Multi-station FCT test system-oriented multi-task parallel test method

Abstract

The invention discloses a multi-task parallel test method for a multi-station FCT test system, and relates to the technical field of parallel tests.The method comprises the steps that a plurality of target PCBA products are loaded to corresponding stations on a needle bed respectively, and starting signals are obtained; scanning the target PCBA product to obtain a plurality of pieces of production work order information and a plurality of test programs; executing a plurality of test programs to obtain a plurality of test signal sets; carrying out test qualification authentication; load working signals are collected; and judging whether the plurality of load working signal feature sets meet a preset requirement or not, and displaying a qualified station of the product test through a machine indicator lamp. The technical problems that PCBA product testing efficiency is low, accuracy is difficult to guarantee and product quality control is not comprehensive in the prior art are solved, and the technical effects that the efficient process of conducting product testing at multiple stations at the same time is achieved, the testing efficiency and accuracy are improved, and the product quality is effectively guaranteed are achieved.

Description

Technical Field

[0001] The present invention relates to the technical field of parallel testing, and in particular to a multi-task parallel testing method for a multi-station FCT testing system. Background Art

[0002] In the field of electronic manufacturing, with the increasing complexity of electronic products and the continuous expansion of production scale, the testing of PCBA (printed circuit board assembly) products faces many challenges. Traditional testing methods often adopt serial testing mode, that is, testing products one by one. This method is extremely inefficient in the face of large-scale production and cannot meet the needs of fast production rhythm. At the same time, traditional testing lacks accuracy, and it is difficult to conduct comprehensive and accurate testing for some complex electrical performance and functional characteristics. For example, when testing parameters such as current, voltage, power and frequency of PCBA products, due to the lack of effective data processing and analysis methods, some subtle but critical performance anomalies may be missed. In addition, in terms of product quality control, traditional testing methods cannot achieve automation and intelligent management of product testing processes. In the process of loading products, it is difficult to ensure the accurate placement of products on the workstation, which can easily lead to inaccurate test results. Moreover, for the data collected during the test, there is a lack of effective feature recognition and analysis capabilities, and it is impossible to deeply explore the product quality information behind the data, and it is impossible to discover potential quality risks in time.

[0003] The existing technology has the following technical problems: low PCBA product testing efficiency, difficulty in ensuring accuracy, and incomplete product quality control. Summary of the invention

[0004] The present application provides a multi-task parallel testing method for a multi-station FCT testing system, which is used to solve the technical problems in the prior art of low PCBA product testing efficiency, difficulty in ensuring accuracy, and incomplete product quality control.

[0005] In view of the above problems, the present application provides a multi-task parallel testing method for a multi-station FCT testing system, the method comprising:

[0006] Load multiple target PCBA products to the corresponding stations on the needle bed respectively, and obtain a start signal after closing the cover; use a barcode scanner to scan the multiple target PCBA products on the multiple stations to obtain multiple production work order information; call the burner to burn the test program for the multiple target PCBA products according to the multiple production work order information to obtain multiple test programs, wherein each test program corresponds to a target PCBA product; obtain test instructions to execute the multiple test programs on the multiple target PCBA products, and use a data acquisition card to traverse the test signals of the multiple stations to obtain multiple test signal sets; transmit the multiple test signal sets to the industrial computer for test qualification certification to obtain a first test qualified target PCBA product set; traverse the first test qualified target PCBA product set to collect load working signals, use the industrial computer to perform feature recognition on the multiple load working signal sets, and obtain multiple load working signal feature sets; determine whether the multiple load working signal feature sets meet the preset requirements, and if so, obtain a second test qualified target PCBA product set; upload the second test qualified target PCBA product set to MES, and display the stations where the product tests are qualified through the machine indicator light.

[0007] One or more technical solutions provided in this application have at least the following technical effects or advantages:

[0008] Load multiple target PCBA products to the corresponding stations on the needle bed respectively, and obtain a start signal after closing the cover; scan multiple target PCBA products on multiple stations to obtain multiple production work order information; call the burner to burn the test program for the multiple target PCBA products to obtain multiple test programs; obtain the test instructions to execute the multiple test programs for the multiple target PCBA products, and use the data acquisition card to traverse the test signals of the multiple stations to obtain multiple test signal sets; transmit the multiple test signal sets to the industrial computer for test qualification certification to obtain the first test qualified target PCBA product set; traverse the first test qualified target PCBA product set to collect load working signals to obtain multiple load working signal feature sets; determine whether the multiple load working signal feature sets meet the preset requirements, and if so, obtain the second test qualified target PCBA product set; upload the second test qualified target PCBA product set to MES, and display the station where the product test is qualified through the machine indicator light. The efficient process of realizing product testing at multiple stations at the same time is achieved, the test efficiency and accuracy are improved, and the technical effect of effectively ensuring product quality is achieved. BRIEF DESCRIPTION OF THE DRAWINGS

[0009] In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the following briefly introduces the drawings required for use in the description of the embodiments. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work.

[0010] Figure 1 A flowchart of a multi-task parallel testing method for a multi-station FCT testing system provided in an embodiment of the present application;

[0011] Figure 2 A schematic diagram of the process of obtaining a start signal for a multi-task parallel testing method for a multi-station FCT testing system provided in an embodiment of the present application. DETAILED DESCRIPTION

[0012] This application provides a multi-task parallel testing method for a multi-station FCT testing system, which is used to solve the technical problems in the prior art of low PCBA product testing efficiency, difficulty in ensuring accuracy, and incomplete product quality control.

[0013] The technical solutions in the embodiments of the present application will be described clearly and completely below in conjunction with the drawings in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, not all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of this application.

