Method and system for testing electrical performance of automobile part
Through multi-dimensional electrical performance data processing and transfer learning algorithms, a dynamic threshold adjustment model is built, which solves the data singularity, adaptability and generalization problems in the electrical performance test of automotive parts, and achieves efficient and accurate electrical performance evaluation.
Patent Information
- Application Number
- CN202510594117.5
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2025-05-09
- Publication Date
- 2025-07-11
- Estimated Expiration
- Not applicable · inactive patent
AI Technical Summary
The existing electrical performance testing methods for automotive parts have problems such as single data dimensions, poor adaptability of static thresholds, limited feature extraction capabilities and insufficient generalization across scenarios, resulting in insufficient detection capabilities and misjudgment and missed inspections.
Multidimensional electrical performance data acquisition and preprocessing are adopted, combined with signal conversion analysis and feature extraction network, a multidimensional fault characterization space is built, and a dynamic threshold adjustment prediction model is built using transfer learning algorithm to realize electrical performance evaluation.
It improves the detection ability of occasional faults, reduces misjudgment and missed detection, improves the efficiency and accuracy of the test, and adapts to the electrical performance evaluation of different environments and models.
Smart Images
Figure CN120294473A_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to the technical field of electrical performance testing, and specifically to a method and system for testing the electrical performance of automotive parts. Background Art
[0002] In the field of electrical performance testing of automotive parts, traditional methods mainly rely on preset fixed thresholds (such as insulation resistance ≥ 100 MΩ, leakage current of withstand voltage test ≤ 10 mA) and manual experience to analyze test data. Such methods have the following technical bottlenecks: Single-dimensional data: Traditional testing equipment usually only collects basic electrical parameters such as voltage, current, and resistance, lacking synchronous analysis of time-domain waveforms (such as transient spikes, signal jitters) and frequency-domain characteristics (such as harmonic distortion, EMI interference), resulting in insufficient detection ability for occasional faults (such as pulse noise caused by poor contact).
[0003] Poor adaptability of static thresholds: Existing technologies adopt fixed threshold judgment criteria without considering the non-linear effects of environmental factors (temperature, humidity, vibration) on electrical performance. For example, the resistance value of insulating materials decreases exponentially with increasing temperature in a high-temperature environment, while traditional methods still mechanically apply normal-temperature thresholds, prone to misjudgment or missed detection.
[0004] Limited feature extraction ability: Conventional preprocessing methods only perform basic operations such as data filtering and normalization, without constructing a correlation feature space for multi-modal data (electrical parameters, waveforms, environmental data), making it difficult to effectively characterize complex fault modes (such as controller logic errors caused by the coupling of EMC interference and power supply ripple).
[0005] Insufficient cross-scenario generalization: For different vehicle models or new parts, traditional testing systems need to manually reset thresholds and testing procedures again, unable to achieve adaptive parameter adjustment through historical data migration, resulting in a long testing protocol development cycle and high costs.
[0006] Therefore, a method and system for testing the electrical performance of automotive parts are provided. Summary of the Invention
[0007] In order to solve the above technical problems, the purpose of the present invention is to provide a method and system for testing the electrical performance of automotive parts.
[0008] In order to achieve the above purpose, the present invention provides the following technical solutions: A method for testing the electrical performance of automotive parts, comprising: Obtaining multi-dimensional electrical performance data of automotive parts in a test environment, and performing preprocessing to obtain multi-dimensional preprocessed electrical performance data, and performing signal conversion analysis on the multi-dimensional preprocessed electrical performance data to obtain a time-domain waveform signal of the electrical performance data and a frequency-domain waveform signal of the electrical performance data; Synchronously input multi-dimensional preprocessed electrical performance data, environmental data, time-domain waveform signals of electrical performance data, and frequency-domain waveform signals of electrical performance data into a feature extraction network to construct a multi-dimensional fault characterization space and generate multi-dimensional feature electrical performance data; Based on a transfer learning algorithm and multi-dimensional feature electrical performance data, construct a threshold adjustment prediction model for electrical performance data to obtain a dynamically adjusted decision threshold for automotive parts under current environmental conditions; According to the dynamically adjusted decision threshold and multi-dimensional preprocessed electrical performance data, construct an electrical performance evaluation prediction model, obtain the electrical performance operation ability value of automotive parts, and further obtain the test results of automotive parts.
[0009] According to one preferred embodiment of the present invention, the process of obtaining multi-dimensional electrical performance data of automotive parts in a test environment includes: Set up a multi-channel data acquisition device, including a data storage unit, and set an acquisition period, where the acquisition period includes several acquisition moments; collect voltage data, current data, impedance data, and power consumption data according to the acquisition moments and transmit them to the data storage unit for storage, denoted as multi-dimensional electrical performance data.
