Intelligent program-controlled analog load system, non-vehicle-mounted charger dynamic efficiency test method, equipment and medium

Abstract

The present application belongs to the technical field of non-vehicle charger efficiency testing, and discloses an intelligent program-controlled analog load system, a non-vehicle charger dynamic efficiency testing method, equipment and a medium to solve the problem of testability. The system comprises, connected in turn: a charging behavior portrait modeling module, which clusters high-dimensional charging behavior data based on a SOM network to generate multiple charging behavior portrait categories; a dynamic load simulation module, which generates a corresponding dynamic load curve for each charging behavior portrait category based on GAN and LSTM; a multi-modal load fusion module, which performs collaborative weighted fusion on the dynamic load curves corresponding to each charging behavior portrait category based on the Shapley value in the cooperative game theory to generate a comprehensive multi-modal load curve; and an intelligent program-controlled load generation module, which converts the multi-modal load curve into a power electronic load signal based on power electronic technology and outputs the signal to a power device to generate a physical load.

Description

Technical Field

[0001] This invention belongs to the field of off-board charger efficiency testing technology, specifically relating to an intelligent programmable simulated load system, a method, equipment, and medium for testing the dynamic efficiency of off-board chargers. Background Technology

[0002] As a core facility for electric vehicle energy replenishment, the performance of off-board chargers directly affects the efficiency of the charging process and the power quality of the grid. A complete charger energy efficiency testing system typically evaluates key indicators such as standby power consumption, operating efficiency, and power factor by simultaneously measuring the AC and DC electrical parameters on its input and output sides.

[0003] Currently, relevant standards (such as GB / T 18487.5-2024 and the IEC 61851 series) provide basic specifications for efficiency testing, typically requiring testing in a stable environment with constant temperature and humidity (e.g., 20℃-30℃, relative humidity 45%-75%). However, these traditional testing methods have significant technical shortcomings. In actual operation, the efficiency of off-board chargers is comprehensively affected by user charging behavior (e.g., initial and final charging amounts, charging power fluctuations), ambient temperature and humidity, and complex transient operating conditions. Existing testing methods often employ static or simple periodic loads, making it difficult to simulate the complex and ever-changing dynamic charging process in the real world. This results in test results that cannot comprehensively and accurately reflect the dynamic efficiency characteristics of off-board chargers in practical applications.

[0004] In summary, the main shortcomings of existing off-board charger energy efficiency testing technologies are: the test conditions are disconnected from the real operating environment, and load simulation methods cannot reproduce dynamic, multimodal user charging behavior, making it difficult to accurately assess and comprehensively characterize the dynamic efficiency of off-board chargers. This limitation restricts the accurate selection and optimization of high-efficiency charging equipment and hinders the improvement of the overall energy efficiency management level of charging networks. Summary of the Invention

[0005] Based on the aforementioned shortcomings and deficiencies in the prior art, one of the objectives of this invention is to at least solve one or more of the aforementioned problems in the prior art. In other words, one of the objectives of this invention is to provide an intelligent programmable simulated load system, a method, equipment, and medium for testing the dynamic efficiency of off-board chargers that meet one or more of the aforementioned requirements, so as to achieve the goal of accurately evaluating and comprehensively characterizing the dynamic efficiency of off-board chargers.

[0006] To achieve the above-mentioned objectives, the present invention adopts the following technical solution:

[0007] In a first aspect, the present invention provides an intelligent programmable simulated load system, comprising, in sequence: a charging behavior profile modeling module, used to cluster high-dimensional charging behavior data based on a self-organizing map neural network to generate multiple charging behavior profile categories; a dynamic load simulation module, used to generate a corresponding dynamic load curve for each of the charging behavior profile categories based on a generative adversarial network and a long short-term memory network; a multimodal load fusion module, used to perform collaborative weighted fusion of the dynamic load curves corresponding to various charging behavior profile categories based on the Shapley value in cooperative game theory to generate a comprehensive multimodal load curve; and an intelligent programmable load generation module, used to convert the multimodal load curve into a power electronic load signal based on power electronics technology and output it to a power device to generate a physical load, thereby performing dynamic efficiency testing of the off-board charger.

[0008] As a preferred embodiment, the charging behavior profiling modeling module performs the following profiling modeling operations: constructing a six-dimensional behavior feature vector, including starting SOC, ending SOC, charging duration, average power, power variance, and peak current ratio; normalizing the six-dimensional behavior feature vector; initializing the parameters of the self-organizing map neural network, including weights, neighborhood, and number of iterations; and inputting the processed charging behavior data into the self-organizing map neural network to form multiple charging behavior profiling categories through competitive learning and weight updates.