[0014] Examples, such as Figure 1 As shown, the present application provides a multi-task parallel testing method for a multi-station FCT test system, the method comprising:

[0015] Step S100: Load multiple target PCBA products to corresponding stations on the needle bed respectively, and obtain a start signal after closing the cover.

[0016] Specifically, when the multi-station FCT test system is started, multiple target PCBA products are placed one by one in the corresponding pre-set stations on the needle bed to ensure that each PCBA product is accurately connected to the electrical connection point of the needle bed, providing a reliable physical connection basis for the subsequent test signal acquisition. This process must strictly follow the operating specifications to ensure the accuracy of the product placement position to avoid problems such as poor contact affecting the test results. After all target PCBA products are loaded, close the cover on the needle bed. The closing action of the cover triggers a signal feedback mechanism, which sends a start signal to the test system through the synergy of mechanical structure and electrical sensor. When the cover is closed, it will trigger the sensor installed at a specific position. The sensor converts the mechanical signal into an electrical signal and transmits the electrical signal to the control unit of the test system through the pre-laid signal transmission line. After receiving the electrical signal, the control unit recognizes it as a start signal, and then starts the subsequent steps of the entire test process, including calling the barcode scanner to scan product information and other operations, which provides the necessary trigger conditions for the subsequent orderly multi-station parallel testing, ensures that the entire test system works according to the predetermined process and timing, and closely cooperates with the overall architecture and working logic of the multi-station FCT test system, laying a solid foundation for the efficient and accurate functional testing of products such as vehicle-mounted pump controller circuit boards.

[0017] Step S200: Scan multiple target PCBA products on multiple workstations using a barcode scanner to obtain multiple production work order information.

[0018] Specifically, after receiving the start signal, this step is entered, and the barcode scanner is used to scan multiple target PCBA products placed on multiple workstations. The barcode scanner uses its built-in optical scanning technology to emit light of a specific frequency and receive reflected light. By processing and analyzing the reflected light signal, it identifies the identification information such as the QR code or barcode attached to the surface of the product. These identification information contains rich production work order information, such as product model, batch number, production process parameters, etc. The barcode scanner transmits the scanned information to the main control unit of the test system through wired or wireless communication. The main control unit decodes and classifies this information to obtain multiple production work order information corresponding to each target PCBA product. This step provides a key basis for the subsequent accurate test program burning and personalized product testing based on the work order information, ensuring that the entire test process is closely centered on the actual production needs and specification requirements of the product, and ensuring the accuracy and effectiveness of the functional testing of products such as the vehicle-mounted pump controller circuit board.

[0019] Step S300: calling a burner to burn test programs for the multiple target PCBA products according to the multiple production work order information to obtain multiple test programs, wherein each test program corresponds to a target PCBA product.

[0020] Specifically, during the operation of the multi-station FCT test system, after obtaining multiple production work order information, the test system accurately calls the burner to work based on this information. The burner first parses the key data such as the specific configuration requirements, functional specifications and corresponding test logic of the target PCBA product in each production work order information. Then, the burner extracts the corresponding test program code fragments from its internal storage or connected external storage device, and writes these specific test program codes accurately into the storage unit of the corresponding target PCBA product, such as flash memory chip, according to the product identification in the work order information. In this way, a unique test program is generated for each target PCBA product to ensure that the test program is highly matched with the actual needs of the product. This process fully reflects the ability to perform differentiated processing based on the production work order information of different products, so that each product can perform comprehensive and accurate functional verification based on its own characteristics during the test process.

[0021] Step S400: Acquire test instructions to execute the multiple test programs on the multiple target PCBA products, and use a data acquisition card to traverse the test signals of the multiple workstations to obtain multiple test signal sets.

[0022] Specifically, after the test program is burned, the test instruction is obtained, which triggers multiple target PCBA products to start their corresponding test programs at the same time. During the running of the test program, the various functional modules inside the target PCBA product work according to the preset logic, generating various electrical signal changes, such as voltage fluctuations, current changes, signal frequency changes, etc. At this time, the data acquisition card begins to play a role. It traverses and collects the test signals of the target PCBA products on multiple stations in turn with a specific sampling frequency and accuracy through the signal acquisition circuit pre-connected to each station. The data acquisition card converts the collected analog signals into digital signals, and classifies and stores them according to the time series and station number, and finally obtains multiple test signal sets. Each test signal set completely records the signal changes of the corresponding target PCBA product during the test process. These signal sets provide a rich data basis for the subsequent accurate judgment of whether the product is qualified, and are an important link in realizing accurate functional testing of products such as vehicle-mounted pump controller circuit boards.

[0023] Step S500: transmitting the plurality of test signal sets to an industrial computer for test qualification certification to obtain a first test qualified target PCBA product set.

[0024] Specifically, when the data acquisition card completes the test signal acquisition of multiple stations and obtains multiple test signal sets, these sets will be stably and efficiently transmitted to the industrial computer through high-speed data transmission lines (such as Ethernet, USB, etc.). As the core data processing unit of the entire test system, the industrial computer has a set of complex and accurate test qualification certification algorithms pre-built in.