[0010] According to one preferred embodiment of the present invention, the process of preprocessing multi-dimensional electrical performance data includes: Obtain the multi-dimensional electrical performance data in the data storage unit; Based on a data processing tool, count the number and positions of missing values in each column of the multi-dimensional electrical performance data; smooth the data by calculating the average value within a data window to reduce the influence of noise; select a window of size n, calculate the average value of the data within the window in turn, and use it as the new value of the data point at the center of the window; draw a box plot of the multi-dimensional electrical performance data, where the range in the box plot is usually 1.5 times the interquartile range, and data points outside this range are regarded as outliers, and based on data smoothing technology, eliminate the outliers; based on Z-score standardization, standardize the multi-dimensional electrical performance data after outlier detection and processing, and denote it as multi-dimensional preprocessed electrical performance data.
[0011] According to one preferred embodiment of the present invention, the process of signal conversion analysis of multi-dimensional preprocessed electrical performance data includes: Each data point in the multi-dimensional preprocessed electrical performance data corresponds to a specific sampling moment, and then a time series is generated; Set the sampling frequency, determine the sampling interval, and based on the sampling interval, generate a corresponding time series for each data point in the multi-dimensional preprocessed electrical performance data; Correspond each electrical performance data in the multi-dimensional preprocessed electrical performance data with the generated time series respectively, and plot a curve that changes with time; taking voltage as an example, the abscissa is time and the ordinate is the voltage value, and the drawn curve is the time-domain waveform signal of the electrical performance data of the voltage. Based on the fast Fourier transform algorithm, process the time-domain waveform signals of the electrical performance data corresponding to each electrical performance data in the multi-dimensional preprocessed electrical performance data, and obtain a graph with frequency as the abscissa and amplitude as the ordinate. The drawn graph is the frequency-domain waveform signal of the electrical performance data.
[0012] According to one preferred embodiment of the present invention, the process of constructing a multi-dimensional fault characterization space includes: Obtain multi-dimensional preprocessed electrical performance data, environmental data, time-domain waveform signals of electrical performance data, and frequency-domain waveform signals of electrical performance data; Perform vectorization processing on the multi-dimensional preprocessed electrical performance data and environmental data. The specific process is as follows: The environmental data includes temperature and humidity; Arrange the multi-dimensional preprocessed electrical performance data and environmental data in sequence to form a one-dimensional vector; The one-dimensional vector can be expressed as , where represents temperature, represents humidity, represents voltage data, represents current data, represents impedance data, represents power consumption data; Perform serialization processing on the time-domain waveform signal of the electrical performance data and the frequency-domain waveform signal of the electrical performance data. The specific process is as follows: Preset a determined fixed length; If the lengths of the time-domain waveform signal of the electrical performance data and the frequency-domain waveform signal of the electrical performance data exceed the fixed length, truncation processing is required. Starting from the starting positions of the time-domain waveform signal of the electrical performance data and the frequency-domain waveform signal of the electrical performance data, intercept a part of the fixed length as the final sequence to obtain the time-domain waveform signal sequence of the electrical performance data and the frequency-domain waveform signal sequence of the electrical performance data; Synchronously input the one-dimensional vector, the time-domain waveform signal sequence of the electrical performance data, and the frequency-domain waveform signal sequence of the electrical performance data into the feature extraction network, construct a multi-dimensional fault characterization space, and generate multi-dimensional feature electrical performance data.
[0013] According to one preferred embodiment of the present invention, the process of constructing an electrical performance data threshold adjustment prediction model includes: Obtain multi-dimensional feature electrical performance data of several historical acquisition periods; And preset standard electrical performance data; Group and label the multi-dimensional characteristic electrical performance data of several groups of historical acquisition cycles, denoted as which is a natural number; Take groups of multi-dimensional characteristic electrical performance data of several groups of historical acquisition cycles and the standard electrical performance data as sample data, and is a natural number less than , and use the sample data to obtain the sample data mean, denoted as the sample set; Take the multi-dimensional characteristic electrical performance data of the remaining several groups of historical acquisition cycles and the standard electrical performance data as the test set; According to the sample set and the test set, form a training sample set; Based on the convolutional neural network, construct a standard prediction model; the standard prediction model includes a first-level standard prediction model and a second-level standard prediction model; And input the training sample set into the first-level standard prediction model, train the first-level standard prediction model, obtain the trained first-level standard prediction model, and denote the trained first-level standard prediction model as the electrical performance data threshold adjustment prediction model, and obtain the dynamic adjustment determination threshold of the automotive parts under the current environmental conditions , the dynamic adjustment determination threshold is: ; where represents the temperature, represents the ideal operating temperature of the automotive part, represents the proportional coefficient of the adjustment temperature influence degree, represents the humidity, represents the voltage data, represents the current data, represents the impedance data, represents the power consumption data.
