Solar power based emergency charging system and method for electric vehicle

Abstract

The application discloses an emergency charging system and method for electric vehicles based on solar power generation, and belongs to the technical field of solar power generation, and solves the problems of poor portability and short service life of the existing emergency charging system for electric vehicles, wherein the system comprises a solar panel array, an interface circuit, a voltage boosting circuit, a voltage stabilizing circuit and a charging gun module; in the application, the emergency charging system for electric vehicles based on solar power generation only needs the solar panel array, the voltage stabilizing circuit and the charging gun module during work, and does not need large-capacity energy storage components, and only needs to maintain the solar panel and the circuit module, so that the portability and service life of the system are remarkably improved; meanwhile, based on the preset voltage value obtained through evaluation, the voltage boosting circuit boosts the low-voltage direct current of the solar panel to a matching value, ensures that the output of the charging gun matches the charging characteristics of the storage battery, and avoids the loss of charging efficiency caused by voltage mismatch.

Description

Technical Field

[0001] This invention belongs to the field of solar power generation technology, specifically relating to an emergency charging system and method for electric vehicles based on solar power generation. Background Technology

[0002] With the increasing global demand for renewable energy, solar energy, as a clean and sustainable energy source, is receiving more and more attention. In the electric vehicle sector, the application of solar power generation technology is becoming increasingly widespread, aiming to provide an environmentally friendly and economical charging solution for electric vehicles. Furthermore, with the rapid development of the global electric vehicle industry, the demand for emergency charging is growing, especially in scenarios without fixed charging facilities. Therefore, efficient and convenient mobile charging systems have become an important research direction.

[0003] In existing technologies, solar-powered emergency charging systems for electric vehicles typically employ a "solar panel energy storage - battery conversion and output" architecture. This involves converting solar energy into electrical energy via solar panels, storing it in a battery, and then controlling the output through a battery management system before finally using an inverter to supply 220V AC power for charging the electric vehicle. In these systems, the battery is the core component for energy storage and output, but its physical characteristics present several challenges.

[0004] 1) High weight and volume costs: Energy storage devices such as lead-acid batteries or lithium batteries are relatively heavy. When integrated into a mobile charging system, they significantly increase the load on the equipment and reduce portability. This is especially true in scenarios such as emergency rescue where frequent handling is required, which greatly increases labor costs and transportation difficulties.

[0005] 2) Significant lifecycle costs: Battery capacity decreases with the number of charge-discharge cycles, replacement cycles are short, and the disposal of used batteries involves environmental pollution risks, increasing the economic and environmental pressures of the system throughout its entire lifecycle. It is difficult to meet the reliable power supply requirements in emergency scenarios.

[0006] To address the issues of poor portability and short lifespan of batteries in existing electric vehicle emergency charging systems, we propose an electric vehicle emergency charging system and method based on solar power generation. Summary of the Invention

[0007] The purpose of this invention is to address the shortcomings of existing technologies by providing an emergency charging system and method for electric vehicles based on solar power generation, which solves the problems of poor portability and short service life of batteries in existing emergency charging systems for electric vehicles.

[0008] This invention is implemented as follows: an emergency charging system for electric vehicles based on solar power generation, the system comprising:

[0009] A solar panel array, comprising at least one set of solar panels, for converting solar energy into direct current electrical energy;

[0010] An interface circuit, electrically connected to the solar panel array, is used to input the DC power output from the solar panels into a boost circuit.

[0011] The boost circuit is electrically connected to the interface circuit. It evaluates and analyzes the power battery based on the battery state parameters, outputs the preset voltage value of the power battery with different access types, and boosts the DC power to the preset voltage value.

[0012] The voltage regulator circuit has its input terminal connected to the boost circuit and its output terminal connected to the charging gun module. It is used to collect the output voltage and light intensity signal of the charging gun module in real time, and generate an error signal based on the voltage error between the output voltage and the preset voltage value. Using the light intensity signal and the error signal as input, it executes the pre-built voltage regulation response model, outputs dynamic PWM control parameters, and performs real-time voltage regulation based on the PWM control parameters.

[0013] The charging gun module is electrically connected to the electric vehicle's power battery. It is used to provide power to the power battery and collect the battery status parameters in real time, feeding the battery status parameters back to the boost circuit.

[0014] Preferably, the solar panel array comprises:

[0015] The signal sampling unit is used to collect the light intensity signal, single-panel output voltage and single-panel output power on the solar panel side in real time.

[0016] At least one set of solar panels, which are connected in parallel in multiple stages via a common ground bus. The solar panels are connected in series with Schottky diodes to prevent reverse connection of the solar panels.

[0017] The isolation protection unit communicates with the signal sampling unit, judges the real-time status of the solar panel based on the collected light intensity signal, single-board output voltage and single-board output power, and triggers the fault board isolation command based on the real-time status judgment result of the solar panel.

[0018] The dynamic current sharing unit is used to obtain the real-time status judgment result of the solar panel and adjust the PWM duty cycle of the solar panel based on the preset dynamic current sharing strategy to achieve current balance.

[0019] Preferably, the method for determining the real-time status of a solar panel based on the collected light intensity signal, single-panel output voltage, and single-panel output power includes:

[0020] The system loads the light intensity signal, single-board output voltage, and single-board output power, and integrates these signals into a solar panel state set.

[0021] The solar panel state set is preprocessed, which involves Kalman filtering and normalization, and the preprocessed solar panel state set is output.

[0022] The parameters within the solar panel state set are timestamped, and a convolutional neural network model is used to extract the feature factors of the solar panel state set, resulting in a feature factor set. These feature factors include a solar power correlation factor, a voltage stability factor, and a power deviation factor. The solar power correlation factor uses the Pearson correlation coefficient between solar illumination and power. This indicates that the voltage stability factor is calculated using the standard deviation of the single-board output voltage within a preset acquisition period. This indicates that the power deviation factor is the deviation rate between the output power and the rated power of a single board. express;

[0023] The feature factor set is loaded, the feature factors are weighted using principal component analysis, and the feature factor set is weighted and summed using the feature factor weight coefficients to calculate the real-time state value of the solar panel.

[0024] The real-time status value of the solar panel is calculated using the following formula:

[0025]

[0026]

[0027]

[0028] in, This indicates the real-time status value of the solar panel. These are the basic coefficients for weighting. These represent the weighting coefficients of the illumination power correlation factor, voltage stability factor, and power deviation factor, respectively. These are the current feature factor value and the target feature factor value, respectively. These represent the allowable deviation threshold and the penalty coefficient, respectively. The Pearson correlation coefficient represents the relationship between illumination and power. This represents the standard deviation of the single-board output voltage within the preset acquisition period. This indicates the deviation rate between the output power and the rated power of a single board.