[0025] First, for each test signal set, the industrial computer performs detailed iterative correlation identification for key parameters such as current, voltage, power, and frequency. Through multiple iterative calculations, the intrinsic correlation between these parameters is deeply explored, and the key data that can represent the performance characteristics of the product are accurately extracted. For example, when analyzing current parameters, a multi-step iterative correlation analysis is performed based on a specific mathematical model to obtain the most accurate current characteristic information. Next, the industrial computer performs deviation analysis on the extracted characteristic data, compares it with the pre-set standard parameter range, and calculates the degree of deviation of each parameter. Then, using the weighted analysis method, the corresponding weights are assigned according to the importance of different parameters to the overall performance of the product, and the comprehensive deviation corresponding to each test signal set is comprehensively calculated. Finally, the industrial computer determines whether each target PCBA product is qualified based on the comprehensive deviation. If the comprehensive deviation is within the allowable error range, the product is judged to be qualified and included in the first test qualified target PCBA product set; otherwise, it is judged to be unqualified. This process strictly follows the test qualification criteria in the document briefing, ensuring that only products such as the on-board pump controller circuit board that truly meet the quality requirements can enter the subsequent load working signal acquisition and further testing stages, effectively ensuring the accuracy and reliability of product testing.

[0026] Step S600: traverse the first test qualified target PCBA product set to collect load working signals, use an industrial computer to perform feature recognition on multiple load working signal sets, and obtain multiple load working signal feature sets.

[0027] Specifically, after obtaining the first test qualified target PCBA product set, start traversing each product in the set. For each product, by simulating the actual working load conditions, it is placed in a load working state. At this time, a high-precision data acquisition device (such as a specific sensor, etc.) is used to comprehensively collect its load working signals, including voltage, current, temperature and other signal types, so as to obtain multiple load working signal sets, each set corresponding to a first test qualified target PCBA product. Then, the industrial computer uses its powerful computing power and pre-trained algorithm model to perform feature recognition on these load working signal sets. First, the industrial computer calculates the mean, variance and other statistics of various signal parameters in each load working signal set to preliminarily summarize the characteristics of the signal's central tendency and discrete degree. Then, these processed signal data are input into the trained feature recognition network layer, which can automatically identify the deep-level features hidden in the signal through a deep learning algorithm, such as the periodicity, change trend, frequency characteristics, etc. of the signal, which can more accurately reflect the performance of the product under the load working state. Finally, after processing by the feature recognition network layer, multiple load working signal feature sets are obtained. These feature sets will provide more accurate and representative data basis for subsequent judgment of whether the product is finally qualified, and help to more accurately screen out high-quality products such as vehicle-mounted pump controller circuit boards.

[0028] Step S700: Determine whether the plurality of load working signal feature sets meet preset requirements, and if so, obtain a second test qualified target PCBA product set.

[0029] Specifically, after obtaining multiple load working signal feature sets, they are judged one by one according to the strict standards and requirements pre-set in the industrial computer. These preset requirements are determined based on the design specifications, performance indicators and actual application scenarios of products such as vehicle-mounted pump controller circuit boards, covering all key aspects of signal characteristics, such as signal stability, accuracy, consistency, etc. For each load working signal feature set, the characteristic parameters therein are carefully compared with the corresponding thresholds and ranges in the preset requirements. For example, check whether the frequency of the signal is within the specified frequency band, whether the change in the signal amplitude meets the expected tolerance range, and whether the phase relationship of the signal meets the design requirements. At the same time, the relationship between multiple characteristic parameters will be comprehensively evaluated to ensure that they conform to the normal working logic of the product as a whole. If all the characteristic parameters in a load working signal feature set meet the preset requirements, the corresponding target PCBA product will be judged as finally qualified and thus included in the second test qualified target PCBA product set. On the contrary, if any characteristic parameter does not meet the preset requirements, the product will be judged as unqualified, ensuring that only products whose performance fully meets the standards under load working conditions can pass the test, effectively guaranteeing the reliability and stability of the product in actual use, and providing solid quality assurance for the product to be put on the market or enter the next production link.

[0030] Step S800: Upload the second test qualified target PCBA product set to MES, and display the workstations where the products have passed the test through the machine indicator lights.

[0031] Specifically, in the final stage of the multi-station FCT test system, after determining the second test qualified target PCBA product set, the test system accurately uploads the product information in the set to MES (manufacturing execution system) through communication interfaces such as Ethernet. As one of the core systems of enterprise production management, MES can receive and record the test qualified data of these products in real time, including detailed information such as product number, test time, and test station, thereby realizing the seamless connection between the test results and the production management process, and facilitating the enterprise to trace the product quality throughout the process and the refined management of the production process. At the same time, the machine indicator light plays its intuitive indication role. For each station, when the target PCBA product on the corresponding station is judged to be qualified and included in the second test qualified product set, the machine indicator light will light up a specific color (such as green) to clearly show the station where the product has passed the test. This visual feedback mechanism enables on-site operators to quickly and intuitively understand the test status of the products at each station, and directly know which stations have successfully passed the test without viewing complex system reports. Operators can perform follow-up operations in a timely manner according to the status of the indicator lights, such as removing qualified products from the needle bed for packaging or entering the next production link, and quickly locating unqualified workstations and conducting problem troubleshooting and product processing. The entire process is highly consistent with the requirements for effective management of test results and convenient on-site operations in the document disclosure, ensuring that the multi-station FCT test system can achieve efficient and accurate product quality control while improving production efficiency.

[0032] In a possible implementation, step S400 further includes:

[0033] Step S410: The data acquisition card performs long-term monitoring on the test signals of the multiple workstations to obtain multiple initial test signal sets, wherein each initial test signal set includes an initial current long-term monitoring parameter sequence, an initial voltage long-term monitoring parameter sequence, an initial power long-term monitoring parameter sequence, and an initial frequency long-term monitoring parameter sequence.

[0034] Step S420: Use the interpolation method to traverse the multiple initial test signal sets to process missing data values, and obtain multiple test signal sets, wherein each test signal set includes a current long-time monitoring parameter sequence, a voltage long-time monitoring parameter sequence, a power long-time monitoring parameter sequence, and a frequency long-time monitoring parameter sequence.