[0009] As a preferred embodiment, the self-organizing mapping neural network consists of an input layer and a competition layer, wherein the competition layer is a two-dimensional planar array composed of neurons; the self-organizing mapping neural network maps high-dimensional charging behavior data to a two-dimensional output plane, forming multiple charging behavior profile categories.

[0010] As a preferred embodiment, the dynamic load simulation module performs the following dynamic load curve generation operation: constructing a generator and a judge, wherein the generator adopts an LSTM network architecture and the judge adopts a multi-scale convolutional layer; inputting a random sequence into the generator to generate a simulated current and voltage sequence; introducing an environmental coupling factor to correct the simulated voltage and current sequence; judging the consistency between the corrected simulated sequence and the real load sequence through the judge, and iteratively training until the judge cannot distinguish them.

[0011] As a preferred embodiment, the multimodal load fusion module performs the following multimodal load curve generation operation: designing a value function, including efficiency indicators, load fluctuation coverage indicators, and energy consumption indicators; based on the value function, calculating the Shapley value of the dynamic load curve corresponding to each type of charging behavior profile, where the Shapley value represents the marginal contribution of that type of load curve in the fusion process; using a dynamic time warping algorithm to perform time-domain alignment on various types of dynamic load curves to eliminate phase offset; and using the calculated Shapley value as a weight, performing a weighted summation on the time-domain aligned load curves to generate the multimodal load curve.

[0012] As a preferred embodiment, the intelligent programmable load generation module performs the following physical load generation operations: using the multimodal load curve as system input; employing a slope detection circuit to monitor the output voltage change rate and dynamically adjusting the dead time and switching frequency based on the change rate; utilizing digital twin technology to calibrate the output of the slope detection circuit in real time; decomposing the multimodal load curve into timing instructions and tracking the load curve in real time through high-frequency PWM modulation; and using fully controllable power devices to generate a physical load according to the timing instructions for dynamic efficiency testing of off-board chargers.

[0013] As a preferred embodiment, the fully controllable power device is a SiC MOSFET.

[0014] Secondly, the present invention provides a method for testing the dynamic efficiency of an off-board charger, characterized in that it employs an intelligent programmable simulated load system as described in the first aspect, the method comprising the steps of: connecting a programmable three-phase AC power supply to the input terminal of the off-board charger; connecting the output terminal of the intelligent programmable simulated load system to the output terminal of the off-board charger as its load; synchronously monitoring the voltage and current data at the input and output terminals of the off-board charger; and calculating the dynamic efficiency of the off-board charger based on the monitored input and output power.

[0015] Thirdly, the present invention provides an electronic device, the computer device including a memory, a processor and a computer program, wherein the computer program, when executed by the processor, implements the dynamic efficiency testing method as described in the second aspect.

[0016] Fourthly, the present invention provides a computer-readable storage medium having a computer program stored thereon, wherein the computer program, when executed by a processor, implements the dynamic efficiency testing method as described in the second aspect.

[0017] Compared with the prior art, the present invention has the following beneficial effects:

[0018] 1. Existing technologies for simulating charging loads struggle to comprehensively cover different charging behavior modes, resulting in discrepancies between simulated and actual load curves. This invention utilizes self-organizing map neural network clustering analysis to generate multimodal load curves, comprehensively considering various charging behavior characteristics to accurately simulate actual load changes, providing reliable load conditions for dynamic efficiency testing of off-board chargers.

[0019] 2. This invention introduces a variety of advanced technologies, including self-organizing map neural network clustering to discover patterns, generative adversarial network to expand the dataset, and long short-term memory network to capture temporal patterns, comprehensively enhancing data processing and analysis capabilities and helping to gain a deeper understanding of charging behavior.

[0020] 3. Traditional multimodal load fusion neglects interaction effects, lacks fairness, and makes it difficult to reasonably determine weights. This invention introduces the Shapley value to calculate the marginal contribution of each type of load curve to determine the weight, fully considering interaction effects, ensuring fair and reasonable fusion, and making the multimodal load curve more accurately reflect the actual comprehensive load.

