Vehicle-mounted charger performance test method and system

Abstract

The invention discloses a vehicle-mounted charger performance testing method and system, and belongs to the field of electronic product manufacturing and automatic testing, and the method comprises the steps: collecting initial parameters of a charger interface to generate an alignment driving instruction; performing docking based on the driving instruction and performing dual verification to obtain a connection result; starting multi-channel detection to generate performance data after successful connection; and finally, based on the performance data, performing intelligent judgment and generating a differential processing instruction. According to the method, virtual simulation and a machine learning model are combined, and a causal analysis and feedback optimization closed loop from a physical process risk to a final quality result is constructed, so that full-process automation and intelligent control of a test process can be realized, and the first pass yield, reliability and intelligent level of a vehicle-mounted charger production test are remarkably improved.

Description

Technical Field

[0001] The present invention relates to the field of electronic product manufacturing and automated testing, and in particular to a vehicle charger performance testing method and system. Background Art

[0002] As a core component of new energy vehicles, the performance and reliability of the on-board charger are directly related to vehicle safety. During manufacturing, performance testing is essential to ensuring product quality and conformity to factory standards. Traditional performance testing is typically conducted at a dedicated test station after product assembly. Key electrical parameters of the charger are measured to verify that its basic functionality meets design requirements.

[0003] As production increases, automated equipment is being used in assembly and testing to improve efficiency. However, existing automation solutions often treat assembly and testing as separate units, resulting in production cycle mismatches and additional logistics. Furthermore, product alignment and electrical connections during testing often rely on universal fixtures, which can lead to insufficient alignment accuracy and unreliable contact, impacting the accuracy and consistency of test results. Existing technologies also lack a mechanism for seamlessly integrating test results with subsequent processes such as labeling and sorting. Summary of the Invention

[0004] To address the above issues, the present invention provides a method and system for testing the performance of an on-board charger. By combining virtual simulation with a machine learning model, and constructing a closed loop of causal analysis and feedback optimization from physical process risks to final quality results, the method and system can achieve full-process automation and intelligent control of the testing process, significantly improving the pass rate, reliability, and intelligence level of on-board charger production testing.

[0005] The above objectives can be achieved through the following solutions: A method for testing the performance of an on-board charger comprises collecting spatial attitude parameters and electrical characteristic parameters of an on-board charger interface, calculating and generating spatial fine-tuning drive instructions; executing docking based on the spatial fine-tuning drive instructions, and performing a consistency check with a preset docking verification benchmark to obtain a verification result indicating successful or failed docking; if the verification result indicates successful docking, initiating a multi-channel performance testing process to generate full-process performance data; and performing a determination based on the full-process performance data and a preset performance determination threshold to generate and execute differentiated handling instructions.

[0006] Optionally, the calculation and generation of the spatial fine-tuning drive instruction includes: obtaining the real-time three-dimensional coordinates of the vehicle charger interface, and performing vector operation to obtain a three-dimensional spatial deviation vector; synchronously collecting the contact impedance and signal response state of the vehicle charger interface to generate electrical characteristic parameters; based on the three-dimensional spatial deviation vector and the electrical characteristic parameters, calculating the X / Y / Z three-axis fine-tuning amount to generate the spatial fine-tuning drive instruction.

[0007] Optionally, the calculating of the X / Y / Z three-axis fine-tuning amount and the generating of the spatial fine-tuning drive instruction includes: performing kinematic solution on the three-dimensional spatial deviation vector to generate a benchmark drive scheme; performing virtual simulation on the benchmark drive scheme to obtain predictive collision and stress distribution parameters; optimizing and correcting the benchmark drive scheme based on the predictive collision and stress distribution parameters and the electrical characteristic parameters, calculating the X / Y / Z three-axis fine-tuning amount as a correction value, and integrating them to generate the spatial fine-tuning drive instruction.

[0008] Optionally, the consistency check with the preset docking verification benchmark includes: extracting the geometric contour features of the vehicle charger interface, and performing similarity calculation with the asymmetric structure template in the docking verification benchmark to generate structural matching parameters; real-time detection of the contact impedance and signal response delay of the vehicle charger interface, and comparing them with the electrical compliance threshold range in the docking verification benchmark to generate electrical compliance parameters; performing logical AND judgment on the structural matching parameters and the electrical compliance parameters to generate a verification result of docking success or failure.

[0009] Optionally, the generation of structural matching parameters includes: acquiring historical interface image data and historical standard feature point coordinates to obtain a historical interface image data set; using the interface image data as input and the standard feature point coordinates as output, establishing and training a neural network model using the historical interface image data set to obtain an interface geometric feature point recognition model; capturing real-time interface image data, inputting the real-time interface image data into the interface geometric feature point recognition model to obtain key geometric feature point coordinates; performing spatial position comparison and topological relationship analysis on the key geometric feature point coordinates and the asymmetric structure template, and calculating and generating structural matching parameters.

[0010] Optionally, generating full-process performance data includes: applying a simulated load to the power bus of the vehicle charger, and collecting voltage and current responses to obtain basic electrical performance parameters; sending protocol test instructions to the data bus of the vehicle charger, and capturing the returned signal waveform to obtain protocol interaction data; aggregating and structuring the basic electrical performance parameters and the protocol interaction data to generate full-process performance data.

[0011] Optionally, the generating and executing of differentiated handling instructions includes: performing a graded comparison of the full-process performance data with the performance judgment threshold to generate a multi-dimensional performance judgment result; parsing the multi-dimensional performance judgment result to obtain a handling level containing qualified, re-inspection required or unqualified information; based on the handling level, constructing identification instructions for qualified products, re-inspection prompt instructions for re-inspected products, and isolation handling instructions for unqualified products to generate differentiated handling instructions.