[0029] The system acquires the real-time status value of the solar panel and determines whether the real-time status value of the solar panel exceeds a preset status threshold. If it does not exceed the preset status threshold, a solar panel operation fault alarm is triggered.

[0030] Preferably, the method for adjusting the PWM duty cycle of the solar panel based on a preset dynamic current sharing strategy includes:

[0031] Obtain the real-time status value of the solar panels in the solar panel array, select the solar panel with the highest current real-time status value as the main output panel, and the remaining solar panels as the secondary output panels, and use the single-board output current of the main output panel as the target output current of the secondary output panels.

[0032] An equivalent circuit model of the solar panel is established based on a preset dynamic current sharing strategy. The equivalent circuit model is used to determine the current deviation between the target output current and the real-time output current of the output sub-panel, and to determine whether the current deviation exceeds the preset current deviation threshold.

[0033] If the current deviation exceeds the preset current deviation threshold, the PWM duty cycle of the output sub-board is increased based on the equivalent circuit model.

[0034] If the current deviation does not exceed the preset current deviation threshold, maintain the current output current of the mainboard. Construct the equivalent circuit model of the dynamic current sharing strategy based on the gradient descent optimization network. First, construct the state input model, as follows:

[0035]

[0036] in, Indicates the first Sub-plate in The output at any given time is the PWM duty cycle adjustment value. , This represents the weight matrix of the gradient descent optimization network. Indicates the first Sub-plate in The state input vector at time t. , The bias matrix of the gradient descent optimization network is represented by the energy function of the sub-board, which is defined based on the current deviation between the sub-board and the main board. The calculation method is as follows:

[0037]

[0038] in, express Time of the first The nonlinear energy function of the sub-plate, express Time of the first The current of the sub-plate, express The motherboard's current at all times, This represents the control deviation factor. Indicates the current deviation factor. express Time of the first PWM duty cycle of the sub-board express The PWM duty cycle of the motherboard is monitored at all times. Based on a nonlinear energy function, a gradient descent optimization network is used to construct a gradient update rule, which is as follows:

[0039]

[0040] in, This represents the parameters of the bias matrix and weight matrix in the state input model. Indicates the learning rate. Indicates the gradient momentum coefficient. This represents the bias matrix and weight matrix parameters in the input model at the previous time step. Indicates the update symbol. Indicates the parameter The gradient, the updated bias matrix and weight matrix parameters in the input model are input again, and the output PWM duty cycle adjustment is used to adjust the PWM duty cycle of the sub-board to achieve current balancing.

[0041] Preferably, the method for evaluating and analyzing power batteries based on battery state parameters includes:

[0042] The battery status parameters are loaded, and the moving average filtering method is used to filter the battery status parameters to eliminate high-frequency noise in the battery status parameters.

[0043] The battery state parameters after eliminating high-frequency noise are expressed as follows:

[0044]

[0045]

[0046] in, These represent the battery status parameters after high-frequency noise elimination and the input values ​​of the battery status parameters, respectively. The bandwidth of the moving average filter. The input frequency for battery state parameters. This represents the frequency response of the moving average filter. Indicates the frequency of the bandwidth;

[0047] The battery state parameters after high-frequency noise elimination are obtained, and the battery state parameters are discretized to obtain discrete values. Discrete cosine transform is used when discretizing the battery state parameters.

[0048] Load at least one set of discrete parameter values, construct a weight vector of the discrete parameter values ​​based on the fuzzy comprehensive scoring model, and define the membership function of the parameter type corresponding to the weight vector. The fuzzy comprehensive scoring model calculates the comprehensive membership of the battery state parameters and the battery type based on fuzzy operation comprehensive decision.

[0049] Select the battery type with the highest comprehensive membership degree with the battery state parameters as the type of power battery to be connected.

[0050] Based on the battery status parameters corresponding to the battery type, the preset voltage value of the power battery of the access type is retrieved, and the preset voltage value of the power battery of different access types is output.

[0051] Preferably, the voltage regulator circuit includes:

[0052] The error extraction unit is used to collect the output voltage and light intensity signals of the charging gun module in real time, and generate an error signal based on the voltage error between the output voltage and the preset voltage value.

[0053] The model building unit is used to build a voltage regulation response model based on the PID algorithm, and iteratively trains the voltage regulation response model using historical error signals and historical illumination intensity signals, and outputs a converged voltage regulation response model.

[0054] The voltage regulation unit is used to take the light intensity signal and error signal as input, execute the pre-built voltage regulation response model, output dynamic PWM control parameters, and perform real-time voltage regulation based on the PWM control parameters.

[0055] Preferably, the voltage regulation response model training method includes:

[0056] Acquire historical error signals and historical illumination intensity signals, and annotate abnormal data in the historical error signals and historical illumination intensity signals;

[0057] Abnormal noise points in historical error signals and historical illumination intensity signals are removed and missing values ​​are added. The historical error signals and historical illumination intensity signals are integrated into a modeling sample set, which is then divided into a training set and a test set.

[0058] With the goal of minimizing voltage regulation error, a joint loss function for the voltage regulation response model is preset;

[0059] Load the pre-built voltage regulation response model, initialize the PID parameters and illumination compensation coefficient of the voltage regulation response model, and preset the training rounds, parameter optimization methods, and early stopping mechanism of the voltage regulation response model;

[0060] Obtain the training set, iteratively train the voltage regulation response model based on the training set until convergence, and output the converged voltage regulation response model;

[0061] The voltage regulation response model is tested using a test set, the test results are output, and the test error between the test results and the actual results is calculated using a joint loss function.

[0062] Determine whether the test error exceeds the preset test accuracy. If it does, output the converged voltage regulation response model.

[0063] Preferably, the joint loss function of the voltage regulation response model is expressed as:

[0064]

[0065] in, This represents the joint loss function of the voltage regulation response model, while These are the error signal weighting coefficient, the illumination intensity weighting coefficient, and the PWM parameter change rate weighting coefficient, respectively. These represent the error signal, illumination intensity, and PWM parameter change rate, respectively. Indicates the sample size.

[0066] On the other hand, the present invention also provides an emergency charging method for electric vehicles based on solar power generation, the method comprising:

[0067] The solar panel array converts light energy into DC power, and collects the light intensity signal, single panel output voltage and single panel output power on the solar panel side in real time. The solar panel current balance is achieved based on the dynamic current sharing strategy, and the DC power output by the solar panel is connected to the boost circuit.

[0068] The power battery is evaluated and analyzed based on battery state parameters, and the preset voltage value of the power battery with different access types is output, and the DC power is boosted to the preset voltage value.