[0021] 4. Existing technologies for dynamic efficiency testing of off-board chargers suffer from inaccurate load simulation or a single load type, resulting in significant discrepancies between the results and actual values, thus affecting reliability. This invention accurately simulates multimodal loads and converts them into physical loads, providing typical reliable load conditions, accurately calculating dynamic efficiency, improving test reliability, and providing an accurate basis for performance evaluation and optimization.

[0022] 5. The technical methods employed in this invention are highly applicable and scalable. Self-organizing map neural networks and other technologies can be widely applied to time series data processing and pattern recognition, while the Shapley value fusion method can be adjusted and optimized as needed. The intelligent programmable simulated load system can be easily integrated into existing charging test equipment, reducing application costs and complexity.

[0023] Further or more detailed beneficial effects will be described in conjunction with specific embodiments in the detailed implementation. Attached Figure Description

[0024] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0025] Figure 1 This is a schematic diagram of the application scenario of the intelligent programmable simulated load system described in Embodiment 1 of the present invention.

[0026] Figure 2 This is a schematic diagram of the topology of the self-organizing mapping neural network described in Embodiment 1 of the present invention.

[0027] Figure 3 This is a schematic diagram of the process for generating the dynamic load curve as described in Embodiment 1 of the present invention.

[0028] Figure 4 This is a schematic diagram of the process for generating the multimodal load curve as described in Embodiment 1 of the present invention.

[0029] Figure 5 This is a schematic diagram of the physical load generation process described in Embodiment 1 of the present invention.

[0030] Figure 6 This is a schematic diagram illustrating the principle of the dynamic efficiency testing method described in Embodiment 2 of the present invention.

[0031] Figure 7 This is a structural diagram of the electronic device provided in the embodiment of the present invention.

[0032] Icon labels:

[0033] 700. Electronic equipment;

[0034] 701. Processor; 702. Communication bus; 703. User interface; 704. Network interface; 705. Memory. Detailed Implementation

[0035] The technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings in the embodiments of the present invention.

[0036] In the following description, several embodiments of the present invention are provided. Different embodiments can be substituted or combined. Therefore, the present invention can also be considered to include all possible combinations of the same and / or different embodiments described. Thus, if one embodiment includes features A, B, and C, and another embodiment includes features B and D, then the present invention should also be considered to include embodiments containing one or more other possible combinations of A, B, C, and D, even if such embodiments are not explicitly described in the following text.

[0037] The following description provides examples and does not limit the scope, applicability, or examples set forth in the claims. Changes may be made to the function and arrangement of the described elements without departing from the scope of the invention. Various processes or components may be appropriately omitted, substituted, or added to the various examples. For example, the described methods may be performed in a different order than described, and various steps may be added, omitted, or combined. Furthermore, features described with respect to some examples may be combined into other examples.

[0038] To facilitate a better understanding of the embodiments of the present invention, its application scenarios will be explained before providing a detailed explanation of the specific implementation methods.

[0039] Please see Figure 1 , Figure 1 A schematic diagram of the application scenario of the intelligent programmable simulated load system is shown.

[0040] The intelligent programmable simulated load system described in the embodiments of this specification is applied to the dynamic efficiency testing process of off-board chargers. In these scenarios, the application of the intelligent programmable simulated load system aims to transform the simulated multimodal load curves into actual electrical loads. By accurately simulating the dynamic changes of electric vehicle loads under different operating conditions, it provides typical and reliable load conditions for the dynamic efficiency testing of off-board chargers, thereby accurately calculating the dynamic efficiency of off-board chargers in real-world usage scenarios.

[0041] The following is a brief explanation of the self-organizing map neural network, high-dimensional charging behavior data, charging behavior profile categories, generative adversarial network, long short-term memory network, dynamic load curve, Shapley value, multimodal load curve, power electronics technology, power electronic load signal, and physical load involved in several embodiments of this specification:

[0042] Self-Organizing Map (SOM) neural networks are a type of unsupervised learning neural network model. They can map high-dimensional data to a low-dimensional space (typically two-dimensional) while preserving the topological relationships between the data. When processing charging behavior data, it can automatically cluster high-dimensional charging behavior data, grouping charging behavior data with similar characteristics into the same category, thereby helping to better understand and analyze the patterns and regularities of charging behavior.

[0043] High-dimensional charging behavior data refers to the dataset collected from multiple dimensions during the electric vehicle charging process. These dimensions may include charging time, charging power, charging voltage, charging current, battery temperature, and battery level at the start and end of charging. High-dimensional data contains rich information, but it also presents challenges for data processing and analysis, requiring appropriate methods for dimensionality reduction and feature extraction.