[0012] Optionally, the method further includes: performing unsupervised cluster analysis on the data with the disposal level of defective products, extracting common failure characteristics, and generating a failure mode feature vector; tracing the spatial posture parameters and electrical characteristic parameters associated with the failure mode feature vector, and constructing and generating a parameter-failure mode association rule set.

[0013] Optionally, the method further includes: inputting the spatial posture parameters and electrical characteristic parameters collected in real time into the parameter-failure mode association rule set for inference to generate a predictive failure risk level; fusing the predictive failure risk level with the predictive collision and stress distribution parameters to generate feedforward adjustment parameters; and using the feedforward adjustment parameters to dynamically optimize the docking verification benchmark and the performance judgment threshold.

[0014] Based on the same inventive concept, the present invention also provides a vehicle charger performance testing system, which includes: an initial parameter perception and instruction generation module, which is used to collect the spatial posture parameters and electrical characteristic parameters of the vehicle charger interface, calculate and generate spatial fine-tuning drive instructions; a docking and verification module, which is used to perform docking based on the spatial fine-tuning drive instructions, and perform consistency verification with a preset docking verification benchmark to obtain a verification result of successful or failed docking; a multi-channel performance detection module, which is used to start a multi-channel performance detection process and generate full-process performance data if the verification result feedback indicates that the docking is successful; a data judgment and differentiated handling module, which is used to make a judgment based on the full-process performance data and a preset performance judgment threshold, and generate and execute differentiated handling instructions.

[0015] Compared with the prior art, the present invention has the following advantages: 1. This invention seamlessly integrates automatic charger alignment, connection verification, multi-channel performance testing, and result-based identification or isolation into a continuous, automated process. This integrated design fundamentally eliminates the material transfer, waiting, and cycle mismatch issues caused by process separation in traditional solutions, significantly shortening production cycles and improving overall manufacturing efficiency. 2. This invention deeply mines historical data on defective products to construct association rules between initial parameters and failure modes. Based on this, it makes real-time, proactive, and dynamic adjustments to the verification benchmarks and decision thresholds for subsequent tests. This closed-loop self-optimization capability enables the test system to learn from experience, automatically identifying and suppressing risk factors that lead to high-frequency failures, thereby achieving continuous improvement in product quality and a steady increase in production first-pass yield.

[0016] Other features and advantages of the present invention will be described in the following description, and in part will become apparent from the description, or will be understood by practicing the present invention. The purpose and other advantages of the present invention can be realized and obtained by the structures pointed out in the description, claims and drawings. BRIEF DESCRIPTION OF THE DRAWINGS

[0017] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the following briefly introduces the drawings required for use in the embodiments or the description of the prior art. Obviously, the drawings described below are some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying any creative work.

[0018] Figure 1 The figure is a flow chart of a method for testing the performance of an on-board charger according to an embodiment of the present invention.

[0019] Figure 2 It is a two-dimensional contour line of the mapping relationship between the adaptive gain coefficient and the electrical initial parameters in the embodiment of the present invention.

[0020] Figure 3 This is a stress and collision risk coupling analysis diagram in the virtual simulation of an embodiment of the present invention.

[0021] Figure 4 3 is a hierarchical clustering heat map of failure modes and initial parameters according to an embodiment of the present invention.

[0022] Figure 5 It is a structural diagram of a vehicle charger performance test according to an embodiment of the present invention. DETAILED DESCRIPTION

[0023] To make the objectives, technical solutions, and advantages of the embodiments of the present invention more clear, the technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative efforts shall fall within the scope of protection of the present invention.

[0024] Reference Figure 1 One embodiment of the present invention proposes a vehicle charger performance testing method. By combining virtual simulation with a machine learning model, and building a causal analysis and feedback optimization closed loop from physical process risks to final quality results, it can achieve full-process automation and intelligent control of the testing process, significantly improving the pass rate, reliability, and intelligence level of vehicle charger production testing.

[0025] The method of this embodiment specifically includes: Collect the spatial posture parameters and electrical characteristic parameters of the vehicle charger interface, calculate and generate spatial fine-tuning drive instructions; Docking is performed based on the spatial fine-tuning drive instruction, and consistency verification is performed with a preset docking verification benchmark to obtain a verification result indicating docking success or failure; If the verification result feedback is successfully connected, the multi-channel performance testing process is started to generate full-process performance data; A judgment is made based on the full-process performance data and a preset performance judgment threshold, and differentiated handling instructions are generated and executed.

[0026] By combining virtual simulation with machine learning models and building a closed loop of causal analysis and feedback optimization from physical process risks to final quality results, it is possible to achieve full automation and intelligent control of the testing process, significantly improving the pass rate, reliability and intelligence level of vehicle charger production testing.

[0027] Optionally, the computing and generating of the spatial fine-tuning drive instruction includes: Obtain the real-time three-dimensional coordinates of the vehicle charger interface and perform vector operations to obtain the three-dimensional space deviation vector; Specifically, in the step of obtaining the real-time three-dimensional coordinates of the on-board charger interface and performing vector operations to obtain a three-dimensional spatial deviation vector, an execution unit can call a high-resolution 3D vision sensor to capture the current position of the interface. The sensor outputs a set of real-time position data containing X, Y, Z coordinates and rotation components. A central processor performs a vector subtraction operation on the real-time position data with a target position data preset in the database that represents the ideal docking position, thereby accurately calculating the deviation between the two, as shown in the formula: , in, is the final three-dimensional space deviation vector, which represents the direction and distance between the current position and the target position; is the preset target pose vector; The current pose vector is collected in real time by the 3D vision sensor. This deviation vector provides an accurate error input for subsequent closed-loop control.