[0069] The system collects the output voltage and light intensity signals of the charging gun module in real time, and generates an error signal based on the voltage error between the output voltage and the preset voltage value. Using the light intensity signal and the error signal as input, it executes a pre-built voltage regulation response model, outputs dynamic PWM control parameters, and performs real-time voltage regulation based on the PWM control parameters.

[0070] The charging gun module is electrically connected to the electric vehicle's power battery, providing power to the battery and collecting the battery status parameters in real time, feeding these parameters back to the boost circuit.

[0071] Preferably, the method for achieving solar panel current balancing based on a dynamic current sharing strategy includes:

[0072] Real-time acquisition of solar panel irradiance signal, single panel output voltage, and single panel output power;

[0073] The real-time status of the solar panel is determined based on the collected light intensity signal, single-board output voltage and single-board output power, and the fault board isolation command is triggered based on the real-time status determination result of the solar panel.

[0074] The system obtains the real-time status judgment results of the solar panel and adjusts the PWM duty cycle of the solar panel based on the preset dynamic current sharing strategy to achieve current balance.

[0075] Compared with the prior art, the embodiments of this application have the following main advantages:

[0076] In this embodiment of the invention, the provided solar-powered electric vehicle emergency charging system only requires a solar panel array, a voltage regulator circuit, and a charging gun module to operate. It does not have a large-capacity energy storage component and only requires maintenance of the solar panel and the circuit module, thereby significantly improving the system's portability and service life. At the same time, based on the preset voltage value obtained from the evaluation, the boost circuit boosts the low-voltage DC power from the solar panel to the matching value, ensuring that the output of the charging gun matches the charging characteristics of the battery and avoiding charging efficiency loss due to voltage mismatch.

[0077] In this embodiment of the invention, a solar panel array is provided. The solar panel array consists of a signal sampling unit, an isolation protection unit, and a dynamic current sharing unit. The signal sampling unit can collect real-time data on illumination, voltage, and power. The isolation protection unit identifies and isolates abnormal panels by judging their status, thus preventing a single faulty panel from affecting the overall power generation efficiency of the solar panel array. The dynamic current sharing unit adjusts the PWM duty cycle according to the status results, making the output current of each solar panel more balanced and avoiding the inefficient mode of strong panels driving weak panels in the solar panel array.

[0078] In this embodiment of the invention, a voltage regulator circuit is provided. Based on a pre-built voltage regulation response model, the voltage regulator circuit can dynamically optimize PID parameters, overcoming the problems of traditional PID parameters being fixed and unable to adapt to the differences in system dynamic characteristics caused by changes in illumination, as well as large voltage fluctuations. At the same time, the voltage regulator circuit consists of an error extraction unit, a model building unit, and a voltage regulation unit. The high-precision sampling of the error extraction unit reduces the noise input of the model building unit, avoiding model mistraining due to data errors. Moreover, the high-frequency PWM regulation of the voltage regulation unit reduces the parameter dependence on components such as inductors and capacitors in the boost circuit, reducing the hardware size and improving system portability. By predicting and feedforward compensating for illumination intensity, the control parameters can be adjusted in advance, avoiding the lag of adjustments made only by the feedback loop, and improving the stability of the charging voltage.

[0079] In this embodiment of the invention, the real-time status of the solar panel is determined by the collected light intensity signal, single-panel output voltage, and single-panel output power. This achieves multi-parameter fusion correlation analysis, which helps to accurately assess the real-time status of the solar panel. Furthermore, through Kalman filtering for noise reduction, timestamp alignment, and feature processing, the real-time status value of the solar panel is rapidly assessed, avoiding power generation efficiency loss due to response lag and improving the system's operational capability in complex environments and industrial control settings.

[0080] In this embodiment of the invention, when adjusting the PWM duty cycle of the solar panel based on a preset dynamic current sharing strategy, the main board with the highest real-time state value is used as the benchmark. By increasing the duty cycle of the secondary board, the current alignment of the main and secondary boards is achieved, thereby ensuring that both the main and secondary boards operate in the optimal current range. This avoids efficiency loss caused by overload of the strong board or underload of the weak board. At the same time, an equivalent circuit model is used to determine the current deviation between the target output current and the real-time output current of the output secondary board. Through the dynamic current sharing strategy, the risk of chain reaction caused by the failure of the secondary board due to uneven current can be avoided in advance. This allows the embodiment of the invention to adapt to different operating conditions and ensure system stability.

[0081] In this embodiment of the invention, the preset voltage value of the power battery of the access type is retrieved based on the battery type corresponding to the battery state parameters. On the one hand, this can reduce charging losses caused by voltage mismatch, and on the other hand, accurate voltage matching can avoid overvoltage or undervoltage, reducing the risk of battery damage. At the same time, when calculating the comprehensive membership degree of battery state parameters and battery type, the membership degree function can describe the partial matching relationship between parameters and type, thereby avoiding the black-and-white misjudgment of traditional threshold method. Attached Figure Description

[0082] Figure 1 This is a schematic diagram of the structure of the solar-powered emergency charging system for electric vehicles provided by the present invention.

[0083] Figure 2 A schematic diagram of the implementation process of a method for determining the real-time status of a solar panel based on the collected light intensity signal, single-panel output voltage, and single-panel output power is shown.

[0084] Figure 3 A schematic diagram of the implementation process of adjusting the PWM duty cycle of a solar panel based on a preset dynamic current sharing strategy is shown.

[0085] Figure 4 A schematic diagram of the implementation process of the power battery evaluation and analysis method based on battery state parameters is shown.

[0086] Figure 5 A schematic diagram of the implementation process of the voltage regulation response model training method is shown.

[0087] Figure 6 A schematic diagram illustrating the implementation process of an emergency charging method for electric vehicles based on solar power generation is shown.

[0088] In the diagram: 100 - Solar panel array, 200 - Interface circuit, 300 - Boost circuit, 400 - Voltage regulator circuit, 500 - Charging gun module. Detailed Implementation

[0089] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this application belongs; the terminology used herein in the specification of the application is for the purpose of describing particular embodiments only and is not intended to be limiting of the application; the terms "comprising" and "having," and any variations thereof, in the specification, claims, and foregoing drawings of this application are intended to cover non-exclusive inclusion. The terms "first," "second," etc., in the specification, claims, or foregoing drawings of this application are used to distinguish different objects, not to describe a particular order.