[0014] According to one preferred embodiment of the present invention, the process of constructing the electrical performance evaluation prediction model includes: Preset the standard determination threshold ; According to the standard determination threshold , obtain the multi-dimensional preprocessed electrical performance data when the actual dynamic adjustment determination threshold is ; Obtain several groups of multi-dimensional preprocessed electrical performance data; Based on the transfer learning algorithm, retain the underlying feature extraction layer of the electrical performance data threshold adjustment prediction model, and use the several groups of multi-dimensional preprocessed electrical performance data and the underlying feature extraction layer as the input layer, and input them into the secondary standard prediction model for training to obtain the trained secondary standard prediction model. Denote the trained secondary standard prediction model as the electrical performance evaluation prediction model, and obtain the electrical performance operation ability value of the automotive part , the electrical performance operation ability value is: Among them, represents the optimal operating temperature of the automotive part, represents the temperature range parameter, represents the maximum value that the environmental humidity may reach, represents the rated power consumption of the automotive part, 、 are weight coefficients.
[0015] According to one preferred embodiment of the present invention, the process of obtaining the test result of the automotive part includes: Preset the standard electrical performance operation ability value ; If the electrical performance operation ability value , then the electrical performance operation ability value of the automotive part meets the standard; If the electrical performance operation ability value , then the electrical performance operation ability value of the automotive part does not meet the standard.
[0016] To achieve at least one of the above invention purposes, the present invention further provides an electrical performance test system for automotive parts. The system executes the above-mentioned electrical performance test method for automotive parts, including: an electrical performance data acquisition module, an electrical performance data processing module, an electrical performance data analysis module, and an intelligent test module; The electrical performance data acquisition module is used to obtain multi-dimensional electrical performance data of the automotive part in the test environment; The electrical performance data processing module is used to preprocess the multi-dimensional electrical performance data to obtain multi-dimensional preprocessed electrical performance data; The electrical performance data analysis module is used to perform signal conversion analysis on the multi-dimensional preprocessed electrical performance data to obtain the time-domain waveform signal of the electrical performance data and the frequency-domain waveform signal of the electrical performance data; The intelligent test module is used to construct a multi-dimensional fault characterization space, generate multi-dimensional characteristic electrical performance data, and construct an electrical performance data threshold adjustment prediction model to obtain the dynamic adjustment decision threshold of the automotive part under the current environmental conditions. According to the dynamic adjustment decision threshold and the multi-dimensional preprocessed electrical performance data, construct an electrical performance evaluation prediction model, and obtain the electrical performance operation ability value of the automotive part.
[0017] The present invention further provides a computer-readable storage medium storing a computer program executable by a processor to implement the above-mentioned method for testing the electrical performance of automotive parts.
[0018] Compared with the prior art, the beneficial effects of the present invention are as follows: obtaining multi-dimensional electrical performance data of automotive parts in a test environment, performing preprocessing to obtain multi-dimensional preprocessed electrical performance data, performing signal conversion analysis on the multi-dimensional preprocessed electrical performance data to obtain a time-domain waveform signal of the electrical performance data and a frequency-domain waveform signal of the electrical performance data; reducing the insufficient detection ability caused by accidental failures (such as pulse noise caused by poor contact); Inputting the multi-dimensional preprocessed electrical performance data, environmental data, the time-domain waveform signal of the electrical performance data, and the frequency-domain waveform signal of the electrical performance data into a feature extraction network, constructing a multi-dimensional fault characterization space, and generating multi-dimensional characteristic electrical performance data; based on a transfer learning algorithm and the multi-dimensional characteristic electrical performance data, constructing an electrical performance data threshold adjustment prediction model to obtain a dynamically adjusted decision threshold; according to the dynamically adjusted decision threshold, constructing an electrical performance evaluation prediction model, obtaining an electrical performance operation ability value, and further obtaining a test result, which characterizes complex fault modes (such as controller logic errors caused by the coupling of EMC interference and power supply ripple), improving efficiency and accuracy. BRIEF DESCRIPTION OF THE DRAWINGS
[0019] In order to more clearly illustrate the technical solutions in the embodiments of the present application or the prior art, the following will briefly introduce the drawings required for use in the embodiments. Obviously, the drawings described below are only some embodiments recorded in the present invention, and those of ordinary skill in the art can also obtain other drawings based on these drawings.
[0020] Figure 1 It is a schematic diagram of a method for testing the electrical performance of automotive parts.
[0021] Figure 2 It is a module diagram of a system for testing the electrical performance of automotive parts. DETAILED DESCRIPTION OF THE EMBODIMENTS
[0022] The following description is used to disclose the present invention so that those skilled in the art can implement the present invention. The preferred embodiments described below are only examples, and those skilled in the art can think of other obvious variations. The basic principles defined in the following description can be applied to other embodiments, variations, improvements, equivalent solutions, and other technical solutions that do not depart from the spirit and scope of the present invention.