[0028] Synchronously collect the contact impedance and signal response status of the vehicle charger interface to generate electrical characteristic parameters; Specifically, in the step of synchronously collecting the contact impedance and signal response status of the vehicle charger interface to generate electrical characteristic parameters, a test control module will use the four-wire method to accurately measure the contact impedance between the terminals when the probe of the alignment mechanism makes initial contact with the charger interface. At the same time, the module will send a weak electrical test pulse and measure the response time or waveform establishment time of its signal loop. These two measurement values, namely the real-time measurement value of the contact impedance and the real-time measurement value of the signal response state, are combined into a data pair to generate an electrical characteristic parameter. This electrical characteristic parameter reflects the potential quality of the electrical connection to be established.

[0029] Based on the three-dimensional spatial deviation vector and the electrical characteristic parameters, the X / Y / Z three-axis fine-tuning amount is calculated to generate a spatial fine-tuning drive instruction.

[0030] Specifically, in the step of calculating the X / Y / Z three-axis fine-tuning amount based on the three-dimensional spatial deviation vector and electrical characteristic parameters to generate spatial fine-tuning drive instructions, an algorithm processing unit adopts an adaptive gain control strategy. This strategy first uses the electrical characteristic parameters generated in the previous step to calculate a normalized electrical quality factor, as shown in the formula: , in, is the electrical quality factor; and is the preset weight coefficient; and It is a real-time measurement value; and is the maximum threshold allowed. Subsequently, the algorithm processing unit uses the electrical quality factor to dynamically adjust the proportional gain of a PID controller to generate an adaptive gain coefficient Finally, the adaptive gain coefficient and the three-dimensional space deviation vector are combined to calculate the final spatial fine-tuning drive instruction including the X, Y, and Z axis fine-tuning amounts, as shown in the formula: , in, is the final generated spatial fine-tuning drive instruction, such as Figure 2 As shown in the figure, the figure intuitively shows in the form of contour lines how the final adaptive gain coefficient is dynamically adjusted nonlinearly and multi-dimensionally according to the real-time combined values ​​of the two core electrical parameters: contact impedance and signal response delay. The high gain area corresponds to the parameter space with better electrical connection quality.

[0031] For example, when a charger is placed on a test station, a 3D vision sensor captures its interface coordinates as (10.1, 5.2, 20.0), while the preset target docking coordinates are (10.0, 5.0, 20.0). A processor calculates the three-dimensional space deviation vector as (-0.1, -0.2, 0.0). At the same time, a test module measures the contact impedance of the interface to be high, resulting in the calculated electrical quality factor Since the electrical quality factor is lower than the normal level, an algorithm processing unit will reduce the adaptive gain coefficient accordingly. , making the physical movement represented by the resulting spatial fine-tuning drive instructions slower and softer. This strategy ensures that the alignment mechanism can push the interface in a prudent manner, avoiding damage while overcoming the high initial impedance through more stable contact, ultimately establishing a reliable electrical connection.

[0032] Optionally, calculating the X / Y / Z three-axis fine-tuning amount and generating the spatial fine-tuning drive instruction includes: Performing kinematics calculation on the three-dimensional space deviation vector to generate a reference driving scheme; Specifically, during the step of kinematically solving the three-dimensional deviation vector to generate a baseline drive solution, a motion planning processor executes an inverse kinematics algorithm. This algorithm aims to resolve the end-position deviation described in a Cartesian coordinate system—the three-dimensional deviation vector—into the specific rotation angles or movement distances required for each joint in the multi-axis alignment mechanism. The mathematical relationship can be abstractly expressed as the formula: , in, is the generated benchmark driving solution, which is usually represented by a set of time series data containing the target position, velocity and acceleration of each joint; Represents the inverse kinematics solution function; is the three-dimensional space deviation vector. This benchmark driving scheme represents the most efficient alignment path in theory without considering any physical constraints or risks.

[0033] Performing virtual simulation on the benchmark driving scheme to obtain predictive collision and stress distribution parameters; Specifically, during the virtual simulation of the baseline drive scenario to obtain predictive crash and stress distribution parameters, a simulation engine executes the baseline drive scenario within a pre-built digital twin model that accurately maps to the physical test fixture. This engine utilizes physical simulation techniques such as Finite Element Analysis (FEA) to simulate the interactions between components during the drive scenario execution. The output of this simulation process can be expressed as: , in, is the obtained predictive collision and stress distribution parameter, which is a set of multiple risk indicators, such as , is the predicted collision probability, is the maximum stress value predicted to be generated on the charger housing or pins; represents the simulation function; is the input benchmark driving scheme; is the digital twin model of the tooling and the product to be tested, such as Figure 3 As shown in the figure, the predicted stress magnitude and the predicted collision probability are coupled and visualized in the form of a three-dimensional surface, allowing managers to simultaneously evaluate two different types of process risks: physical stress and geometric interference in one view.

[0034] The reference drive scheme is optimized and corrected based on the predicted collision and stress distribution parameters and the electrical characteristic parameters, and the X / Y / Z three-axis fine-tuning amounts serving as correction values ​​are calculated and integrated to generate spatial fine-tuning drive instructions.