[0090] To address the issues of poor portability and short lifespan of batteries in existing electric vehicle emergency charging systems, we propose an electric vehicle emergency charging system and method based on solar power generation. In short, the system consists of a solar panel array 100, an interface circuit 200, a boost circuit 300, a voltage regulator circuit 400, and a charging gun module 500. During operation, the solar panel array 100 converts solar energy into DC power. The interface circuit 200 connects the DC power output from the solar panels to the boost circuit 300. The boost circuit 300 evaluates and analyzes the power battery based on battery state parameters, outputs preset voltage values ​​for different battery connection types, and boosts the DC power to the preset voltage values. The voltage regulator circuit 400 collects the output voltage and light intensity signals from the charging gun module 500 in real time, and generates an error signal based on the voltage error between the output voltage and the preset voltage value. Using the light intensity signal and the error signal as input, it executes a pre-constructed voltage regulation response model, outputs dynamic PWM control parameters, and performs real-time voltage regulation based on the PWM control parameters. Finally, the charging gun module 500 provides power to the power battery. In this embodiment of the invention, the provided solar-powered electric vehicle emergency charging system only requires a solar panel array 100, a voltage regulator circuit 400, and a charging gun module 500 when in operation. It does not have a large-capacity energy storage component and only requires maintenance of the solar panel and the circuit module, thereby significantly improving the portability and service life of the system. At the same time, based on the preset voltage value obtained from the evaluation, the boost circuit 300 boosts the low-voltage DC power of the solar panel to the matching value, ensuring that the output of the charging gun matches the charging characteristics of the battery and avoiding charging efficiency loss due to voltage mismatch.

[0091] This invention provides a schematic diagram of the implementation process of an emergency charging system for electric vehicles based on solar power generation. Figure 1 A schematic diagram illustrating the implementation process of a solar-powered emergency charging system for electric vehicles is shown. Specifically, the solar-powered emergency charging system for electric vehicles includes:

[0092] Solar panel array 100 includes at least one set of solar panels for converting solar energy into direct current electrical energy;

[0093] In this embodiment of the invention, the solar panel array 100 includes:

[0094] The signal sampling unit is used to collect the light intensity signal, single-panel output voltage and single-panel output power on the solar panel side in real time.

[0095] At least one set of solar panels, which are connected in parallel in multiple stages via a common ground bus. The solar panels are connected in series with Schottky diodes to prevent reverse connection of the solar panels.

[0096] It should be noted that the signal sampling unit is embedded in the side of the solar panel, and the signal sampling unit includes a voltage sensor, a current sensor, and a light sensor. The light intensity is collected by the light sensor, while the output voltage of the single panel is collected in real time by the voltage sensor in conjunction with the analog-to-digital converter.

[0097] The isolation protection unit is communicatively connected to the signal sampling unit. Based on the collected light intensity signal, single-board output voltage, and single-board output power, it determines the real-time status of the solar panel and triggers a faulty board isolation command based on the real-time status determination result of the solar panel. In this embodiment of the invention, the signal sampling unit can collect light, voltage, and power data in real time, while the isolation protection unit identifies and isolates abnormal boards through status determination, avoiding the impact of a single group of faulty boards on the overall power generation efficiency of the solar panel array 100. Furthermore, the dynamic current sharing unit adjusts the PWM duty cycle according to the status result, making the output current of each solar panel more balanced and avoiding the inefficient mode of strong boards driving weak boards in the solar panel array 100.

[0098] The dynamic current sharing unit is used to obtain the real-time status judgment result of the solar panel and adjust the PWM duty cycle of the solar panel based on the preset dynamic current sharing strategy to achieve current balance.

[0099] In this embodiment of the invention, a solar panel array 100 is provided. The solar panel array 100 consists of a signal sampling unit, an isolation protection unit, and a dynamic current sharing unit. The signal sampling unit can collect real-time data on light, voltage, and power. The isolation protection unit identifies and isolates abnormal panels by judging their status, thus preventing a single faulty panel from affecting the overall power generation efficiency of the solar panel array 100. The dynamic current sharing unit adjusts the PWM duty cycle according to the status results, making the output current of each solar panel more balanced, thus avoiding the inefficient mode of strong panels driving weak panels in the solar panel array 100.

[0100] The interface circuit 200 is electrically connected to the solar panel array 100 and is used to input the DC power output by the solar panel into the boost circuit 300.

[0101] The boost circuit 300 is electrically connected to the interface circuit 200. It evaluates and analyzes the power battery based on the battery status parameters, outputs the preset voltage value of the power battery with different access types, and boosts the DC power to the preset voltage value.

[0102] The voltage regulator circuit 400 has its input terminal connected to the boost circuit 300 and its output terminal connected to the charging gun module 500. It is used to collect the output voltage and light intensity signal of the charging gun module 500 in real time, and generate an error signal based on the voltage error between the output voltage and the preset voltage value. Using the light intensity signal and the error signal as input, it executes the pre-built voltage regulation response model, outputs dynamic PWM control parameters, and performs real-time voltage regulation based on the PWM control parameters.

[0103] The voltage regulator circuit 400 includes:

[0104] The error extraction unit is used to collect the output voltage and light intensity signals of the charging gun module 500 in real time, and generate an error signal based on the voltage error between the output voltage and the preset voltage value.

[0105] The model building unit is used to build a voltage regulation response model based on the PID algorithm, and iteratively trains the voltage regulation response model using historical error signals and historical illumination intensity signals, and outputs a converged voltage regulation response model.

[0106] The voltage regulation unit is used to take the light intensity signal and error signal as input, execute the pre-built voltage regulation response model, output dynamic PWM control parameters, and perform real-time voltage regulation based on the PWM control parameters.

[0107] In this embodiment of the invention, a voltage regulator circuit 400 is provided. The voltage regulator circuit 400 can dynamically optimize PID parameters based on a pre-built voltage regulation response model, which can overcome the problems of traditional PID parameters being fixed and difficult to adapt to the differences in system dynamic characteristics caused by changes in illumination and large voltage fluctuations. At the same time, the voltage regulator circuit 400 is composed of an error extraction unit, a model building unit, and a voltage regulation unit. The high-precision sampling of the error extraction unit reduces the noise input of the model building unit and avoids model mistraining caused by data errors. Moreover, the high-frequency PWM regulation of the voltage regulation unit reduces the parameter dependence on components such as inductors and capacitors of the boost circuit 300, reduces the hardware size, and improves the portability of the system.

[0108] The charging gun module 500 is electrically connected to the electric vehicle's power battery. It is used to provide power to the power battery and collect the battery status parameters of the power battery in real time, and feed the battery status parameters back to the boost circuit 300.

[0109] In this embodiment of the invention, the provided solar-powered electric vehicle emergency charging system only requires a solar panel array 100, a voltage regulator circuit 400, and a charging gun module 500 when in operation. It does not have a large-capacity energy storage component and only requires maintenance of the solar panel and the circuit module, thereby significantly improving the portability and service life of the system. At the same time, based on the preset voltage value obtained from the evaluation, the boost circuit 300 boosts the low-voltage DC power of the solar panel to the matching value, ensuring that the output of the charging gun matches the charging characteristics of the battery and avoiding charging efficiency loss due to voltage mismatch.