[0023] It is understood that the term "a" should be understood as "at least one" or "one or more". That is, in one embodiment, the number of an element can be one, while in other embodiments, the number of the element can be multiple. The term "a" should not be understood as a limitation on the quantity.
[0024] As Figure 1 shown, an electrical performance testing method for automotive parts includes the following steps: Obtain multi-dimensional electrical performance data of automotive parts in a test environment, and perform preprocessing to obtain multi-dimensional preprocessed electrical performance data. Perform signal conversion analysis on the multi-dimensional preprocessed electrical performance data to obtain the time-domain waveform signal and frequency-domain waveform signal of the electrical performance data; Synchronously input the multi-dimensional preprocessed electrical performance data, environmental data, time-domain waveform signal of the electrical performance data, and frequency-domain waveform signal of the electrical performance data into a feature extraction network to construct a multi-dimensional fault characterization space and generate multi-dimensional characteristic electrical performance data; Based on the transfer learning algorithm and the multi-dimensional characteristic electrical performance data, construct an electrical performance data threshold adjustment prediction model to obtain the dynamic adjustment decision threshold of automotive parts under the current environmental conditions; According to the dynamic adjustment decision threshold and the multi-dimensional preprocessed electrical performance data, construct an electrical performance evaluation prediction model, and obtain the electrical performance operation ability value of the automotive parts, and then obtain the test result of the automotive parts.
[0025] It should be further noted that in the specific implementation process, the specific process of obtaining the multi-dimensional electrical performance data of automotive parts in a test environment includes: According to the test standards and requirements, apply different voltage and current excitations to the automotive parts to be tested through a power supply device to simulate the power supply states of the vehicle under various conditions such as starting, normal driving, accelerating, decelerating, and idling; Set up a multi-channel data acquisition device, including a data storage unit, a voltage acquisition unit, a current acquisition unit, an impedance acquisition unit, and a power consumption acquisition unit, and set an acquisition period, and the acquisition period includes several acquisition moments; The voltage acquisition unit connects a high-precision voltmeter in parallel with the power input terminal or a specific voltage test point of the automotive parts to be tested, and measures the input voltage and the voltage data of the internal key nodes of the parts in real time under different excitation conditions, and transmits the voltage data to the data storage unit for storage; The current acquisition unit connects an ammeter in series in the circuit to ensure that all the current can flow through the ammeter, accurately measure the magnitude of the current passing through the automotive parts, and record the current data in real time during the process of applying different excitation conditions, and transmit the current data to the data storage unit for storage; The impedance acquisition unit uses an impedance analyzer. According to the operation instructions of the instrument, the test probe is connected to the relevant ports of the automotive parts to be tested, and the impedance values of the parts under different excitation conditions are measured. The impedance analyzer will automatically calculate and output parameters such as the real part, imaginary part, amplitude, and phase of the impedance, which are recorded as impedance data, and transmit the impedance data to the data storage unit for storage; The power consumption acquisition unit calculates the power consumption data of the automotive parts in real time based on the corresponding voltage and current data measured under the same excitation condition, and synchronously transmits the power consumption data to the data storage unit for storage; It should be further noted that the voltage data includes operating voltage, transient voltage, ripple voltage, and insulation voltage, etc.; the current data includes steady-state current, surge current, leakage current, and short-circuit current; the impedance data includes DC resistance, AC impedance, and insulation impedance; the power consumption data includes active power, reactive power, charge-discharge efficiency, and energy loss; The voltage data, current data, impedance data, and power consumption data are recorded as multi-dimensional electrical performance data.
[0026] It should be further noted that in the specific implementation process, the specific process of preprocessing the multi-dimensional electrical performance data to obtain multi-dimensional preprocessed electrical performance data includes: Obtain the multi-dimensional electrical performance data in the data storage unit; Based on a data processing tool, count the number and position of missing values in each column (such as columns of voltage, current, impedance, power consumption, etc.) of the multi-dimensional electrical performance data; smooth the data by calculating the average value within the data window to reduce the influence of noise; select a window of size n, and calculate the average value of the data within the window in turn, and use it as the new value of the data point at the center of the window; draw a box plot of the multi-dimensional electrical performance data. The range of the whiskers (i.e., the line segments at both ends of the box plot) in the box plot is usually 1.5 times the interquartile range (IQR, that is, the difference between the upper quartile Q3 and the lower quartile Q1). Data points outside this range are regarded as outliers, and based on data smoothing technology, the outliers are removed to make it more conform to the overall trend of the data; Based on Z-score standardization, standardize the multi-dimensional electrical performance data after outlier detection and processing, and record it as multi-dimensional preprocessed electrical performance data.