[0044] Charging behavior profiles are categorized based on the analysis and processing of high-dimensional charging behavior data. Using clustering methods such as self-organizing map neural networks, data with similar charging behavior characteristics are grouped into one category, each representing a different charging behavior profile category. For example, the fast charging behavior category might be characterized by high charging power and short charging time, while the slow charging behavior category might be characterized by low charging power and long charging time.

[0045] Generative Adversarial Networks (GANs) are deep learning models consisting of a generator and a discriminator. The generator aims to produce fake data that resembles real data, while the discriminator aims to distinguish real data from the fake data generated by the generator. In charging behavior-related applications, GANs can be used to generate simulated charging behavior data to expand datasets or simulate different charging scenarios.

[0046] Long Short-Term Memory (LSTM) networks are a special type of recurrent neural network (RNN). They are capable of processing and predicting long-term dependencies in time-series data. In charging behavior analysis, the data during the charging process is sequential data that changes over time. LSTM can effectively capture the temporal characteristics and patterns in this data, such as predicting changes in charging power over a future period.

[0047] A dynamic load curve describes how the load changes over time during charging. It reflects the power demand of the charging equipment at different times. In off-board charger dynamic efficiency testing, different types of charging behaviors will produce different dynamic load curves, and these curves are crucial for evaluating the charger's performance and efficiency.

[0048] Shapley value is an important concept in cooperative game theory, used to fairly distribute the benefits of cooperation. In this invention, Shapley value is introduced into multimodal load fusion. By calculating the marginal contribution of each type of load curve, the weight of each type of load in the fusion process is determined, solving the problems of traditional fusion methods ignoring interaction effects and lacking fairness.

[0049] The multimodal load curve is a comprehensive load curve obtained by collaboratively weighting and fusing the dynamic load curves corresponding to different charging behavior profiles. It comprehensively considers the characteristics of various charging behaviors, and can more comprehensively and accurately reflect the load situation in actual charging scenarios, providing a more reliable basis for the dynamic efficiency testing of off-board chargers.

[0050] Power electronics technology is a technology that uses power electronic devices to convert and control electrical energy. In this invention, power electronics technology is used to convert simulated multimodal load curves into actual power electronic load signals, and to achieve accurate simulation of the load by controlling the switching states of power devices.

[0051] Power electronic load signals are electrical signals generated using power electronics technology to simulate actual loads. They contain information such as load voltage and current, accurately reflecting the load's characteristics. In off-board charger dynamic efficiency testing, the power electronic load signal is output to power devices to generate a physical load.

[0052] A physical load refers to an actual device or apparatus that consumes electrical energy. In this invention, power electronic load signals are converted into physical loads using power electronics technology, thus constructing an actual power electronic load device, such as a load constructed using fully controlled power devices (e.g., SiC MOSFETs), for dynamic efficiency testing of off-board chargers.

[0053] Example 1:

[0054] This embodiment provides an intelligent programmable simulated load system, which includes the following modules connected in sequence:

[0055] The charging behavior profiling module is used to cluster high-dimensional charging behavior data based on a self-organizing map (SOM) neural network, generating multiple charging behavior profile categories. The efficiency of off-board chargers is significantly affected by differences in user charging behavior, such as high-frequency fast charging for ride-hailing vehicles and periodic charging for private cars. Traditional static load testing cannot reflect the dynamic characteristics of real-world scenarios, and charging behavior features that may affect the efficiency of off-board chargers include initial State of Charge (SOC), final SOC, charging duration, average power, power variance, and peak current ratio. To simulate the dynamic load of off-board chargers, profiling models of different charging behaviors are necessary. This embodiment uses a self-organizing map (SOM) neural network to implement this clustering and profiling function.

[0056] SOM, an unsupervised learning algorithm for clustering and high-dimensional visualization, was developed by simulating the signal processing characteristics of the human brain. Figure 2 As shown, it consists of an input layer and a competition layer (or output layer), where the competition layer is a two-dimensional planar array composed of neurons. In this embodiment, SOM maps high-dimensional behavioral data (such as charging periods, SOC interval probabilities, and power fluctuations) to the two-dimensional planar array of the output layer through competitive learning and topology preservation, forming interpretable user profile categories.