[0035] Specifically, as the core device for signal acquisition, the data acquisition card conducts long-term monitoring of the test signals of the target PCBA products on multiple stations. It continuously samples the signals of each station at precise time intervals. In this process, for each station, the data acquisition card will build an initial current long-term monitoring parameter sequence, which records in detail the continuous change of the current of the product at this station over time during the long-term test, and provides key data for the subsequent analysis of the energy consumption characteristics of the product; at the same time, it builds an initial voltage long-term monitoring parameter sequence to reflect the dynamic fluctuation of the voltage, which helps to judge the stability of the circuit; the initial power long-term monitoring parameter sequence comprehensively reflects the power consumption state of the product at different times, which is crucial for evaluating the performance efficiency of the product; there is also an initial frequency long-term monitoring parameter sequence, and the recorded frequency change information can be used to analyze the operation of the frequency-related functions in the product. Through such comprehensive and detailed long-term monitoring, multiple initial test signal sets are finally obtained, each of which contains rich initial signal feature information of the corresponding station product, laying a solid foundation for subsequent further data processing and product test certification.

[0036] In the signal processing process of the multi-station FCT test system, after obtaining multiple initial test signal sets containing initial current, voltage, power and frequency long-time monitoring parameter sequences, there may be data missing due to the complex actual acquisition environment. At this time, the interpolation method begins to play a role. The interpolation method traverses each initial test signal set in a mathematical calculation manner. For the missing data values ​​in the initial current long-time monitoring parameter sequence, it will use a specific interpolation algorithm (such as linear interpolation, polynomial interpolation, etc.) to calculate based on the numerical change trend of adjacent data points in the sequence and the distribution law of the overall data, so as to fill in the missing data points and form a complete current long-time monitoring parameter sequence. Similarly, for the initial voltage long-time monitoring parameter sequence, the initial power long-time monitoring parameter sequence and the initial frequency long-time monitoring parameter sequence, the interpolation method will also process the missing values ​​according to their respective data characteristics and laws. After such a meticulous processing process, multiple initial test signal sets with missing data were converted into multiple complete test signal sets, each of which contains accurate current long-time monitoring parameter sequences, voltage long-time monitoring parameter sequences, power long-time monitoring parameter sequences, and frequency long-time monitoring parameter sequences. These complete test signal sets provide a solid and reliable data foundation for subsequent precise test and certification in industrial computers, ensuring the accuracy and reliability of the test results, and helping to more accurately determine whether the functions of products such as vehicle-mounted pump controller circuit boards meet the requirements, thereby ensuring the effectiveness of the entire test process and the stability of product quality.

[0037] In a possible implementation, step S500 further includes:

[0038] Step S510: extracting a first test signal set from the multiple test signal sets.

[0039] Step S520: In the data processing unit of the industrial computer, iteratively associate and identify the first current long-time monitoring parameter sequence, the first voltage long-time monitoring parameter sequence, the first power long-time monitoring parameter sequence and the first frequency long-time monitoring parameter sequence of the first test signal set to determine the first iterative associated current monitoring parameters, the first iterative associated voltage monitoring parameters, the first iterative associated power monitoring parameters and the first iterative associated frequency monitoring parameters.

[0040] Step S530: Perform deviation analysis on the first iteration-associated current monitoring parameter, the first iteration-associated voltage monitoring parameter, the first iteration-associated power monitoring parameter and the first iteration-associated frequency monitoring parameter from the corresponding standard values ​​to obtain the first iteration-associated current monitoring parameter deviation, the first iteration-associated voltage monitoring parameter deviation, the first iteration-associated power monitoring parameter deviation and the first iteration-associated frequency monitoring parameter deviation.

[0041] Step S540: performing a weighted analysis on the first iteration-associated current monitoring parameter deviation, the first iteration-associated voltage monitoring parameter deviation, the first iteration-associated power monitoring parameter deviation, and the first iteration-associated frequency monitoring parameter deviation to obtain a first comprehensive deviation.

[0042] Step S550: determine whether the first comprehensive deviation meets the preset requirement. If so, the test result of the first target PCBA product at the workstation corresponding to the first test signal set is qualified.

[0043] Step S560: Utilize the data processing unit to perform test qualification authentication on the plurality of test signal sets, and add the plurality of target PCBA products that have passed the test into the first set of target PCBA products that have passed the test.

[0044] Specifically, it is the beginning of a detailed analysis of a single product. At this point, among the multiple test signal sets that have been acquired, one of the sets is accurately selected as the first test signal set according to a pre-set rule or order. This selection process may be based on factors such as the station number, product number, or test time to ensure that each test signal set can be extracted in an orderly manner for separate evaluation. For example, if the extraction is performed in the order of the station number, the first selected test signal set is the set of all test signal data generated by the product corresponding to the first station during the test process, including long time-series monitoring parameter sequences such as current, voltage, power, and frequency. This step provides a clear object for subsequent in-depth iterative association identification, deviation analysis, and comprehensive deviation calculation operations on a single product, and is a basic step for accurate judgment of each target PCBA product during the entire test qualification certification process.