[0035] Specifically, the baseline drive scheme is optimized and corrected based on the predicted collision and stress distribution parameters and electrical characteristic parameters, and an optimization decision unit executes a multi-objective optimization algorithm. The algorithm aims to minimize the alignment time while ensuring that the predicted collision probability and maximum stress do not exceed the preset safety threshold, and will refer to the electrical characteristic parameters to adjust the strategy of the final contact stage. For example, when the predicted maximum stress value is high, the algorithm will reduce the terminal speed in the baseline drive scheme; when the generated electrical characteristic parameters show poor contact quality, the algorithm may add a low-speed, constant-force "nudge" action at the end of the drive scheme. The output of this optimization and correction process is the final spatial fine-tuning drive instruction, which is a directly executable drive sequence that has been optimized for both safety and quality.

[0036] For example, assuming that the initial position of a charger to be tested results in a large three-dimensional spatial deviation vector, the baseline drive solution calculated by a motion planning processor is a high-speed, direct linear movement. However, a simulation engine discovered in a virtual simulation of the solution that the end point of the high-speed movement path will cause a stress concentration exceeding the safety threshold between a corner of the fixture and the charger casing. At this time, an optimization decision unit will intervene and automatically optimize and correct the baseline drive solution based on the high stress risk, adjusting the original high-speed linear movement to an "L"-shaped compound path that first moves rapidly and then presses vertically downward at a low speed. The final generated spatial fine-tuning drive instruction successfully avoids manufacturing risks that may cause damage to the product appearance while ensuring efficient alignment.

[0037] Optionally, the performing consistency check with a preset docking verification benchmark includes: Extracting geometric contour features of the on-board charger interface and performing similarity calculation with the asymmetric structure template in the docking verification benchmark to generate a structural matching parameter; Specifically, in the step of extracting the geometric contour features of the on-board charger interface and performing similarity calculations to generate structural matching parameters, an image processing unit first uses a high-resolution industrial camera to capture a real-time image of the interface. Subsequently, an interface geometric feature point recognition model extracts a set of key geometric feature point vectors describing the current interface profile. The image processing unit then uses the cosine similarity algorithm to calculate the similarity between the real-time feature vector and the standard feature vector corresponding to the asymmetric structure template preset in the docking verification benchmark. The calculation process is shown in the formula: , in, It is the final structural matching parameter, and its value range is [-1,1]. The closer it is to 1, the higher the geometric matching degree. is the key geometric feature point vector extracted from the real-time image; is the standard feature vector stored in the docking validation benchmark; Represents the calculation of the Euclidean norm of a vector.

[0038] Real-time detection of the contact impedance and signal response delay of the on-board charger interface, and comparison of the impedance and signal response delay with the electrical compliance threshold range in the docking verification benchmark to generate electrical compliance parameters; Specifically, during the step of real-time detection of the interface's contact impedance and signal response delay to generate electrical compliance parameters, an electrical detection unit is responsible for quantifying the electrical quality of the connection. This unit compares the real-time measured contact impedance and signal response delay with the electrical compliance threshold range defined in the docking verification benchmark. To obtain a normalized parameter, a scoring function can be used as shown in the formula: , in, is the generated electrical compliance parameter; and is the preset weight; A step function, when the measured value At the threshold If it is within the range, it outputs 1; if it is outside the range, it outputs a value between 0 and 1.

[0039] A logical AND judgment is performed on the structural matching parameter and the electrical compliance parameter to generate a verification result indicating whether the docking is successful or failed.

[0040] Specifically, in the step of performing a logical AND judgment on the structural matching parameters and the electrical compliance parameters to generate a verification result indicating whether the docking is successful or not, a decision logic unit performs a Boolean logic judgment. This unit compares the structural matching parameters and electrical compliance parameters generated in the first two steps with their respective judgment thresholds. Only when both conditions are met is the docking determined to be successful. The judgment logic is shown in the formula: , in, The final verification result of whether the connection is successful or failed is a Boolean value. and Parameters generated for the previous steps; and These are the preset judgment thresholds for structural matching and electrical compliance, respectively.

[0041] For example, when a charger interface is operated by the alignment mechanism, an image processing unit captures its image and extracts the feature vector, and the calculated structural matching parameter is 0.99. At the same time, an electrical detection unit measures its contact impedance to be 30 milliohms, the signal response delay to be 8 nanoseconds, and the calculated electrical compliance parameter to be 0.97. Assume that the preset structural matching judgment threshold is 0.95 and the electrical compliance judgment threshold is 0.90. A decision logic unit makes a judgment based on the formula. Since 0.99 ≥ 0.95 and 0.97 ≥ 0.90, both conditions are met, so the final verification result is "docking successful", and the subsequent performance test process will be triggered.

[0042] Optionally, generating the structural matching parameters includes: Obtain historical interface image data and historical standard feature point coordinates to obtain a historical interface image dataset; Specifically, in the step of obtaining historical interface image data and historical standard feature point coordinates to obtain a historical interface image dataset, data collection is carried out in a controlled lighting environment. The objects of collection are multiple standard charger interface samples with perfect physical dimensions that have been confirmed by metrology. These samples are photographed from different angles and distances using a high-resolution industrial camera to form an original dataset containing thousands or even tens of thousands of images. Subsequently, engineering technicians use professional image annotation software to accurately mark the pixel coordinates of preset key geometric feature points on each historical interface image based on the product's computer-aided design drawings. Each image and its corresponding coordinate annotation together constitute a data pair, and the collection of all data pairs ultimately constitutes the historical interface image dataset used for model training.