[0057] Specifically, the charging behavior profiling module performs the following profiling operations:

[0058] 1. Define features: Construct a 6-dimensional behavior vector, including initial SOC, ending SOC, charging time, average power, power variance, and peak current ratio;

[0059] 2. Normalization: Modular normalization is used to eliminate dimensional differences and improve the robustness of distance calculation;

[0060] 3. Parameter initialization: Set the weights w ij Assign small, random initial values; set a larger initial neighborhood. N c And set the number of loops in the network. T ;

[0061] 4. Charging behavior input: Input each charging behavior X k ={ x 1k , x 2k , … , x 7k} Input them sequentially into the network;

[0062] 5. Charging behavior clustering: Calculate charging behavior X k Distance to all output neurons d jk , and select X k The neuron with the smallest distance c , c For the winning neuron, i.e. X k The corresponding portrait;

[0063] 6. Weight Update: Update node weights c The connection weights of the node and its neighboring nodes are updated using the formula (1):

[0064] (1)

[0065] In equation (1), η ( t The ) represents the learning rate, which decreases as the number of iterations increases;

[0066] 7. Looping and Iterating: Let t = t +1, return to step 4, until... t = T until.

[0067] The SOM network constructed in this embodiment has topology preservation and dynamic noise resistance capabilities. It can map high-dimensional electricity consumption behavior data to a two-dimensional semantic grid. While preserving the original data spatial structure, it automatically identifies the boundaries of charging behavior patterns, overcomes the shortcomings of traditional K-means clustering that is sensitive to initial centers and cannot handle nonlinear distributions, and improves the interpretability of charging behavior profiles.

[0068] The dynamic load simulation module generates corresponding dynamic load curves for each charging behavior profile category based on generative adversarial networks (GANs) and long short-term memory (LSTM) networks. The charging process from an off-board charger to an electric vehicle involves complex current and voltage timing characteristics, which traditional resistive loads cannot represent in a real charging process. This embodiment employs a combination of GANs and LSTMs, and introduces an environmental coupling factor for correction, to generate a dynamic voltage and current sequence that conforms to the real charging process.

[0069] The core idea of ​​GAN comes from Nash equilibrium in game theory. Its main structure includes a generator and a judge. In this embodiment, the generator generates simulated voltage and current sequences based on random sequences. The input of the judge is the simulated sequence and the real voltage and current sequence. Through continuous training, when the judge cannot distinguish between the simulated sequence and the real sequence, it is considered that the model can reflect the dynamic characteristics of the real load.

[0070] See Figure 3 The specific operations performed by the dynamic load simulation module to generate dynamic load are as follows:

[0071] 1. Parameter determination: Determine the structure and parameters of the generator and the judge. The generator is mainly based on an LSTM network, and the judge uses multi-scale convolutional layers to determine the consistency of time series sequences.

[0072] 2. Generate a simulated sequence: Input a noise sequence z Generate simulated current and voltage sequences G ( z );

[0073] 3. Environmental Coupling Factor Correction: Introducing environmental coupling factors (including temperature, humidity, etc.) to the analog current-voltage sequence. G ( z The corrected analog current-voltage sequence is obtained by performing corrections. G c ( z );

[0074] 4. Consistency Comparison: The judge compares the corrected analog current and voltage sequences. G c ( z ) and the actual current and voltage sequence x Check if they match. If they do not match, return to step 2. If they match, end the loop.

[0075] The GAN-LSTM method proposed in this embodiment has high transient response accuracy and strong generalization ability, which makes up for the shortcomings of traditional analog circuits such as slow frequency response and narrow operating condition coverage, and realizes high-precision simulation of the dynamic load of off-board charger charging.

[0076] The multimodal load fusion module, based on the Shapley value in cooperative game theory, performs collaborative weighted fusion of dynamic load curves corresponding to various charging behavior profiles to generate a comprehensive multimodal load curve. In the dynamic efficiency test of off-board chargers, different charging behaviors correspond to differentiated load curves. To obtain a comprehensive multimodal load for various charging behaviors, traditional methods typically use arithmetic averaging or subjective weighting to fuse the simulated curves corresponding to various charging behaviors. However, these traditional load fusion methods have significant drawbacks: on the one hand, they do not consider the nonlinear coupling relationship between different load curves; on the other hand, fixed weights cannot dynamically reflect the actual contribution of various loads under different operating conditions, which may lead to efficiency test results deviating from the real scenario. Therefore, this embodiment introduces the Shapley value as a weighting basis, effectively solving the problems of traditional fusion methods ignoring interaction effects and lacking fairness.