[0045] After the first test signal set is extracted, the data processing unit of the industrial computer begins to iteratively associate and identify the first current long-time monitoring parameter sequence, the first voltage long-time monitoring parameter sequence, the first power long-time monitoring parameter sequence, and the first frequency long-time monitoring parameter sequence it contains. The data processing unit uses a neural network model to first analyze the first current long-time monitoring parameter sequence, taking into account factors such as the trend of current changes over time and the fluctuation amplitude. At the same time, combined with the first voltage long-time monitoring parameter sequence, because there is a close physical relationship between current and voltage, through multiple iterative calculations, the influence of voltage changes on current is explored, thereby determining a first iterative associated current monitoring parameter that can more accurately reflect the current characteristics of the product under actual working conditions. Then, for the first voltage long-time monitoring parameter sequence, the information of other related parameter sequences such as current and power is also integrated, and the stability of voltage under different working conditions, the law of coordinated changes with other parameters, etc. are analyzed during the iteration process, and then the first iterative associated voltage monitoring parameter is determined, which can more accurately reflect the performance characteristics of the product in terms of voltage. For the first power long-time monitoring parameter sequence, since power is closely related to current and voltage, the data processing unit fully considers the impact of changes in current and voltage on power, as well as the changing trend of power itself over time, in the iterative calculation, and determines the first iterative associated power monitoring parameters through complex correlation analysis, so that it can better represent the power performance of the product. Finally, when analyzing the first frequency long-time monitoring parameter sequence, combined with the product circuit characteristics and changes in other parameters, after multiple iterative correlation calculations, the first iterative associated frequency monitoring parameters are determined, which can accurately reflect the product's operating characteristics in terms of frequency. Through such a comprehensive and in-depth iterative correlation identification process, a more representative and accurate parameter basis is provided for the subsequent accurate judgment of whether the product is qualified, which helps to improve the reliability of the test results of products such as vehicle-mounted pump controller circuit boards.

[0046] After the first iteration-associated current monitoring parameters, the first iteration-associated voltage monitoring parameters, the first iteration-associated power monitoring parameters, and the first iteration-associated frequency monitoring parameters are determined, these parameters are analyzed for deviation from the pre-set ideal standard values ​​corresponding to products such as vehicle-mounted pump controller circuit boards. For the first iteration-associated current monitoring parameters, the difference between it and the standard current value is calculated. This difference reflects the deviation between the actual current characteristics of the product and the standard requirements. The difference is normalized within a certain range to obtain the deviation of the first iteration-associated current monitoring parameters. Similarly, for the first iteration-associated voltage monitoring parameters, the deviation from the standard voltage value is calculated and converted into the deviation of the first iteration-associated voltage monitoring parameters. This deviation can intuitively display the degree of deviation of the voltage parameters. For power and frequency, the first iteration-associated power monitoring parameters are compared with the standard power value to calculate the deviation, and the first iteration-associated power monitoring parameter deviation is obtained; the first iteration-associated frequency monitoring parameters are compared with the standard frequency value to calculate the deviation of the first iteration-associated frequency monitoring parameters. These deviation values ​​accurately quantify the differences between the product and the standard values ​​in various key parameters, providing basic data for subsequent weighted analysis, helping to accurately determine whether the product meets quality requirements, thereby ensuring the accuracy and reliability of product performance evaluation throughout the entire testing process.

[0047] After obtaining the deviation of the first iteration-related current monitoring parameter, the deviation of the first iteration-related voltage monitoring parameter, the deviation of the first iteration-related power monitoring parameter, and the deviation of the first iteration-related frequency monitoring parameter, these deviations are weighted and analyzed according to a predetermined weight allocation scheme. Since in products such as the vehicle-mounted pump controller circuit board, the degree of influence of parameters such as current, voltage, power, and frequency on the overall performance of the product varies, for example, the current may have a greater impact on the energy consumption and stability of the product, and the voltage may be closely related to the normal working range of the circuit. Therefore, according to the design requirements and actual application scenarios of the product, a corresponding weight value is assigned to each parameter deviation. Then, the deviation of the first iteration-related current monitoring parameter is multiplied by its corresponding weight, the deviation of the first iteration-related voltage monitoring parameter is multiplied by its weight, the deviation of the first iteration-related power monitoring parameter is multiplied by its weight, and the deviation of the first iteration-related frequency monitoring parameter is multiplied by its weight. Finally, these weighted deviations are summed up to obtain the first comprehensive deviation. This first comprehensive deviation comprehensively considers the deviation of each key parameter and its importance to product performance. It can more comprehensively and accurately reflect the overall difference between the product and the standard performance requirements, and provides a comprehensive quantitative indicator for subsequent judgment of whether the product is qualified, ensuring the scientificity and reliability of the test result evaluation.

[0048] Determine whether the first comprehensive deviation meets the preset requirements. If so, it means that the deviation of the product from the standard in various performance indicators is within an acceptable range, and the test result of the first target PCBA product at the workstation corresponding to the first test signal set is determined to be qualified. Otherwise, if the preset requirements are not met, it is determined to be unqualified.

[0049] The data processing unit repeats the above steps for multiple test signal sets to conduct comprehensive test and qualification certification, and adds all the multiple target PCBA products that have passed the test one by one into the first test qualified target PCBA product set, providing screened high-quality product samples for subsequent further load working signal acquisition and final product qualification judgment, ensuring the accuracy and effectiveness of the entire test process, and helping to ensure the high-quality production of products such as vehicle-mounted pump controller circuit boards.

[0050] In a possible implementation, step S520 further includes:

[0051] Step S521: extracting the first long-time-series current monitoring parameter at the first position and the second long-time-series current monitoring parameter at the second position in the first long-time-series current monitoring parameter sequence, and performing iterative correlation analysis to obtain a first iterative current monitoring parameter.

[0052] Step S522: performing iterative correlation identification on the first iterative current monitoring parameter and the first current long-time monitoring parameter located at the third position to obtain a second iterative current monitoring parameter.

[0053] Step S523: after multiple iterations, until the last first current long-time monitoring parameter in the first current long-time monitoring parameter sequence completes the iteration, a first iteration-associated current monitoring parameter is obtained.