[0043] Taking interface image data as input and standard feature point coordinates as output, a neural network model is established and trained using the historical interface image data set to obtain an interface geometric feature point recognition model; Specifically, in the step of establishing and training a neural network model to obtain an interface geometric feature point recognition model, a convolutional neural network (CNN) architecture, such as U-Net or its variants, is used because it performs well in image segmentation and key point localization tasks. The training process uses image data from a historical interface image dataset as the input of the model and the corresponding standard feature point coordinates as the expected output of the model. Through the backpropagation algorithm, the weight parameters of the network are iteratively optimized with the goal of minimizing the mean square error loss function between the model's predicted coordinates and the true labeled coordinates. The loss function can be expressed as the formula: , in, is the mean square error loss value; is the total number of samples in the training batch; is the coordinate vector of the key geometric feature points predicted by the model for the i-th sample image; is the coordinate vector of the true standard feature point corresponding to the i-th sample. When the loss value converges to a sufficiently small preset value, the training is completed, and the solidified network weight parameters constitute the final interface geometric feature point recognition model.

[0044] Capturing real-time interface image data, inputting the real-time interface image data into the interface geometric feature point recognition model to obtain coordinates of key geometric feature points; Specifically, in the step of capturing real-time interface image data and inputting it into the model to obtain the coordinates of key geometric feature points, when a charger under test enters the test station, the industrial camera deployed at that station is triggered to capture one or more frames of real-time images of the charger interface. After preprocessing, these images are input as a tensor into the interface geometric feature point recognition model trained in the previous step. The model performs a forward propagation, and its output layer is directly decoded into a set of real-time coordinate values ​​of key geometric feature points. This process is fully automated and typically takes milliseconds.

[0045] The coordinates of the key geometric feature points are compared with the asymmetric structure template in terms of spatial position and topological relationship, and structural matching parameters are calculated and generated.

[0046] Specifically, in the step of comparing and analyzing the coordinates of key geometric feature points with the asymmetric structure template to calculate and generate the structural matching parameters, an analysis unit compares the real-time key geometric feature point coordinates output by the previous model with the asymmetric structure template stored in the docking verification benchmark. In order to comprehensively evaluate the similarity of the overall structure rather than the error of a single point, the cosine similarity algorithm can be used. By flattening the two sets of coordinate points into high-dimensional vectors and calculating the cosine value of their angle, a quantitative structural matching parameter ranging from -1 to 1 can be obtained. This parameter can robustly evaluate the accuracy of the overall contour, size and internal topological relationship of the interface.

[0047] For example, a charger to be tested is placed on the production line. An industrial camera on top captures a 1024×1024 pixel image of the interface. This image is fed into a pre-trained U-Net model, which quickly outputs the coordinates of eight key vertices, such as [(102,305),(410,306),...]. Simultaneously, an analysis unit reads the standard asymmetric structural template for this model of charger from a database, which also defines the ideal coordinates of these eight vertices. By forming two 16-dimensional vectors from the coordinates of these two sets of eight points and calculating the cosine similarity between them, the final structural matching parameter is 0.9992. This high score indicates that the geometry of the current charger interface is highly consistent with the standard design.

[0048] Optionally, generating full-process performance data includes: Apply a simulated load to the power bus of the on-board charger and collect voltage and current responses to obtain basic electrical performance parameters; Specifically, in the step of applying a simulated load to the power bus of the on-board charger to obtain basic electrical performance parameters, a programmable DC electronic load integrated in the test fixture is activated. This electronic load simulates different working conditions during the charging process, such as no-load, 50% rated load, and 100% full load, according to the preset test sequence. Under each load state, a high-precision data acquisition module will synchronously measure the output voltage of the charger. and output current , and monitor its input power A calculation unit can calculate the product's charging efficiency in real time based on these measured values, as shown in the formula: , in, The voltage, current, and charging efficiency data collected under different loads together constitute the basic electrical performance parameters that reflect its core power conversion capabilities.

[0049] Send protocol test instructions to the data bus of the vehicle charger and capture the returned signal waveform to obtain protocol interaction data; Specifically, in the step of sending a protocol test command to the data bus of the on-board charger to obtain protocol interaction data, a protocol analyzer or simulated electronic control unit module sends a standard communication handshake or data request command to the charger under test via the controller area network bus or local interconnect network bus. A high-speed oscilloscope or data acquisition card will synchronously capture the complete digital waveform of the response signal returned by the charger. A time analysis unit accurately calculates the time difference between sending the command and receiving the valid response, that is, the protocol response delay, as shown in the formula: , in, The protocol response is delayed; The timestamp of receiving a valid response signal; This is the timestamp of when the test command was sent. This response delay, along with the captured signal waveform data, constitutes the protocol interaction data used to assess whether the charger's communication function is normal and whether its response is timely.

[0050] The basic electrical performance parameters and the protocol interaction data are aggregated and structured to generate full-process performance data.

[0051] Specifically, in the step of aggregating and structuring the basic electrical performance parameters and protocol interaction data to generate full-process performance data, a data processing module aggregates and formats all the discrete data generated in the first two steps. This module integrates basic electrical performance parameters such as voltage, current, and charging efficiency under different loads with protocol interaction data such as protocol response delay and signal waveform data into a unified, structured data record, such as a JSON object or XML file. This complete data record is the final generated full-process performance data, which comprehensively describes the performance of the product in this test and provides a complete and objective basis for subsequent quality judgment links.

[0052] For example, when a charger enters the performance test process, an electronic load is first loaded at 50% of its rated power. At this time, its output voltage is measured to be 14.4V, the output current is 20A, and the calculated charging efficiency is 94.5%. These data are recorded as part of the basic electrical performance parameters. Next, a CAN bus analyzer sends a request command to read the battery temperature to the charger and receives a correctly formatted response frame 5 milliseconds later. This response time and the waveform data of the response frame are recorded as protocol interaction data. Finally, a data processing module integrates this information into a JSON object. This complete JSON object is the full-process performance data generated by this test and is transmitted to the next judgment unit.