[0077] Shapley values, as an allocation tool in cooperative game theory, solve the problem of fair allocation in multimodal load fusion by calculating the marginal contribution of each type of load curve. Mathematically, it is an allocation scheme that satisfies efficiency (total contribution equals fusion effect), symmetry (equivalent curves have the same weight), and virtuality (ineffective curves have zero weight).

[0078] See Figure 4 The specific operations performed by the multimodal load fusion module to generate multimodal load curves are as follows:

[0079] 1. Value Function Design: Let... N For all charging behavior categories, S For any N A subset of the value function (i.e., the fusion benefit). v ( S This can be represented as:

[0080] (2)

[0081] In equation (2), α , β and γ These represent the efficiency weight, load fluctuation coverage weight, and energy consumption penalty coefficient, respectively, and their values ​​can be determined by the multimodal load fusion requirements. Efficiency ( S () is an efficiency indicator, representing the average efficiency of the testing process. Coverage ( SThe term "coverage index" refers to the load fluctuation entropy. A higher entropy value indicates more comprehensive operating condition coverage. EnergyCost ( S () is an energy consumption indicator, and its physical meaning is the power consumption measured.

[0082] 2. Shapley Value Calculation: The Shapley value for each type of charging behavior is calculated based on the value function, using the following formula:

[0083] (3)

[0084] In equation (4), The weight coefficients represent subsets. S The probability of occurrence Marginal contribution indicates joining the first... i The increase in fusion benefits corresponding to the charging-like behavior and the load. φ i Indicates the first i The Shapley value of the load corresponding to the charging behavior;

[0085] 3. Dynamic Time-Domain Alignment: Dynamic Time Warping (DTW) is used to eliminate phase shifts between curves. The calculation formula is as follows:

[0086] (4)

[0087] In equation (4), L i ( t ) indicates the first i Load curves corresponding to charging behavior;

[0088] 4. Weighted fusion: The fused load curve after time-shift compensation is obtained by weighted calculation. The calculation formula is as follows:

[0089] (5).

[0090] In this embodiment, the Shapley value collaborative weighting effectively solves the problems of fairness and interaction effects in multimodal load fusion through marginal contribution quantification and dynamic weight allocation. The generated multimodal load can reflect the load characteristics after weighting various charging behaviors.

[0091] An intelligent programmable load generation module is used to convert the multimodal load curve into a power electronic load signal based on power electronics technology, and output it to power devices to generate a physical load for dynamic efficiency testing of off-board chargers. To use the simulated multimodal load curve for actual off-board charger dynamic efficiency testing, the load curve needs to be converted into an actual electrical load. This embodiment combines power electronics and digital twin technology to propose a method for generating an intelligent programmable simulated load, see reference [link to relevant documentation]. Figure 5 The specific implementation steps are as follows:

[0092] 1. Load curve input: Use the simulated multimodal load curve as the system input;

[0093] 2. Slope detection: A slope detection circuit is used to monitor the rate of change of the output voltage (dV / dt). When a steep rise / fall is detected, the dead time and switching frequency are dynamically adjusted to suppress overshoot voltage.

[0094] 3. Real-time calibration: The output of the slope detection circuit is calibrated in real time using digital twin technology;

[0095] 4. Curve Analysis: The load curve is decomposed into timing instructions, and the load curve is tracked in real time through high-frequency PWM modulation;

[0096] 5. Load output: Power electronic loads are constructed using fully controlled power devices (such as SiC MOSFETs) as the system output.

[0097] The programmable simulated load obtained through the above steps in this embodiment can accurately characterize the dynamic changes of electric vehicle load under different operating conditions, and is typical, so it can be directly used for dynamic efficiency testing of off-board chargers. During the test, a programmable three-phase AC power supply is connected to the input terminal of the off-board charger, and the intelligent programmable simulated load is used as the output terminal of the off-board charger. The voltage and current data of the input and output terminals are monitored respectively, the input and output power are calculated, and then the dynamic efficiency of the off-board charger is obtained.

[0098] Example 2:

[0099] This embodiment provides a dynamic efficiency testing method for off-board chargers, employing the intelligent programmable simulated load system described in Embodiment 1. The principle of this method is explained in [reference needed]. Figure 6 The method includes the following steps:

[0100] Connect the programmable three-phase AC power supply to the input terminal of the off-board charger;

[0101] Connect the output of the intelligent programmable simulated load system to the output of the off-board charger as its load.