[0110] This invention provides a method for determining the real-time status of a solar panel based on the acquired light intensity signal, panel output voltage, and panel output power. Figure 2 A schematic diagram illustrating the implementation process of a method for determining the real-time status of a solar panel based on acquired light intensity signals, single-panel output voltage, and single-panel output power is shown. The method specifically includes:

[0111] S101, load the light intensity signal, single-board output voltage and single-board output power, and integrate the light intensity signal, single-board output voltage and single-board output power into a solar panel state set. In this embodiment, the solar panel state set integrates the light intensity signal, single-board output voltage and single-board output power, thereby realizing multi-source data fusion, and thus being able to comprehensively capture and reflect the state characteristics of the solar panel.

[0112] S102, preprocess the solar panel state set, wherein the preprocessing method is Kalman filtering and normalization, and output the preprocessed solar panel state set;

[0113] S103, timestamp alignment is performed on the parameters within the solar panel state set to ensure simultaneous feature extraction and calculation of multi-source data parameters, avoiding feature distortion caused by asynchronous sampling. A convolutional neural network model is used to extract feature factors from the solar panel state set, resulting in a feature factor set. These feature factors include a solar power correlation factor, a voltage stability factor, and a power deviation factor. The solar power correlation factor uses the Pearson correlation coefficient between solar power and solar energy. This indicates that the voltage stability factor is calculated using the standard deviation of the single-board output voltage within a preset acquisition period. This indicates that the power deviation factor is the deviation rate between the output power and the rated power of a single board. express;

[0114] S104, Load the feature factor set, use principal component analysis to assign weights to the feature factors, and combine the feature factor weight coefficients to perform a weighted summation of the feature factor set to calculate the real-time state value of the solar panel.

[0115] The real-time status value of the solar panel is calculated using the following formula:

[0116]

[0117]

[0118]

[0119] in, This indicates the real-time status value of the solar panel. These are the basic coefficients for weighting. These represent the weighting coefficients of the illumination power correlation factor, voltage stability factor, and power deviation factor, respectively. These are the current feature factor value and the target feature factor value, respectively. These represent the allowable deviation threshold and the penalty coefficient, respectively. The Pearson correlation coefficient represents the relationship between illumination and power. This represents the standard deviation of the single-board output voltage within the preset acquisition period. This indicates the deviation rate between the output power and the rated power of a single board.

[0120] S105, obtain the real-time status value of the solar panel, determine whether the real-time status value of the solar panel exceeds the preset status threshold, if it does not exceed the preset status threshold (in this embodiment, the preset status threshold can be 0.6-0.65), trigger the solar panel operation fault alarm.

[0121] In this embodiment of the invention, the real-time status of the solar panel is determined by the collected light intensity signal, single-panel output voltage, and single-panel output power. This achieves multi-parameter fusion correlation analysis, which helps to accurately assess the real-time status of the solar panel. Furthermore, through Kalman filtering for noise reduction, timestamp alignment, and feature processing, the real-time status value of the solar panel is rapidly assessed, avoiding power generation efficiency loss due to response lag and improving the system's operational capability in complex environments and industrial control settings.

[0122] This invention provides a method for adjusting the PWM duty cycle of a solar panel based on a preset dynamic current sharing strategy. Figure 3 A schematic diagram illustrating the implementation process of a method for adjusting the PWM duty cycle of a solar panel based on a preset dynamic current sharing strategy is shown. The method specifically includes:

[0123] S201, obtain the real-time status value of the solar panels in the solar panel array 100, select the solar panel with the highest current real-time status value as the main output panel, and the remaining solar panels as the secondary output panels, and use the single-board output current of the main output panel as the target output current of the secondary output panels.

[0124] S202, Based on the preset dynamic current sharing strategy, establish the equivalent circuit model of the solar panel, and use the equivalent circuit model to determine the current deviation between the target output current and the real-time output current of the output sub-board.

[0125] S203, determine whether the current deviation exceeds the preset current deviation threshold;

[0126] S204, if the current deviation exceeds the preset current deviation threshold, the PWM duty cycle of the output sub-board is increased based on the equivalent circuit model. The equivalent circuit model of the dynamic current sharing strategy is constructed based on the gradient descent optimization network. First, the state input model is constructed, as follows:

[0127]

[0128] in, Indicates the first Sub-plate in The output at any given time is the PWM duty cycle adjustment value. , This represents the weight matrix of the gradient descent optimization network. Indicates the first Sub-plate in The state input vector at time t. , The bias matrix of the gradient descent optimization network is represented by the energy function of the sub-board, which is defined based on the current deviation between the sub-board and the main board. The calculation method is as follows:

[0129]

[0130] in, express Time of the first The nonlinear energy function of the sub-plate, express Time of the first The current of the sub-plate, express The motherboard's current at all times, This represents the control deviation factor. Indicates the current deviation factor. express Time of the first PWM duty cycle of the sub-board express The PWM duty cycle of the motherboard is monitored at all times. Based on a nonlinear energy function, a gradient descent optimization network is used to construct a gradient update rule, which is as follows:

[0131]

[0132] in, This represents the parameters of the bias matrix and weight matrix in the state input model. Indicates the learning rate. Indicates the gradient momentum coefficient. This represents the bias matrix and weight matrix parameters in the input model at the previous time step. Indicates the update symbol. Indicates the parameter The gradient is updated, and the bias matrix and weight matrix parameters in the updated input model are input again to output the PWM duty cycle adjustment amount, which is used to adjust the PWM duty cycle of the sub-board to achieve current balance.

[0133] S205, if the current deviation does not exceed the preset current deviation threshold, maintain the current single-board output current of the output motherboard.

[0134] In this embodiment of the invention, when adjusting the PWM duty cycle of the solar panel based on a preset dynamic current sharing strategy, the main board with the highest real-time state value is used as the benchmark. By increasing the duty cycle of the secondary board, the current alignment of the main and secondary boards is achieved, thereby ensuring that both the main and secondary boards operate in the optimal current range. This avoids efficiency loss caused by overload of the strong board or underload of the weak board. At the same time, an equivalent circuit model is used to determine the current deviation between the target output current and the real-time output current of the output secondary board. Through the dynamic current sharing strategy, the risk of chain reaction caused by the failure of the secondary board due to uneven current can be avoided in advance. This allows the embodiment of the invention to adapt to different operating conditions and ensure system stability.