[0027] It should be further noted that in the specific implementation process, the specific process of signal conversion analysis for the multi-dimensional preprocessed electrical performance data includes: Each data point in the multi-dimensional preprocessed electrical performance data corresponds to a specific sampling moment, and then a time series is generated; Set the sampling frequency, determine the sampling interval, and based on the sampling interval, generate a corresponding time series for each data point in the multi-dimensional preprocessed electrical performance data; it should be further noted that the time series is like a time axis, providing a reference framework for subsequent analysis of the change of multi-dimensional preprocessed electrical performance data over time; Subsequently, obtain the time-domain waveform signal. Correlate each electrical performance data in the multi-dimensional preprocessed electrical performance data with the generated time series respectively, and draw a curve that changes over time; taking voltage as an example, the abscissa is time and the ordinate is the voltage value, and the drawn curve is the time-domain waveform signal of the electrical performance data of voltage; It should be further noted that according to the time-domain waveform signal, it is possible to intuitively understand the numerical values and change trends of electrical performance parameters at different times, such as whether there are periodic fluctuations, mutation points, etc.; Based on the Fast Fourier Transform (FFT) algorithm, process the time-domain waveform signals of the electrical performance data corresponding to each electrical performance data in the multi-dimensional preprocessed electrical performance data, and obtain a graph with frequency as the abscissa and amplitude as the ordinate. The drawn graph is the frequency-domain waveform signal of the electrical performance data; it should be further noted that the frequency-domain waveform signal of the electrical performance data can reveal the different frequency components and their relative intensities contained in the signal, helping us discover the periodic characteristics or interference components of specific frequencies hidden in the time-domain signal.
[0028] It should be further noted that in the specific implementation process, the specific process of constructing the multi-dimensional fault characterization space includes: Obtain multi-dimensional preprocessed electrical performance data, environmental data, time-domain waveform signals of electrical performance data, and frequency-domain waveform signals of electrical performance data; Perform vectorization processing on the multi-dimensional preprocessed electrical performance data and environmental data. The specific process is as follows: The environmental data includes temperature, humidity, etc.; Determine the order of each data in the vector according to certain rules; Arrange the multi-dimensional preprocessed electrical performance data and environmental data with the sorted order in sequence to form a one-dimensional vector; The one-dimensional vector can be expressed as , where represents temperature, represents humidity, represents voltage data, represents current data, represents impedance data, represents power consumption data; It should be further noted that if more data items are provided, they are added to the vector in sequence to form a feature vector containing all relevant information for subsequent feature extraction and analysis operations; Serialize the time-domain waveform signal and frequency-domain waveform signal of the electrical performance data. The specific process is as follows: Preset a fixed length. If the lengths of the time-domain waveform signal and frequency-domain waveform signal of the electrical performance data exceed the fixed length, truncation processing is required. Starting from the starting positions of the time-domain waveform signal and frequency-domain waveform signal of the electrical performance data, intercept a part with the fixed length as the final sequence to obtain the time-domain waveform signal sequence and frequency-domain waveform signal sequence of the electrical performance data. When the lengths of the time-domain waveform signal and frequency-domain waveform signal of the electrical performance data are less than the fixed length, padding operations are required. Add several 0s at the end of the signal to make its length reach the fixed length. For example, if the length of the time-domain waveform signal or frequency-domain waveform signal of the electrical performance data is 80 and the fixed length is 100, then add 20 0s after the signal. Synchronously input the one-dimensional vector, the time-domain waveform signal sequence, and the frequency-domain waveform signal sequence of the electrical performance data into the feature extraction network, construct a multi-dimensional fault characterization space, and generate multi-dimensional characteristic electrical performance data.
[0029] It should be further noted that in the specific implementation process, the specific process of constructing an electrical performance data threshold adjustment prediction model based on the transfer learning algorithm and multi-dimensional characteristic electrical performance data includes: Obtain multi-dimensional characteristic electrical performance data for several historical acquisition cycles. And preset standard electrical performance data. Group and label the multi-dimensional characteristic electrical performance data for several historical acquisition cycles, denoted as where is a natural number. Take groups of multi-dimensional characteristic electrical performance data for several historical acquisition cycles and the standard electrical performance data as sample data, and where is a natural number less than and use the sample data to obtain the sample data mean, denoted as the sample set. Take the remaining multi-dimensional characteristic electrical performance data for several historical acquisition cycles and the standard electrical performance data as the test set. According to the sample set and the test set, form a training sample set. Based on the convolutional neural network, construct a standard prediction model. It should be further noted that the standard prediction model includes a first-level standard prediction model and a second-level standard prediction model. Input the training sample set into the first-level standard prediction model, train the first-level standard prediction model, obtain the trained first-level standard prediction model, denote the trained first-level standard prediction model as the electrical performance data threshold adjustment prediction model, and obtain the dynamic adjustment determination threshold of the automotive parts under the current environmental conditions , the dynamic adjustment determination threshold is: ; where represents temperature, represents the ideal operating temperature of the automotive part, represents the proportionality coefficient for adjusting the influence degree of temperature, represents humidity, represents voltage data, represents current data, represents impedance data, represents power consumption data.