[0102] Simultaneously monitor the voltage and current data at the input and output terminals of the off-board charger;

[0103] The dynamic efficiency of the off-board charger is calculated based on the monitored input and output power.

[0104] Example 3:

[0105] like Figure 7As shown, this embodiment provides an electronic device, which may include: at least one processor, at least one network interface, a user interface, a memory, and at least one communication bus.

[0106] The communication bus can be used to enable communication between the various components mentioned above.

[0107] The user interface may include buttons, and optional user interfaces may also include standard wired interfaces and wireless interfaces.

[0108] The network interface may include, but is not limited to, Bluetooth modules, NFC modules, Wi-Fi modules, etc.

[0109] The processor may include one or more processing cores. It connects various parts of the electronic device via various interfaces and lines, executing instructions, programs, code sets, or instruction sets stored in memory, and accessing data stored in memory to perform various functions and process data. Optionally, the processor can be implemented using at least one hardware form of DSP, FPGA, or PLA. The processor may integrate one or more of the following: CPU, GPU, and modem. The CPU primarily handles the operating system, user interface, and applications; the GPU is responsible for rendering and drawing the content required for display; and the modem handles wireless communication. It is understood that the modem may also be implemented as a separate chip without being integrated into the processor.

[0110] The memory may include RAM or ROM. Optionally, the memory may include a non-transitory computer-readable medium. The memory can be used to store instructions, programs, code, code sets, or instruction sets. The memory may include a program storage area and a data storage area, wherein the program storage area may store instructions for implementing an operating system, instructions for at least one function (such as touch function, sound playback function, image playback function, etc.), instructions for implementing the above-described method embodiments, etc.; the data storage area may store data involved in the above-described method embodiments, etc. Optionally, the memory may also be at least one storage device located remotely from the aforementioned processor. The memory, as a computer storage medium, may include an operating system, a network communication module, a user interface module, and a test application. The processor can be used to call the test application stored in the memory and execute the steps of the dynamic efficiency testing method mentioned in the foregoing embodiments.

[0111] Example 4:

[0112] This embodiment provides a computer-readable storage medium storing instructions that, when executed on a computer or processor, cause the computer or processor to perform one or more steps as described in Embodiment 2. If the constituent modules of the above-described electronic device are implemented as software functional units and sold or used as independent products, they can be stored in the computer-readable storage medium.

[0113] In the above embodiments, implementation can be achieved, in whole or in part, through software, hardware, firmware, or any combination thereof. When implemented in software, it can be implemented, in whole or in part, as a computer program product. The computer program product includes one or more computer instructions. When the computer program instructions are loaded and executed on a computer, all or part of the processes or functions described in the embodiments of this specification are generated. The computer can be a general-purpose computer, a special-purpose computer, a computer network, or other programmable device. The computer instructions can be stored in or transmitted through a computer-readable storage medium. The computer instructions can be transmitted from one website, computer, server, or data center to another website, computer, server, or data center via wired (e.g., coaxial cable, fiber optic, Digital Subscriber Line (DSL)) or wireless (e.g., infrared, wireless, microwave, etc.) means. The computer-readable storage medium can be any available medium accessible to a computer or a data storage device such as a server or data center that integrates one or more available media. The available media may be magnetic media (e.g., floppy disks, hard disks, magnetic tapes), optical media (e.g., Digital Versatile Discs (DVDs)), or semiconductor media (e.g., Solid State Disks (SSDs)).

[0114] Those skilled in the art will understand that all or part of the processes in the methods of Embodiment 2 described above can be implemented by a computer program instructing related hardware. This program can be stored in a computer-readable storage medium, and when executed, it can include the processes of the embodiments of the methods described above. The aforementioned storage medium includes various media capable of storing program code, such as ROM, RAM, magnetic disks, or optical disks. Unless otherwise specified, the technical features of this embodiment and the implementation scheme can be combined arbitrarily.

[0115] It should be noted that, for the sake of simplicity, the foregoing method 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, because according to the present invention, some steps can be performed in other orders or simultaneously. 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.

[0116] In the above embodiments, the descriptions of each embodiment have different focuses. For parts not described in detail in a certain embodiment, please refer to the relevant descriptions in other embodiments.

[0117] The above description is merely an exemplary embodiment of the present invention and should not be construed as limiting the scope of the invention. Any equivalent changes and modifications made in accordance with the teachings of this invention are still within the scope of this invention. Those skilled in the art will readily conceive of embodiments of the invention upon considering the specification and practicing the disclosure herein. This invention is intended to cover any variations, uses, or adaptations of the invention that follow the general principles of the invention and include common knowledge or customary techniques in the art not described herein. The specification and embodiments are to be considered exemplary only, and the scope and spirit of the invention are defined by the claims.