[0135] This invention provides a method for evaluating and analyzing power batteries based on battery state parameters. Figure 4 This diagram illustrates the implementation flow of a method for evaluating and analyzing power batteries based on battery state parameters. The method specifically includes:

[0136] S301, loads battery state parameters, and uses the moving average filtering method to filter the battery state parameters to eliminate high-frequency noise in the battery state parameters. The moving average filtering method balances noise suppression and signal resolution, which can eliminate random noise and retain the true trend of battery state changes, thereby avoiding feature loss due to excessive smoothing.

[0137] The battery state parameters after eliminating high-frequency noise are expressed as follows:

[0138]

[0139]

[0140] in, These represent the battery status parameters after high-frequency noise elimination and the input values ​​of the battery status parameters, respectively. The bandwidth of the moving average filter. The input frequency for battery state parameters. This represents the frequency response of the moving average filter. Indicates the frequency of the bandwidth;

[0141] S302, obtain the battery state parameters after eliminating high-frequency noise, and perform discretization processing on the battery state parameters to obtain discrete values ​​of the parameters. When discretizing the battery state parameters, the Discrete Cosine Transform (DCT) is used. The DCT is more sensitive to the energy distribution of the signal and can concentrate the main changes of the battery state parameters in a few discrete coefficients, avoiding the loss of key information caused by equidistant segmentation, and reducing the error between the discrete values ​​of the parameters and the original continuous values.

[0142] S303, load at least one set of discrete parameter values, construct a weight vector of the discrete parameter values ​​based on the fuzzy comprehensive scoring model, and define the membership function of the parameter type corresponding to the weight vector. The fuzzy comprehensive scoring model calculates the comprehensive membership degree of the battery state parameters and the battery type based on fuzzy operation comprehensive decision. In this embodiment of the invention, the membership function can describe the partial matching relationship between the parameters and the type, thereby avoiding the black-and-white misjudgment of the traditional threshold method.

[0143] S304, Select the battery type with the highest comprehensive membership degree with the battery state parameters as the type of power battery to be connected;

[0144] S305 retrieves the preset voltage value of the power battery of the access type based on the battery type corresponding to the battery status parameters, and outputs the preset voltage value of the power battery of different access types.

[0145] In this embodiment of the invention, the preset voltage value of the power battery of the access type is retrieved based on the battery type corresponding to the battery state parameters. On the one hand, this can reduce charging losses caused by voltage mismatch, and on the other hand, accurate voltage matching can avoid overvoltage or undervoltage, reducing the risk of battery damage. At the same time, when calculating the comprehensive membership degree of battery state parameters and battery type, the membership degree function can describe the partial matching relationship between parameters and type, thereby avoiding the black-and-white misjudgment of traditional threshold method.

[0146] This invention provides a method for training a voltage regulation response model. Figure 5 The diagram illustrates the implementation flow of the voltage regulation response model training method, which specifically includes:

[0147] S401: Acquire historical error signals and historical illumination intensity signals, and annotate abnormal data in the historical error signals and historical illumination intensity signals;

[0148] S402, remove abnormal noise points from historical error signals and historical illumination intensity signals, and add missing values. Integrate historical error signals and historical illumination intensity signals into a modeling sample set. Divide the modeling sample set into a training set and a test set. The ratio of the training set to the test set can be 3:1.

[0149] S403, with the goal of minimizing voltage regulation error, presets the joint loss function of the voltage regulation response model;

[0150] S404: Load the pre-built voltage regulation response model, initialize the PID parameters and illumination compensation coefficient of the voltage regulation response model, and preset the training rounds, parameter optimization methods, and early stopping mechanism of the voltage regulation response model. The training rounds can be 100-120 times, and the parameter optimization method can be the Adam optimizer.

[0151] S405, Obtain the training set, iteratively train the voltage regulation response model based on the training set until convergence, and output the converged voltage regulation response model;

[0152] S406 uses a test set to test the voltage regulation response model, outputs the test results, and calculates the test error between the test results and the actual results through a joint loss function;

[0153] S407, determine whether the test error exceeds the preset test accuracy;

[0154] S408, if the test accuracy exceeds the preset limit, outputs a converged regulated voltage response model;

[0155] If the preset test accuracy is not exceeded, return to S405 and continue iterative training of the model.

[0156] In this embodiment, the joint loss function of the voltage regulation response model is expressed as:

[0157]

[0158] in, This represents the joint loss function of the voltage regulation response model, while These are the error signal weighting coefficient, the illumination intensity weighting coefficient, and the PWM parameter change rate weighting coefficient, respectively. These represent the error signal, illumination intensity, and PWM parameter change rate, respectively. Indicates the sample size.

[0159] In this embodiment of the invention, a training method for a voltage regulation response model is provided. The voltage regulation response model dynamically optimizes PID parameters through training with historical data. Based on real-time acquisition of the output voltage and light intensity signal of the charging gun module 500, the voltage regulation response model can accurately calculate the error between the output voltage and the preset voltage, and generate an error signal accordingly to achieve precise voltage control. It can also dynamically adjust the PWM (pulse width modulation) control parameters according to the change of light intensity to adapt to different lighting conditions and ensure the stability of the output voltage.

[0160] On the other hand, embodiments of the present invention also provide an emergency charging method for electric vehicles based on solar power generation. Figure 6 A schematic diagram illustrating the implementation process of an emergency charging method for electric vehicles based on solar power generation is shown. Specifically, the emergency charging method for electric vehicles based on solar power generation includes:

[0161] S10 converts light energy into DC power through solar panel array 100, collects the light intensity signal, single panel output voltage and single panel output power on the solar panel side in real time, realizes solar panel current balance based on dynamic current sharing strategy, and connects the DC power output by the solar panel to boost circuit 300.

[0162] S20 evaluates and analyzes the power battery based on battery state parameters, outputs preset voltage values ​​for power batteries with different access types, and boosts the DC power to the preset voltage value;

[0163] S30: Real-time acquisition of output voltage and light intensity signals of charging gun module 500, and generation of error signal based on voltage error between output voltage and preset voltage value. Using light intensity signal and error signal as input, execute pre-built voltage regulation response model, output dynamic PWM control parameters, and perform real-time voltage regulation based on PWM control parameters.

[0164] S40, the charging gun module 500 is electrically connected to the electric vehicle power battery, provides power to the power battery, and collects the battery status parameters of the power battery in real time, feeding the battery status parameters back to the boost circuit 300.

[0165] In this embodiment, the method for achieving solar panel current balancing based on a dynamic current sharing strategy includes:

[0166] S501 collects real-time light intensity signals, single-panel output voltage, and single-panel output power from the solar panel side.

[0167] S502 determines the real-time status of the solar panel based on the collected light intensity signal, single-board output voltage and single-board output power, and triggers the fault board isolation command based on the real-time status determination result of the solar panel.