[0030] It should be further noted that in the specific implementation process, according to the dynamic adjustment determination threshold and the multi-dimensional preprocessed electrical performance data, an electrical performance evaluation prediction model is constructed, and the electrical performance operation ability value of the automotive part is obtained. Furthermore, the specific process of obtaining the test result of the automotive part includes: Preset the standard determination threshold ; According to the standard determination threshold , obtain the multi-dimensional preprocessed electrical performance data when the actual dynamic adjustment determination threshold is; Obtain several groups of multi-dimensional preprocessed electrical performance data; Based on the transfer learning algorithm, retain the underlying feature extraction layer of the electrical performance data threshold adjustment prediction model, and use the several groups of multi-dimensional preprocessed electrical performance data and the underlying feature extraction layer as the input layer, input them into the second-level standard prediction model for training, obtain the trained second-level standard prediction model, denote the trained second-level standard prediction model as the electrical performance evaluation prediction model, and obtain the electrical performance operation ability value of the automotive part , the electrical performance operation ability value is: where represents the best operating temperature of the automotive part, represents the temperature range parameter, represents the maximum value that the environmental humidity may reach, represents the rated power consumption of the automotive part, , are weight coefficients; Preset the standard electrical performance operation ability value ; If the electrical performance operation ability value , then the electrical performance operation ability value of the automotive part meets the standard; If the electrical performance operation ability value , then the electrical performance operation ability value of the automotive part does not meet the standard; As Figure 2 shown, an electrical performance test system for automotive parts includes: an electrical performance data acquisition module, an electrical performance data processing module, an electrical performance data analysis module, and an intelligent test module; The electrical performance data acquisition module is used to obtain multi-dimensional electrical performance data of the automotive part under the test environment; The electrical performance data processing module is used to preprocess the multi-dimensional electrical performance data to obtain multi-dimensional preprocessed electrical performance data; The electrical performance data analysis module is used to perform signal conversion analysis on the multi-dimensional preprocessed electrical performance data to obtain the time-domain waveform signal of the electrical performance data and the frequency-domain waveform signal of the electrical performance data; The intelligent test module is used to construct a multi-dimensional fault characterization space, generate multi-dimensional characteristic electrical performance data, and construct an electrical performance data threshold adjustment prediction model to obtain the dynamic adjustment decision threshold of the automotive part under the current environmental conditions. According to the dynamic adjustment decision threshold and the multi-dimensional preprocessed electrical performance data, an electrical performance evaluation prediction model is constructed, and the electrical performance operation ability value of the automotive part is obtained.
[0031] Embodiments disclosed in the present invention, the processes described above with reference to the flowcharts can be implemented as computer software programs. Embodiments disclosed in the present invention include a computer program product, which includes a computer program carried on a computer-readable medium, and the computer program contains program codes for executing the methods shown in the flowcharts. In such an embodiment, the computer program can be downloaded and installed from a network through a communication part, and / or installed from a removable medium. When the computer program is executed by a central processing unit (CPU), the above functions defined in the methods of the present application are executed. It should be noted that the computer-readable medium in the present application can be a computer-readable signal medium or a computer-readable storage medium or any combination of the two. A computer-readable storage medium can be, for example, but not limited to, an electrical, magnetic, optical, electromagnetic, infrared, or semiconductor system, apparatus, or device, or any combination of the above. More specific examples of a computer-readable storage medium can include, but are not limited to: an electrical connection with one or more wire segments, a portable computer disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the above. In the present application, a computer-readable storage medium can be any tangible medium that contains or stores a program, and the program can be used by or combined with an instruction execution system, apparatus, or device. In the present application, a computer-readable signal medium can include a data signal propagated in a baseband or as part of a carrier wave, which carries computer-readable program codes. Such a propagated data signal can take various forms, including but not limited to electromagnetic signals, optical signals, or any suitable combination of the above. A computer-readable signal medium can also be any computer-readable medium other than a computer-readable storage medium, and the computer-readable medium can send, propagate, or transmit a program for use by or combined with an instruction execution system, apparatus, or device. The program codes contained on the computer-readable medium can be transmitted by any appropriate medium, including but not limited to: wireless, wire segments, optical cables, RF, etc., or any suitable combination of the above.