Claims

1. An intelligent programmable simulated load system, characterized in that, Including those connected sequentially: The charging behavior profile modeling module is used to cluster high-dimensional charging behavior data based on a self-organizing map neural network to generate multiple charging behavior profile categories. The dynamic load simulation module is used to generate a corresponding dynamic load curve for each charging behavior profile category based on generative adversarial networks and long short-term memory networks. The multimodal load fusion module is used to perform collaborative weighted fusion of dynamic load curves corresponding to various charging behavior profile categories based on the Shapley value in cooperative game theory, and generate a comprehensive multimodal load curve. The multimodal load fusion module performs the following multimodal load curve generation operation: Design a value function, including efficiency metrics, load fluctuation coverage metrics, and energy consumption metrics; Based on the value function, the Shapley value of the dynamic load curve corresponding to each type of charging behavior profile is calculated. The Shapley value represents the marginal contribution of the load curve in the fusion process. Dynamic time warping algorithm is used to align various dynamic load curves in the time domain to eliminate phase offset; Using the calculated Shapley value as a weight, the time-domain aligned load curves are weighted and summed to generate the multimodal load curves. The intelligent programmable load generation module is used to convert the multimodal load curve into a power electronic load signal based on power electronics technology, and output it to the power device to generate a physical load, thereby conducting dynamic efficiency testing of the off-board charger.

2. The intelligent programmable simulated load system according to claim 1, characterized in that, The charging behavior profiling module performs the following profiling operations: Construct a six-dimensional behavioral feature vector, including initial SOC, ending SOC, charging time, average power, power variance, and peak current ratio; The six-dimensional behavioral feature vector is normalized. Initialize the parameters of the self-organizing map neural network, including weights, neighborhood, and number of iterations; The processed charging behavior data is input into a self-organizing map neural network, and multiple charging behavior profile categories are formed through competitive learning and weight updates.

3. The intelligent programmable simulated load system according to claim 2, characterized in that: The self-organizing map neural network consists of an input layer and a competition layer, wherein the competition layer is a two-dimensional planar array composed of neurons. The self-organizing mapping neural network maps high-dimensional charging behavior data to a two-dimensional output plane, forming multiple charging behavior profile categories.

4. The intelligent programmable simulated load system according to claim 3, characterized in that, The dynamic load simulation module performs the following dynamic load curve generation operation: Construct a generator and a judge, wherein the generator adopts an LSTM network architecture and the judge adopts a multi-scale convolutional layer; Input a random sequence into the generator to generate a simulated current-voltage sequence; An environmental coupling factor is introduced to correct the simulated voltage and current sequence; The discriminator determines the consistency between the corrected simulated sequence and the real load sequence, and iterates the training until the discriminator can no longer distinguish between them.

5. The intelligent programmable simulated load system according to claim 4, characterized in that, The intelligent programmable load generation module performs the following physical load generation operations: Use the multimodal load curve as the system input; A slope detection circuit is used to monitor the rate of change of the output voltage, and the dead time and switching frequency are dynamically adjusted according to the rate of change. The output of the slope detection circuit is calibrated in real time using digital twin technology; The multimodal load curve is decomposed into timing instructions, and the load curve is tracked in real time by high-frequency PWM modulation; A fully controlled power device is used to generate a physical load according to the timing instructions, which is used for dynamic efficiency testing of off-board chargers.

6. The intelligent programmable simulated load system according to claim 5, characterized in that: The fully controllable power device is a SiC MOSFET.

7. A method for testing the dynamic efficiency of a non-vehicle-mounted charger, characterized in that, The method using the intelligent programmable simulated load system as described in claim 6 includes the following steps: Connect the programmable three-phase AC power supply to the input terminal of the off-board charger; Connect the output of the intelligent programmable simulated load system to the output of the off-board charger as its load. Simultaneously monitor the voltage and current data at the input and output terminals of the off-board charger; The dynamic efficiency of the off-board charger is calculated based on the monitored input and output power.

8. A computer device, the computer device comprising a memory, a processor, and a computer program, characterized in that, When the computer program is executed by the processor, it implements the dynamic efficiency testing method as described in claim 7.

9. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by the processor, it implements the dynamic efficiency testing method as described in claim 7.

Images