[0168] S503 obtains the real-time status judgment result of the solar panel, and adjusts the PWM duty cycle of the solar panel based on the preset dynamic current sharing strategy to achieve current balance.

[0169] In summary, this invention provides an emergency charging system and method for electric vehicles based on solar power generation. In the embodiments of this invention, the provided emergency charging system for electric vehicles based on solar power generation only requires a solar panel array 100, a voltage regulator circuit 400, and a charging gun module 500 to operate. There are no large-capacity energy storage components, and only the solar panels and circuit modules need to be maintained, thereby significantly improving the portability and service life of the system. At the same time, based on the preset voltage value obtained from the evaluation, the boost circuit 300 boosts the low-voltage DC power from the solar panel to the matching value, ensuring that the output of the charging gun matches the charging characteristics of the battery and avoiding charging efficiency loss due to voltage mismatch.

[0170] It should be noted that, for the sake of simplicity, the foregoing embodiments are all described as a series of actions. However, those skilled in the art should understand that the present invention is not limited to the described order of actions, as some steps may be performed in other orders or simultaneously according to the present invention. Furthermore, those skilled in the art should also understand that the embodiments described in the specification are preferred embodiments, and the actions and modules involved are not necessarily essential to the present invention.

[0171] The above embodiments are only used to illustrate the technical solutions of the present invention, and are not intended to limit the scope of protection of the invention. Obviously, the described embodiments are only some embodiments of the present invention, not all embodiments. Based on these embodiments, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention. Although the present invention has been described in detail with reference to the above embodiments, those skilled in the art can still combine, add, delete, or otherwise adjust the features of the various embodiments of the present invention according to the circumstances without conflict or creative effort, thereby obtaining different technical solutions that do not fundamentally depart from the concept of the present invention. These technical solutions also fall within the scope of protection of the present invention.

Claims

1. An emergency charging system for electric vehicles based on solar power generation, characterized in that, The system includes: A solar panel array, comprising at least one set of solar panels, for converting solar energy into direct current electrical energy; An interface circuit, electrically connected to the solar panel array, is used to input the DC power output from the solar panels into a boost circuit. The boost circuit is electrically connected to the interface circuit. It evaluates and analyzes the power battery based on the battery state parameters, outputs the preset voltage value of the power battery with different access types, and boosts the DC power to the preset voltage value. The voltage regulator circuit has its input terminal connected to the boost circuit and its output terminal connected to the charging gun module. It is used to collect the output voltage and light intensity signal of the charging gun module in real time, and generate an error signal based on the voltage error between the output voltage and the preset voltage value. Using the light intensity signal and the error signal as input, it executes the pre-built voltage regulation response model, outputs dynamic PWM control parameters, and performs real-time voltage regulation based on the PWM control parameters. The charging gun module is electrically connected to the electric vehicle's power battery. It is used to provide power to the power battery and collect the battery status parameters in real time, feeding the battery status parameters back to the boost circuit.

2. The emergency charging system for electric vehicles based on solar power generation as described in claim 1, characterized in that: The solar panel array includes: The signal sampling unit is used to collect the light intensity signal, single-panel output voltage and single-panel output power on the solar panel side in real time. At least one set of solar panels, which are connected in parallel in multiple stages via a common ground bus. The solar panels are connected in series with Schottky diodes to prevent reverse connection of the solar panels. The isolation protection unit communicates with the signal sampling unit, judges the real-time status of the solar panel based on the collected light intensity signal, single-board output voltage and single-board output power, and triggers the fault board isolation command based on the real-time status judgment result of the solar panel. The dynamic current sharing unit is used to obtain the real-time status judgment result of the solar panel and adjust the PWM duty cycle of the solar panel based on the preset dynamic current sharing strategy to achieve current balance.

3. The emergency charging system for electric vehicles based on solar power generation as described in claim 2, characterized in that: The method for determining the real-time status of a solar panel based on the collected light intensity signal, single-panel output voltage, and single-panel output power includes: The system loads the light intensity signal, single-board output voltage, and single-board output power, and integrates these signals into a solar panel state set. The solar panel state set is preprocessed, which involves Kalman filtering and normalization, and the preprocessed solar panel state set is output. The parameters within the solar panel state set are timestamped, and a convolutional neural network model is used to extract the feature factors of the solar panel state set, resulting in a feature factor set. These feature factors include a solar power correlation factor, a voltage stability factor, and a power deviation factor. The solar power correlation factor uses the Pearson correlation coefficient between solar illumination and power. This indicates that the voltage stability factor is calculated using the standard deviation of the single-board output voltage within a preset acquisition period. This indicates that the power deviation factor is the deviation rate between the output power and the rated power of a single board. express; The feature factor set is loaded, the feature factors are weighted using principal component analysis, and the feature factor set is weighted and summed using the feature factor weight coefficients to calculate the real-time state value of the solar panel. The system acquires the real-time status value of the solar panel and determines whether the real-time status value of the solar panel exceeds a preset status threshold. If it does not exceed the preset status threshold, a solar panel operation fault alarm is triggered.

4. The emergency charging system for electric vehicles based on solar power generation as described in claim 3, characterized in that: The method for adjusting the PWM duty cycle of the solar panel based on a preset dynamic current sharing strategy includes: Obtain the real-time status value of the solar panels in the solar panel array, select the solar panel with the highest current real-time status value as the main output panel, and the remaining solar panels as the secondary output panels, and use the single-board output current of the main output panel as the target output current of the secondary output panels. An equivalent circuit model of the solar panel is established based on a preset dynamic current sharing strategy. The equivalent circuit model is used to determine the current deviation between the target output current and the real-time output current of the output sub-panel, and to determine whether the current deviation exceeds the preset current deviation threshold. If the current deviation exceeds the preset current deviation threshold, the PWM duty cycle of the output sub-board is increased based on the equivalent circuit model. If the current deviation does not exceed the preset current deviation threshold, maintain the current single-board output current of the output motherboard; The equivalent circuit model of the dynamic current sharing strategy is constructed based on the gradient descent optimization network. First, the state input model is constructed as follows: in, Indicates the first Sub-plate in The output at any given time is the PWM duty cycle adjustment value. , This represents the weight matrix of the gradient descent optimization network. Indicates the first Sub-plate in The state input vector at time t. , The bias matrix of the gradient descent optimization network is represented by the energy function of the sub-board, which is defined based on the current deviation between the sub-board and the main board. The calculation method is as follows: in, express Time of the first The nonlinear energy function of the sub-plate, express Time of the first The current of the sub-plate, express The motherboard's current at all times, This represents the control deviation factor. Indicates the current deviation factor. express Time of the first PWM duty cycle of the sub-board express The PWM duty cycle of the motherboard is monitored at all times. Based on a nonlinear energy function, a gradient descent optimization network is used to construct a gradient update rule, which is as follows: in, This represents the parameters of the bias matrix and weight matrix in the state input model. Indicates the learning rate. Indicates the gradient momentum coefficient. This represents the bias matrix and weight matrix parameters in the input model at the previous time step. Indicates the update symbol. Indicates the parameter The gradient is used to update the bias matrix and weight matrix parameters in the input model. The output PWM duty cycle adjustment is then used to adjust the PWM duty cycle of the sub-board to achieve current balance.

