the i-th time node, V out (t i ) represents the true value of the voltage at the i-th time node; And convergence is completed when the loss function is less than a preset first threshold or the number of iterations is greater than a preset second threshold.

5. The capacitance calibration method for a supercapacitor energy supply system in a vehicle according to claim 1, characterized in that: The step of calculating a final calibration result of the supercapacitor system capacitance based on the plurality of preliminary calibration results of the supercapacitor system capacitance includes: The average of all preliminary calibration results of the supercapacitor system capacitance within a day is used as the final calibration result of the supercapacitor system capacitance on that day; According to the final calibration results of the supercapacitor system capacitance value for each day in a month, the average value is calculated as the final calibration result of the supercapacitor system capacitance value for that month; According to the final calibration results of the supercapacitor system capacitance value in each month of the year, the average value is calculated as the final calibration result of the supercapacitor system capacitance value in the year.

Images