[0053] Optionally, generating and executing a differential handling instruction includes: Performing a hierarchical comparison of the full-process performance data with the performance judgment threshold to generate a multi-dimensional performance judgment result; Specifically, in the step of performing a hierarchical comparison of the full-process performance data with the performance judgment threshold to generate a multi-dimensional performance judgment result, a judgment analysis unit sets multiple threshold intervals for each key performance parameter, such as "optimal interval", "acceptable interval" and "failure interval", to achieve a refined hierarchical evaluation. This unit compares the measured values ​​in the full-process performance data with these grading thresholds item by item, and generates a qualitative rating for each performance parameter. All individual ratings are aggregated into a data structure, which together constitute the multi-dimensional performance judgment result, and its structure can be expressed as a formula: , in, It is the generated multi-dimensional performance judgment result; The key name representing the nth performance parameter, such as "charging efficiency"; The performance parameter is rated, such as "optimal" or "failed." This step provides detailed, multi-dimensional input for subsequent comprehensive decision-making.

[0054] Analyze the multi-dimensional performance determination results to obtain a disposition level including qualified, retest required, or unqualified information; Specifically, in the step of parsing the multi-dimensional performance judgment results to obtain the disposal level, a rule engine is responsible for executing a set of preset business logic rules. These rules define how to determine the final disposal level of the product based on the combination of each individual rating. For example, the rule can be set as follows: when all individual ratings are "optimal" or "acceptable", the final disposal level is judged to be "qualified"; when there is no "failed" rating but there is at least one "needs attention" rating, it is judged to be "re-inspected"; if any rating is "failed", it is directly judged to be "unqualified". This parsing process can be abstracted into a function as shown in the formula: , in, It is the final output, containing the disposition level of qualified, retested or unqualified information; Represents the decision function executed by the rule engine; It is the multi-dimensional performance judgment result of the input.

[0055] Based on the disposal levels, identification instructions for qualified products, re-inspection prompt instructions for products requiring re-inspection, and isolation disposal instructions for unqualified products are constructed respectively to generate differentiated disposal instructions.

[0056] Specifically, in the step of constructing different instructions based on the disposal level to generate differentiated disposal instructions, an instruction generation module is responsible for converting the abstract disposal level obtained in the previous step into machine instructions that can be directly executed by downstream devices. This module matches and constructs a data packet containing a clear action, target address and related data based on the input disposal level. For example, an identification instruction for qualified products can be a data structure containing the target device address "Laser_1" and the data to be printed "SN12345"; and an isolation instruction for unqualified products can be a data structure containing the target device address "Reject_Bin_A" and the fault code "E05". The final generated data packet is the differentiated disposal instruction, and its generation process can be expressed as the formula: , in, is the generated differentiated disposal instruction; Represents the instruction construction function; is the input disposal level.

[0057] For example, after a charger test is completed, its multi-dimensional performance judgment result is determined to be {voltage stability: "optimal", communication response: "needs attention"}. A rule engine interprets this judgment result as a handling level of "needs re-inspection" based on internal rules. Subsequently, an instruction generation module receives the "needs re-inspection" level and immediately constructs a differentiated handling instruction with the content of {action: "route", destination: "re-inspect_station_02", reason_code: "CAN_TIMEOUT"}. The instruction is sent to the production line control system, and a physical sorting baffle acts according to the instruction to guide the charger to the No. 2 re-inspection station for technicians to conduct targeted inspections.

[0058] Optionally, the method further includes: Performing unsupervised cluster analysis on the data of the defective products, extracting common failure features, and generating failure mode feature vectors; Specifically, in the step of performing unsupervised cluster analysis on data with a disposal level of non-conforming products to generate failure mode feature vectors, a data analysis module regularly collects the full-process performance data of all products judged as "non-conforming" from the historical database. This module uses an unsupervised clustering analysis algorithm, such as K-means clustering, to aggregate these high-dimensional performance data. The goal of this algorithm is to minimize the sum of the squares of the distances from each data point to the centroid of the cluster to which it belongs, thereby automatically classifying products with similar failure performance into one category. Its objective function can be expressed as the formula: , in, is the total within-cluster sum of squares of the clustering algorithm; is the preset number of failure mode clusters; represents the i-th failure mode cluster; is the full-process performance data vector of a defective product; is the centroid of the ith cluster. The centroid is an extracted common failure feature, which mathematically represents a typical failure mode and is defined as a specific failure mode feature vector.

[0059] The spatial posture parameters and electrical characteristic parameters associated with the failure mode feature vector are traced back to construct and generate a parameter-failure mode association rule set.

[0060] Specifically, in the step of tracing the associated initial parameters to construct and generate a parameter-failure mode association rule set, a rule mining engine performs a data mining task. For each failure mode cluster clustered in the previous step, the engine traces back the initial parameters collected at the beginning of the test for all products in the cluster, including spatial posture parameters and electrical characteristic parameters. This constitutes a large historical paired data set of "(initial parameter combination)-(failure mode)". Then, the engine uses an association rule mining algorithm, such as the Apriori algorithm, to train the data set to discover frequently occurring "item sets" and generate association rules in the form of the formula: , in, is a generated association rule; Represents the initial parameter condition combination that leads to failure, such as "contact impedance > 150mΩ" and "Z-axis deviation > 0.2mm"; This is the failure mode that is frequently associated with this combination of conditions, such as "communication timeout failure"; is the support of the rule, indicating the frequency of the combination in the total sample; is the confidence of the rule, which indicates the probability that the result will occur when the condition occurs. All valid rules that meet the minimum support and confidence thresholds together constitute the final parameter-failure mode association rule set, such as Figure 4 As shown in the figure, the graph is in the form of a hierarchical clustering heat map. It not only intuitively shows the correlation strength between a specific initial parameter combination and a specific failure mode through the depth of the color block, but also automatically reveals the similarities between different failure modes and the inherent correlation between different risk sources through the tree-like clustering diagram on both sides of the rows and columns, providing deeper insights for the aggregate analysis of the root causes.