[0032] The flowcharts and block diagrams in the accompanying drawings illustrate the possible architectures, functions, and operations of systems, methods, and computer program products according to various embodiments of the present invention. In this regard, each block in the flowchart or block diagram may represent a module, a segment of a program, or a part of code, which contains one or more executable instructions for implementing a specified logical function. It should also be noted that in some alternative implementations, the functions marked in the blocks may occur in a different order than that marked in the accompanying drawings. For example, two consecutive blocks shown may actually be executed substantially in parallel, and they may sometimes be executed in the reverse order, depending on the functions involved. It should also be noted that each block in the block diagram and / or flowchart, and combinations of blocks in the block diagram and / or flowchart, can be implemented by a dedicated hardware-based system that performs the specified functions or operations, or can be implemented by a combination of dedicated hardware and computer instructions.
[0033] Those skilled in the art should understand that the embodiments of the present invention described above and shown in the accompanying drawings are only examples and do not limit the present invention. The objectives of the present invention have been fully and effectively achieved. The functions and structural principles of the present invention have been demonstrated and explained in the embodiments. Without departing from the said principles, the embodiments of the present invention may have any variations or modifications.
Claims
1. An electrical performance testing method for automotive parts, characterized in that, Including: Obtain multi-dimensional electrical performance data of automotive parts in a test environment, and perform preprocessing to obtain multi-dimensional preprocessed electrical performance data. Conduct signal conversion analysis on the multi-dimensional preprocessed electrical performance data to obtain the time-domain waveform signal of the electrical performance data and the frequency-domain waveform signal of the electrical performance data; Synchronously input the multi-dimensional preprocessed electrical performance data, environmental data, the time-domain waveform signal of the electrical performance data, and the frequency-domain waveform signal of the electrical performance data into a feature extraction network to construct a multi-dimensional fault characterization space and generate multi-dimensional characteristic electrical performance data; Based on the transfer learning algorithm and the multi-dimensional characteristic electrical performance data, construct an electrical performance data threshold adjustment prediction model to obtain the dynamic adjustment decision threshold of automotive parts under the current environmental conditions; According to the dynamic adjustment decision threshold and the multi-dimensional preprocessed electrical performance data, construct an electrical performance evaluation prediction model, obtain the electrical performance operation ability value of automotive parts, and further obtain the test results of automotive parts.
2. The electrical performance testing method for an automotive part according to claim 1, wherein The process of obtaining multi-dimensional electrical performance data of automotive parts in a test environment includes: Set up a multi-channel data acquisition device, including a data storage unit, and set an acquisition period, where the acquisition period contains several acquisition moments; collect voltage data, current data, impedance data, and power consumption data according to the acquisition moments and transmit them to the data storage unit for storage, denoted as multi-dimensional electrical performance data.
3. The method for testing the electrical performance of an automotive part according to claim 2, characterized in that, The process of preprocessing multi-dimensional electrical performance data includes: Obtain the multi-dimensional electrical performance data in the data storage unit; based on a data processing tool, calculate the average value within the calculation data window to smooth the data, draw a box plot of the multi-dimensional electrical performance data, and perform preprocessing on the multi-dimensional electrical performance data based on data smoothing technology and Z-score standardization, denoted as multi-dimensional preprocessed electrical performance data.
4. The electrical performance testing method for an automotive part according to claim 3, characterized in that The process of signal conversion analysis on multi-dimensional preprocessed electrical performance data includes: Each data point in the multi-dimensional preprocessed electrical performance data corresponds to a specific sampling moment, and then a time series is generated; Set the sampling frequency, determine the sampling interval, and generate a corresponding time series for each data point in the multi-dimensional preprocessed electrical performance data; Correspond each electrical performance data in the multi-dimensional preprocessed electrical performance data with the generated time series respectively, and draw a curve that changes with time, which is the time-domain waveform signal of the electrical performance data; Based on the fast Fourier transform algorithm, process the time-domain waveform signal of the electrical performance data corresponding to each electrical performance data in the multi-dimensional preprocessed electrical performance data, which is the frequency-domain waveform signal of the electrical performance data.
5. A method for testing the electrical performance of an automotive part according to claim 4, characterized in that, The process of constructing a multi-dimensional fault characterization space includes: Obtain environmental data; perform vectorization processing on the multi-dimensional preprocessed electrical performance data and environmental data. The process is as follows: The environmental data includes temperature and humidity; Arrange the multi-dimensional preprocessed electrical performance data and environmental data in sequence to form a one-dimensional vector; The one-dimensional vector can be expressed as , where represents temperature, represents humidity, represents voltage data, represents current data, represents impedance data, represents power consumption data; The process of serializing the time-domain waveform signal of the electrical performance data and the frequency-domain waveform signal of the electrical performance data is as follows: Preset and determine a fixed length; If the lengths of the time-domain waveform signal of the electrical performance data and the frequency-domain waveform signal of the electrical performance data exceed the fixed length, perform truncation processing to obtain the time-domain waveform signal sequence of the electrical performance data and the frequency-domain waveform signal sequence of the electrical performance data. Synchronously input the one-dimensional vector, the time-domain waveform signal sequence of electrical performance data, and the frequency-domain waveform signal sequence of electrical performance data into the feature extraction network, construct a multi-dimensional fault characterization space, and generate multi-dimensional characteristic electrical performance data.