[0054] Step S524: Iteratively associate and identify the first voltage long-time monitoring parameter sequence, the first power long-time monitoring parameter sequence and the first frequency long-time monitoring parameter sequence to obtain the first iterative associated voltage monitoring parameters, the first iterative associated power monitoring parameters and the first iterative associated frequency monitoring parameters.

[0055] Specifically, for the first current long-time monitoring parameter sequence, first select the first current long-time monitoring parameter in the first position and the first current long-time monitoring parameter in the second position. Then, based on the initial understanding of the product current characteristics and past technical experience and other factors, determine the initial importance weights for these two parameters. Next, multiply the first current long-time monitoring parameter in the first position by its corresponding initial importance weight, and then multiply the first current long-time monitoring parameter in the second position by its corresponding initial importance weight, and finally add the results of the two multiplication operations to obtain the first iterative current monitoring parameter. This process is based on the initial judgment of the importance of the first two data points at the beginning of the current change. A weighted comprehensive consideration is made to preliminarily depict the possible change trend of the current.

[0056] Next, the first iterative current monitoring parameter obtained in the previous step is iteratively associated with the first current long-time monitoring parameter located in the third position. At this time, based on the technical association between the newly added parameters and the existing parameters, the new importance weights of each parameter are re-determined. First, the first iterative current monitoring parameter is multiplied by its new importance weight, and then the first current long-time monitoring parameter located in the third position is multiplied by its new importance weight. Finally, the results of these two multiplication operations are added together to obtain the second iterative current monitoring parameter. In this step, the importance of the parameters to the overall trend judgment is re-evaluated based on the new data points during the current change process, and the understanding of the current characteristics is optimized through weighted calculation.

[0057] The iterative process continues, and in each iteration, the algorithm dynamically updates the importance weights based on the current parameter status and the newly added parameters. For example, in a certain iteration, the parameters obtained in the previous iteration are first multiplied by their corresponding new importance weights, and then the new first current long-time monitoring parameters are multiplied by their corresponding new importance weights, and finally the results of these two multiplication operations are added together to obtain a new iteration result. This cycle is repeated until the last parameter in the first current long-time monitoring parameter sequence completes the iteration, and the first iteration associated current monitoring parameter is obtained, which comprehensively reflects the entire current long-time change situation and can accurately characterize the product current performance.

[0058] Finally, the weighted progressive analysis iterative association identification method is used to process the first voltage long-time monitoring parameter sequence, the first power long-time monitoring parameter sequence and the first frequency long-time monitoring parameter sequence respectively. For the voltage sequence, in each iteration, the voltage parameters at different positions are calculated according to a similar importance weight adjustment strategy, and the first iteration associated voltage monitoring parameters are gradually obtained, which can accurately reflect the performance characteristics of the voltage during the long-time test process. For the power sequence, this algorithm is also used to integrate the information of the power parameters at each position through continuous iteration and adjustment based on the importance weight, so as to determine the first iteration associated power monitoring parameters, which are used to accurately reflect the power characteristics of the product. In terms of the frequency sequence, the same operation is performed, and the first iteration associated frequency monitoring parameters are finally obtained, so that it can well represent the operating status of the product in terms of frequency. Through this series of steps, the weighted progressive analysis algorithm is used to comprehensively and deeply explore the intrinsic relationship between each parameter sequence in the first test signal set, providing reliable and representative key monitoring parameters for subsequent deviation analysis and comprehensive deviation calculation, which helps to accurately determine whether products such as vehicle-mounted pump controller circuit boards meet quality requirements.

[0059] In a possible implementation, step S521 further includes:

[0060] Step S5211: perform parameter component similarity identification on the first current long-term monitoring parameter at the first position and the second current long-term monitoring parameter at the second position to obtain a first iterative current parameter component similarity set.

[0061] Step S5212: normalize the first iterative current parameter component similarity set to obtain a first iterative coefficient set.

[0062] Step S5213: Use a graph convolutional network to perform convolution calculation on the first iteration coefficient set and the second-order first current long-term monitoring parameter to obtain a first iteration current monitoring parameter.

[0063] Specifically, a detailed similarity identification work is carried out on the two first current long-time monitoring parameters at the beginning. For the current amplitude, the amplitude information in the first and second first current long-time monitoring parameters is extracted respectively. Then, the cosine similarity calculation formula is used to regard the two amplitude data as vectors, and the cosine similarity between them is calculated. By calculating the dot product of the two amplitude vectors divided by the product of their modulus lengths, a value between -1 and 1 is obtained. The closer the value is to 1, the higher the similarity in the current amplitude. In terms of current waveform, the current waveforms of the two parameters are digitized and converted into a series of discrete data points. Then, the cosine similarity calculation method is also used to substitute the vector composed of these data points into the formula for calculation. By comparing the calculation results with the set similarity threshold, the similarity between the two in the current waveform is judged. For the current frequency, the two parameters are first subjected to spectrum analysis to obtain their frequency component data. Afterwards, these frequency data are used as vectors, and the cosine similarity formula is used to calculate the similarity between them, which helps to accurately evaluate the closeness of the two parameters in the current frequency characteristics. Finally, for the current peak, the current peak values ​​of the first and second parameters are extracted, constructed into vector form, and then their similarity is calculated using cosine similarity. By calculating the four parameter components of current amplitude, current waveform, current frequency and current peak respectively based on cosine similarity, a first iteration current parameter component similarity set that fully reflects the similarity of the two parameters in various key aspects is finally obtained.