[0061] For example, after analyzing data from 500 defective products accumulated over the past week, a data analysis module discovered three major failure clusters using the K-means clustering algorithm. One of these clusters was identified as a failure mode characterized by "excessive voltage drop under high load." A rule mining engine then traced the initial test parameters of the 150 products belonging to this cluster and ran the Apriori algorithm. The algorithm discovered a high-confidence rule: If a product's "initial contact impedance" is high and the Y-axis component of its "three-dimensional spatial deviation vector" is negative, there is an 85% probability that the product will experience an "excessive voltage drop under high load" failure. This rule is automatically stored in the parameter-failure mode association rule set, providing valuable knowledge for subsequent analysis.

[0062] Optionally, the method further includes: Inputting the spatial posture parameters and electrical characteristic parameters collected in real time into the parameter-failure mode association rule set for inference to generate a predictive failure risk level; Specifically, at the start of a new round of testing, after initial parameters are collected in real time, a risk inference engine immediately invokes the parameter-failure mode association rule set. This engine uses the currently collected spatial posture parameters and electrical characteristic parameters as input, performing high-speed queries and matching within the rule set. If the current parameter combination matches one or more high-confidence association rules, the engine calculates a comprehensive, quantitative predictive failure risk level based on the confidence scores of these rules. This process can be abstracted into the formula: , in, It is the generated predictive failure risk level, which can be a value between 0 and 1, with a higher value indicating a greater predicted failure probability; represents the risk inference function; It is the initial parameter set containing real-time spatial attitude and electrical characteristics; It is a parameter-failure mode association rule set.

[0063] performing a fusion calculation on the predictive failure risk level and the predictive collision and stress distribution parameters to generate a feedforward adjustment parameter; Specifically, during the step of merging the predictive failure risk level with the predictive collision and stress distribution parameters to generate feedforward adjustment parameters, a decision fusion unit is responsible for balancing two different dimensions of forward-looking risks: the "predictive collision and stress distribution parameters" representing physical process safety, and the "predictive failure risk level" generated in the previous step, representing product quality trends. This unit uses a multi-objective decision function to perform a weighted fusion of the two risks to generate a set of feedforward adjustment parameters for proactive intervention. The calculation process can be expressed as the formula: , in, is the final set of feedforward adjustment parameters; and are the preset weights for quality risk and physical process risk respectively; and It is a function that maps various risk levels or parameters to specific adjustment values; and These are two types of risk data input.

[0064] The feedforward adjustment parameters are used to dynamically optimize the docking verification benchmark and the performance judgment threshold.

[0065] Specifically, in the step of dynamically optimizing the test benchmark and threshold using feedforward adjustment parameters, a parameter configuration module receives the feedforward adjustment parameters generated in the previous step. Based on these parameters, the module makes a one-time, dynamic, and personalized adjustment to the relevant standards in this round of testing. For example, if the feedforward adjustment parameters indicate that the voltage stability judgment needs to be tightened, the module will temporarily adjust the allowable range of voltage fluctuations in the performance judgment threshold from ±1% to ±0.5%. Similarly, if it is indicated that the physical verification standard needs to be adjusted, the relevant parameters in the docking verification benchmark will be modified. In this way, the test standard is no longer fixed, but is dynamically optimized "tailor-made" for each product to be tested.

[0066] For example, a charger under test entered a workstation. Its initial parameters, collected in real time, matched a high-confidence rule in the parameter-failure mode association rule set, indicating that this parameter combination was highly correlated with "prone to failure in the communication protocol test." A risk inference engine generated a predictive failure risk rating of 0.9 based on this information. Simultaneously, virtual simulations indicated that the physical risk of this alignment process was low. A decision fusion unit, after comprehensive assessment, determined that quality risk was the primary concern during this test and generated a set of feedforward adjustment parameters designed to strictly inspect communication functionality. Upon receiving this instruction, a parameter configuration module temporarily dynamically optimized and tightened the performance threshold for "CAN bus response delay" from "≤10ms" to "≤5ms" during this test. This reduced the "tolerance" for communication performance in this round of testing, increasing the probability of detecting potential defects predicted by the initial parameters.

[0067] Based on the same inventive concept, the present invention also provides a vehicle charger performance test system, such as Figure 5 As shown, the system includes: The initial parameter perception and command generation module is used to collect the spatial posture parameters and electrical characteristic parameters of the vehicle charger interface, calculate and generate spatial fine-tuning drive instructions; A docking and verification module, configured to execute docking based on the spatial fine-tuning drive instruction and perform consistency verification with a preset docking verification benchmark to obtain a verification result indicating docking success or failure; A multi-channel performance detection module is used to start the multi-channel performance detection process and generate full-process performance data if the verification result feedback indicates successful docking; The data judgment and differentiated processing module is used to make judgments based on the full-process performance data and a preset performance judgment threshold, and to generate and execute differentiated processing instructions.

[0068] It should be noted that the functional division and information interaction between the aforementioned modules are logical. Physically, they can be integrated into the same software platform or deployed in a distributed manner. The connections between them represent data and control flows, designed to collaboratively achieve the dynamic optimization of building energy consumption of the present invention. The foregoing description is merely an exemplary embodiment of the present invention and is not intended to limit its scope.