6. The electrical performance testing method for an automotive part according to claim 5, wherein, The process of constructing a prediction model for adjusting the threshold of electrical performance data includes: Obtain multi-dimensional characteristic electrical performance data for a number of historical acquisition cycles; And preset standard electrical performance data; Group and label the multi-dimensional characteristic electrical performance data of several groups of historical acquisition cycles, denoted as is a natural number; Put A number of groups of multi-dimensional characteristic electrical performance data and standard electrical performance data of historical acquisition cycles are used as sample data, and is a natural number less than Using the sample data, the mean of the sample data is obtained and denoted as the sample set; Use the multi-dimensional characteristic electrical performance data for the remaining several historical acquisition cycles and the standard electrical performance data as the test set; According to the sample set and the test set, form a training sample set; Based on the convolutional neural network, construct a standard prediction model; The standard prediction model includes a first-level standard prediction model and a second-level standard prediction model; And input the training sample set into the first-level standard prediction model, train the first-level standard prediction model, obtain the trained first-level standard prediction model, denote the trained first-level standard prediction model as the electrical performance data threshold adjustment prediction model, and obtain the dynamic adjustment determination threshold of the automotive parts under the current environmental conditions , the dynamic adjustment determination threshold is:[[]] ; Among them, represents temperature, represents the ideal operating temperature of the automotive part, represents the proportionality coefficient for adjusting the degree of temperature influence, represents humidity, represents voltage data, represents current data, represents impedance data, represents power consumption data.
7. A method for testing the electrical performance of an automotive part according to claim 6, characterized in that, The process of constructing an electrical performance evaluation prediction model includes: Preset standard determination threshold ; According to the standard determination threshold , obtain the actual dynamically adjusted determination threshold when the multi-dimensional preprocessed electrical performance data; Obtain a number of multi-dimensional preprocessed electrical performance data; Based on the transfer learning algorithm, retain the underlying feature extraction layer of the electrical performance data threshold adjustment prediction model, and use the several groups of multi-dimensional preprocessed electrical performance data and the underlying feature extraction layer as the input layer, and input them into the secondary standard prediction model for training to obtain the trained secondary standard prediction model, and record the trained secondary standard prediction model as the electrical performance evaluation prediction model, and obtain the electrical performance operation ability value of the automotive parts , the electrical performance operation ability value is: Among them, represents the optimal operating temperature of the automotive parts, represents the temperature range parameter, represents the maximum value that the environmental humidity may reach, represents the rated power consumption of the automotive parts, and is the weight coefficient.
8. A method for testing the electrical performance of an automotive part according to claim 7, characterized in that, The process of obtaining the test results of automotive parts includes: presetting the standard electrical performance operation ability value ; If the electrical performance operation ability value , then the electrical performance operation ability value of the automotive parts meets the standard; If the electrical performance operation ability value , the electrical performance operation ability value of the automotive parts does not meet the standard.
9. An electrical performance testing system for automotive parts, characterized in that, The system executes a method for testing the electrical performance of automotive parts according to any one of claims 1-8 above, including: an electrical performance data acquisition module, an electrical performance data processing module, an electrical performance data analysis module, and an intelligent testing module; The electrical performance data acquisition module is used to obtain multi-dimensional electrical performance data of automotive parts in a test environment; The electrical performance data processing module is used to preprocess the multi-dimensional electrical performance data to obtain multi-dimensional preprocessed electrical performance data; The electrical performance data analysis module is used to perform signal conversion analysis on the multi-dimensional preprocessed electrical performance data to obtain a time-domain waveform signal of electrical performance data and a frequency-domain waveform signal of electrical performance data; The intelligent testing module is used to construct a multi-dimensional fault characterization space, generate multi-dimensional characteristic electrical performance data, construct a prediction model for adjusting the threshold of electrical performance data, obtain a dynamically adjusted determination threshold of automotive parts under the current environmental conditions, and construct an electrical performance evaluation prediction model according to the dynamically adjusted determination threshold and the multi-dimensional preprocessed electrical performance data, and obtain the electrical performance operation ability value of automotive parts.
10. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores a computer program, and the computer program can be executed by a processor to implement a method for testing the electrical performance of automotive parts according to any one of claims 1-8 above.