[0064] After obtaining the similarity set of the first iterative current parameter components, it needs to be normalized. The purpose of normalization is to map the values ​​of the similarities of each component to a specific range (for example, between 0 and 1) so that there is a unified standard and comparability in subsequent calculations. Through the normalization algorithm, the similarity values ​​of components such as current amplitude, current waveform, current frequency and current peak are converted. Using the minimum-maximum normalization method, the similarity of each component is compared and calculated with the minimum and maximum values ​​in the set, so that they present a reasonable distribution within the new numerical range. After normalization, each coefficient in the obtained first iterative coefficient set can more accurately reflect its relative importance in the overall similarity assessment, providing a suitable data basis for subsequent convolution calculations.

[0065] The graph convolution network, a powerful tool, is used to perform convolution calculations on the first iteration coefficient set and the second first current long-time monitoring parameter. The graph convolution network can effectively capture the local and global feature relationships in the data. The coefficients in the first iteration coefficient set are used as weights and convolved with the second first current long-time monitoring parameter. During the convolution process, the different features in the second first current long-time monitoring parameter are weighted and fused according to the size of the coefficient. For example, for current parameter components with higher coefficient values ​​(i.e., higher similarity weights) after normalization, they will be given greater influence in the convolution calculation, so that the final first iteration current monitoring parameters can comprehensively consider the similarity between the current parameter components and the characteristics of the second first current long-time monitoring parameter. Through such convolution calculations, the first iteration current monitoring parameters obtained can more accurately reflect the trend of current characteristic changes in the iterative correlation analysis, provide more reliable data support for subsequent iteration steps, and help improve the accuracy and effectiveness of testing products such as vehicle-mounted pump controller circuit boards.

[0066] In a possible implementation, step S600 further includes:

[0067] Step S610: traverse the multiple load working signal sets to perform parameter mean calculations to obtain multiple load working signal parameter mean sets.

[0068] Step S620: traverse the multiple load working signal sets to perform parameter fluctuation variance calculation to obtain multiple load working signal fluctuation variance sets.

[0069] Step S630: using a feature recognition network layer to perform feature recognition on the plurality of load working signal parameter mean sets and the plurality of load working signal fluctuation variance sets to obtain the plurality of load working signal feature sets.

[0070] Specifically, a traversal operation is started for multiple load working signal sets. For each type of parameter such as current, voltage, power and frequency in each load working signal set, the mean is calculated one by one. For example, when calculating the mean of the current parameter, all current data points in the set are added and then divided by the total number of data points to obtain the mean of the current parameter of the load working signal set. The same calculation method is used for parameters such as voltage, power and frequency, and finally multiple load working signal parameter mean sets are obtained. These mean sets can reflect the average level of various parameters when the load is working to a certain extent, and are a preliminary and holistic quantitative description of the load working signal.

[0071] Similarly, multiple sets of load working signals are traversed. This time, the focus is on calculating the fluctuation variance of various parameters in each set. Taking the calculation of voltage parameter fluctuation variance as an example, the difference between each voltage data point and the voltage mean in the set is first calculated, and these differences are squared, and then the average is calculated. The result is the voltage fluctuation variance of the load working signal set. The current, power, frequency and other parameters are also calculated in this way, and then multiple sets of load working signal fluctuation variances are obtained. The fluctuation variance can reflect the fluctuation degree and stability of the load working signal during operation, which provides an important basis for subsequent more in-depth analysis.

[0072] After obtaining multiple load working signal parameter mean sets and multiple load working signal fluctuation variance sets, the feature recognition network layer begins to process them in depth. The feature recognition network layer first integrates the input mean set and variance set, pairs and organizes the data points associated with the two, and forms a data set format suitable for analysis. Then, the neural network algorithm is used to start mining the hidden feature information in the data set. For each parameter mean in the mean set, it analyzes the relationship between different parameter means, such as whether there is a certain proportional relationship between the current mean and the voltage mean, and the law of coordinated change of the power mean and the frequency mean. At the same time, combined with the data in the fluctuation variance set, the influence of parameter fluctuation variance on the overall characteristics is explored. For example, a larger current fluctuation variance may be associated with certain specific power mean and voltage mean combinations, which may imply the instability of the load operation. Through the analysis and learning of a large number of data points, the feature recognition network layer gradually identifies a series of feature patterns that can characterize the characteristics of the load working signal. These patterns include the typical performance of different parameters under specific mean and variance combinations, as well as their trend characteristics over time. Finally, these identified characteristic patterns are sorted and summarized to form multiple load working signal feature sets. These feature sets can accurately reflect various states of load working, such as stable features under normal working conditions, abnormal features when there may be hidden faults, etc., providing a valuable basis for subsequent accurate judgment of the performance of products such as vehicle-mounted pump controller circuit boards under load working conditions.

[0073] In a possible implementation, step S630 further includes:

[0074] Step S631: Acquire a plurality of sample load working signal parameter mean value sets and a plurality of sample load working signal fluctuation variance sets as training data sets.

[0075] Step S632: Use the training data set to perform supervised training on the framework built based on the convolutional neural network until convergence, and obtain a trained feature recognition network layer.

[0076] Specifically, multiple sample load working signal parameter mean sets and multiple sample load working signal fluctuation variance sets are collected and sorted as training data sets. These sample data are widely representative, covering the signal characteristics of products such as vehicle-mounted pump controller circuit boards under various normal and potentially abnormal load working conditions. For example, it includes data related to load working signals under different load intensities, different working hours, and different environmental conditions. By collecting and processing a large number of actual samples, it is ensured that the training data set can fully reflect the various changes in load working signals, providing a solid data foundation for subsequent network training.