5. The emergency charging system for electric vehicles based on solar power generation as described in claim 1, characterized in that: The method for evaluating and analyzing power batteries based on battery state parameters includes: The battery status parameters are loaded, and the moving average filtering method is used to filter the battery status parameters to eliminate high-frequency noise in the battery status parameters. The battery state parameters after high-frequency noise elimination are obtained, and the battery state parameters are discretized to obtain discrete values. Discrete cosine transform is used when discretizing the battery state parameters. Load at least one set of discrete parameter values, construct a weight vector of the discrete parameter values ​​based on the fuzzy comprehensive scoring model, and define the membership function of the parameter type corresponding to the weight vector. The fuzzy comprehensive scoring model calculates the comprehensive membership of the battery state parameters and the battery type based on fuzzy operation comprehensive decision. Select the battery type with the highest comprehensive membership degree with the battery state parameters as the type of power battery to be connected. Based on the battery status parameters corresponding to the battery type, the preset voltage value of the power battery of the access type is retrieved, and the preset voltage value of the power battery of different access types is output.

6. The emergency charging system for electric vehicles based on solar power generation as described in claim 5, characterized in that: The voltage regulator circuit includes: The error extraction unit is used to collect the output voltage and light intensity signals of the charging gun module in real time, and generate an error signal based on the voltage error between the output voltage and the preset voltage value. The model building unit is used to build a voltage regulation response model based on the PID algorithm, and iteratively trains the voltage regulation response model using historical error signals and historical illumination intensity signals, and outputs a converged voltage regulation response model. The voltage regulation unit is used to take the light intensity signal and error signal as input, execute the pre-built voltage regulation response model, output dynamic PWM control parameters, and perform real-time voltage regulation based on the PWM control parameters.

7. The solar-powered emergency charging system for electric vehicles as described in claim 6, characterized in that: The voltage regulation response model training method includes: Historical error signals and historical illumination intensity signals are acquired, and abnormal data in these signals are denoised. Based on illumination intensity signals within different sampling periods, mean denoising is performed on the historical illumination intensity signals, as shown in the following formula: in, express The light intensity signal at any given time, express The light intensity signal at any given time, The total sampling time is represented by the formula below. After removing abnormal noise points from historical error signals and historical illumination intensity signals, and supplementing missing values, an illumination prediction function is proposed based on the denoised historical illumination intensity to predict the illumination signal intensity at subsequent times. in, express Predicted light intensity at any given time express Predicted light intensity at any given time express The light intensity signal at any given time, Indicates the historical influence coefficient. This represents the trend enhancement factor. For predicted light intensity values, when there is a difference between the predicted and current light intensity values, a voltage correction is performed in advance. The correction formula is as follows: in, express Voltage correction amount at any given time. This represents the voltage correction coefficient. Based on this voltage correction coefficient as feedforward compensation, the PWM control parameters are adjusted using the following formula: in, express PWM control parameters at any given time. express PWM control parameters at any given time. This represents the control coefficient. These represent the PID parameters, express Voltage error at time, express The voltage error at any given time is used to regulate the voltage response model based on the PWM control parameter adjustment method. The voltage regulation circuit is controlled by predicting the light intensity. The historical error signal and the light intensity signal are integrated into a modeling sample set, which is then divided into a training set and a test set. The joint loss function of the voltage response model is preset with the goal of minimizing the voltage regulation error. Load the pre-built voltage regulation response model, initialize the PID parameters and illumination compensation coefficient of the voltage regulation response model, and preset the training rounds, parameter optimization methods, and early stopping mechanism of the voltage regulation response model; Obtain the training set, iteratively train the voltage regulation response model based on the training set until convergence, and output the converged voltage regulation response model; The voltage regulation response model is tested using a test set, the test results are output, and the test error between the test results and the actual results is calculated using a joint loss function. Determine whether the test error exceeds the preset test accuracy. If it does, output the converged voltage regulation response model.

8. The solar-powered emergency charging system for electric vehicles as described in claim 7, characterized in that: The joint loss function of the voltage regulation response model is expressed as: in, This represents the joint loss function of the voltage regulation response model, while These are the error signal weighting coefficient, the illumination intensity weighting coefficient, and the PWM parameter change rate weighting coefficient, respectively. These represent the error signal, illumination intensity, and PWM parameter change rate, respectively. Indicates the sample size.

9. A method for emergency charging of electric vehicles based on solar power generation, implemented using an emergency charging system for electric vehicles based on solar power generation as described in any one of claims 1-8, characterized in that: The solar-powered emergency charging method for electric vehicles includes: The solar panel array converts light energy into DC power, and collects the light intensity signal, single panel output voltage and single panel output power on the solar panel side in real time. The solar panel current balance is achieved based on the dynamic current sharing strategy, and the DC power output by the solar panel is connected to the boost circuit. The power battery is evaluated and analyzed based on battery state parameters, and the preset voltage value of the power battery with different access types is output, and the DC power is boosted to the preset voltage value. The system collects the output voltage and light intensity signals of the charging gun module in real time, and generates an error signal based on the voltage error between the output voltage and the preset voltage value. Using the light intensity signal and the error signal as input, it executes a pre-built voltage regulation response model, outputs dynamic PWM control parameters, and performs real-time voltage regulation based on the PWM control parameters. The charging gun module is electrically connected to the electric vehicle's power battery, providing power to the battery and collecting the battery status parameters in real time, feeding these parameters back to the boost circuit.

10. The emergency charging method for electric vehicles based on solar power generation as described in claim 9, characterized in that: The method for achieving solar panel current balancing based on a dynamic current sharing strategy includes: Real-time acquisition of solar panel irradiance signal, single panel output voltage, and single panel output power; The real-time status of the solar panel is determined based on the collected light intensity signal, single-board output voltage and single-board output power, and the fault board isolation command is triggered based on the real-time status determination result of the solar panel. The system obtains the real-time status judgment results of the solar panel and adjusts the PWM duty cycle of the solar panel based on the preset dynamic current sharing strategy to achieve current balance.

Images