Claims

1. A vehicle charger performance testing method, characterized in that: The method comprises: Collect the spatial posture parameters and electrical characteristic parameters of the vehicle charger interface, calculate and generate spatial fine-tuning drive instructions; Docking is performed based on the spatial fine-tuning drive instruction, and consistency verification is performed with a preset docking verification benchmark to obtain a verification result indicating docking success or failure; If the verification result feedback is successfully connected, the multi-channel performance testing process is started to generate full-process performance data; A judgment is made based on the full-process performance data and a preset performance judgment threshold, and differentiated handling instructions are generated and executed.

2. A vehicle charger performance testing method according to claim 1, characterized in that: The calculation and generation of the spatial fine-tuning drive instruction includes: Obtain the real-time three-dimensional coordinates of the vehicle charger interface and perform vector operations to obtain the three-dimensional space deviation vector; Synchronously collect the contact impedance and signal response status of the vehicle charger interface to generate electrical characteristic parameters; Based on the three-dimensional spatial deviation vector and the electrical characteristic parameters, the X / Y / Z three-axis fine-tuning amount is calculated to generate a spatial fine-tuning drive instruction.

3. A vehicle charger performance testing method according to claim 2, characterized in that: Calculating the X / Y / Z three-axis fine-tuning amount and generating the spatial fine-tuning drive instruction includes: Performing kinematics calculation on the three-dimensional space deviation vector to generate a reference driving scheme; Performing virtual simulation on the benchmark driving scheme to obtain predictive collision and stress distribution parameters; The reference drive scheme is optimized and corrected based on the predicted collision and stress distribution parameters and the electrical characteristic parameters, and the X / Y / Z three-axis fine-tuning amounts serving as correction values ​​are calculated and integrated to generate spatial fine-tuning drive instructions.

4. A vehicle charger performance testing method according to claim 1, characterized in that: The consistency check with the preset docking verification benchmark includes: Extracting geometric contour features of the on-board charger interface and performing similarity calculation with the asymmetric structure template in the docking verification benchmark to generate a structural matching parameter; Real-time detection of the contact impedance and signal response delay of the on-board charger interface, and comparison of the impedance and signal response delay with the electrical compliance threshold range in the docking verification benchmark to generate electrical compliance parameters; A logical AND judgment is performed on the structural matching parameter and the electrical compliance parameter to generate a verification result indicating whether the docking is successful or failed.

5. A vehicle charger performance testing method according to claim 4, characterized in that: The generated structure matching parameters include: Obtain historical interface image data and historical standard feature point coordinates to obtain a historical interface image dataset; Taking interface image data as input and standard feature point coordinates as output, a neural network model is established and trained using the historical interface image data set to obtain an interface geometric feature point recognition model; Capturing real-time interface image data, inputting the real-time interface image data into the interface geometric feature point recognition model to obtain coordinates of key geometric feature points; The coordinates of the key geometric feature points are compared with the asymmetric structure template in terms of spatial position and topological relationship, and structural matching parameters are calculated and generated.

6. A vehicle charger performance testing method according to claim 1, characterized in that: Generating full-process performance data includes: Apply a simulated load to the power bus of the on-board charger and collect voltage and current responses to obtain basic electrical performance parameters; Send protocol test instructions to the data bus of the vehicle charger and capture the returned signal waveform to obtain protocol interaction data; The basic electrical performance parameters and the protocol interaction data are aggregated and structured to generate full-process performance data.

7. A vehicle charger performance testing method according to claim 3, characterized in that: The generating and executing the differential handling instruction includes: Performing a hierarchical comparison of the full-process performance data with the performance judgment threshold to generate a multi-dimensional performance judgment result; Analyze the multi-dimensional performance determination results to obtain a disposition level including qualified, retest required, or unqualified information; Based on the disposal levels, identification instructions for qualified products, re-inspection prompt instructions for products requiring re-inspection, and isolation disposal instructions for unqualified products are constructed respectively to generate differentiated disposal instructions.

8. A vehicle charger performance testing method according to claim 7, characterized in that: The method further comprises: Performing unsupervised cluster analysis on the data of the defective products, extracting common failure features, and generating failure mode feature vectors; The spatial posture parameters and electrical characteristic parameters associated with the failure mode feature vector are traced back to construct and generate a parameter-failure mode association rule set.

9. A vehicle charger performance testing method according to claim 8, characterized in that: The method further comprises: Inputting the spatial posture parameters and electrical characteristic parameters collected in real time into the parameter-failure mode association rule set for inference to generate a predictive failure risk level; performing a fusion calculation on the predictive failure risk level and the predictive collision and stress distribution parameters to generate a feedforward adjustment parameter; The feedforward adjustment parameters are used to dynamically optimize the docking verification benchmark and the performance judgment threshold.

10. A vehicle charger performance test system, applied to a vehicle charger performance test method according to any one of claims 1 to 9, characterized in that: The system comprises: The initial parameter perception and command generation module is used to collect the spatial posture parameters and electrical characteristic parameters of the vehicle charger interface, calculate and generate spatial fine-tuning drive commands; A docking and verification module, configured to execute docking based on the spatial fine-tuning drive instruction and perform consistency verification with a preset docking verification benchmark to obtain a verification result indicating docking success or failure; A multi-channel performance detection module is used to start the multi-channel performance detection process and generate full-process performance data if the verification result feedback indicates successful docking; The data judgment and differentiated processing module is used to make judgments based on the full-process performance data and a preset performance judgment threshold, and to generate and execute differentiated processing instructions.

Images