[0077] The obtained training data set is used to supervise the training of the framework built based on the convolutional neural network. The convolutional neural network framework has powerful feature extraction and pattern recognition capabilities. During the training process, the mean set and fluctuation variance set of the sample load working signal parameters in the training data set are input into the network framework. The network continuously optimizes the processing and feature extraction capabilities of the input data by adjusting the convolution kernel parameters, neuron connection weights, etc. in the network according to the pre-set objective function (such as minimizing the classification error or maximizing the feature recognition accuracy, etc.). In each iterative training, the difference between the output of the network and the true label corresponding to the sample data (such as normal working state or specific fault type, etc.) is calculated, and the network parameters are adjusted by back propagation based on this difference. This process continues until the performance indicators of the network (such as accuracy, loss function value, etc.) reach a stable state, that is, convergence. After such a training process, the trained feature recognition network layer is finally obtained. The network layer can accurately identify key features from the mean set and fluctuation variance set of the load working signal parameters, providing reliable support for judging the performance status of the product when working under load.

[0078] In one possible implementation, Figure 2 As shown, step S100 also includes:

[0079] Step S110: After loading is completed, the visual inspection module is used to collect product images to obtain a target loading image.

[0080] Step S120: using a workstation positioning identifier to perform positioning identification on the target loading image to obtain a positioning identification result.

[0081] Step S130: When the positioning recognition result is that the positioning recognition is passed, a start signal is obtained after closing the cover.

[0082] Specifically, in the operation process of the multi-station FCT test system, when the product is loaded, the image acquisition phase begins. Through the visual inspection module, the loaded product is captured in all directions to obtain the target loading image. The visual inspection module has high-resolution imaging capabilities and can clearly record various detailed information such as the appearance and position of the product at the loading station, providing a rich data foundation for subsequent accurate analysis.

[0083] The station location identifier (neural network model) is used to locate and identify the acquired target loading image. The neural network model has been trained with a large amount of sample data and has strong image recognition capabilities. It will conduct in-depth analysis of key information such as product features and position relationships in the target loading image, and compare the elements in the image with the preset standard template or model to determine whether the product is in the correct position on the loading station, and whether its posture, direction, etc. meet the requirements, and finally obtain the positioning and recognition results.

[0084] When the positioning recognition result is passed, it means that the position and posture of the product on the loading station meet the requirements of the test system. At this time, the cover is closed, and after the cover is successfully closed, the corresponding sensor or control logic is triggered to obtain a start signal. This start signal will serve as a trigger point for subsequent test processes, such as starting a series of operations such as electrical performance testing and functional testing of the product, to ensure that the entire test process is carried out in an orderly manner based on the correct positioning of the product, and to improve the accuracy and reliability of the test results.

[0085] It should be noted that the above-mentioned sequence of the embodiments of the present application is only for description and does not represent the advantages and disadvantages of the embodiments. And the above-mentioned specific embodiments of this specification are described. In addition, the processes depicted in the accompanying drawings do not necessarily require the specific order or continuous order shown to achieve the desired results. In some embodiments, multitasking and parallel processing are also possible or may be advantageous.

[0086] The above description is only a preferred embodiment of the present application and is not intended to limit the present application. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present application should be included in the protection scope of the present application.

[0087] This specification and drawings are merely exemplary illustrations of the present application and are deemed to cover any and all modifications, variations, combinations or equivalents within the scope of the present application. Obviously, a person skilled in the art may make various modifications and variations to the present application without departing from the scope of the present application. Thus, if these modifications and variations of the present application fall within the scope of the present application and its equivalents, the present application intends to include these modifications and variations.

Claims

1. A multi-task parallel testing method for a multi-station FCT test system, characterized in that: The method comprises: Load multiple target PCBA products to the corresponding stations on the needle bed respectively, and obtain a start signal after closing the cover; Use a barcode scanner to scan multiple target PCBA products on multiple workstations to obtain multiple production work order information; Calling a burner to burn the test programs for the multiple target PCBA products according to the multiple production work order information to obtain multiple test programs, wherein each test program corresponds to a target PCBA product; Obtaining test instructions to execute the multiple test programs on the multiple target PCBA products, and using a data acquisition card to traverse the test signals of the multiple workstations to obtain multiple test signal sets; Transmitting the plurality of test signal sets to an industrial computer for test qualification certification to obtain a first test qualified target PCBA product set; Traversing the first test qualified target PCBA product set to collect load working signals, using an industrial computer to perform feature recognition on multiple load working signal sets to obtain multiple load working signal feature sets; Determine whether the plurality of load working signal feature sets meet preset requirements, and if so, obtain a second test qualified target PCBA product set; The target PCBA product set that has passed the second test is uploaded to the MES, and the workstations where the products have passed the test are displayed through the machine indicator lights.

2. The multi-task parallel testing method for a multi-station FCT testing system as claimed in claim 1, characterized in that: Acquire test instructions to execute the multiple test programs on the multiple target PCBA products, and use a data acquisition card to traverse the test signals of the multiple stations to obtain multiple test signal sets, including: The data acquisition card performs long-term monitoring on the test signals of the multiple workstations to obtain multiple initial test signal sets, wherein each initial test signal set includes an initial current long-term monitoring parameter sequence, an initial voltage long-term monitoring parameter sequence, an initial power long-term monitoring parameter sequence, and an initial frequency long-term monitoring parameter sequence; The multiple initial test signal sets are traversed by interpolation method to process missing data values, and multiple test signal sets are obtained, wherein each test signal set includes a current long-time monitoring parameter sequence, a voltage long-time monitoring parameter sequence, a power long-time monitoring parameter sequence and a frequency long-time monitoring parameter sequence.

3. The multi-task parallel testing method for a multi-station FCT testing system as claimed in claim 2, characterized in that: The plurality of test signal sets are transmitted to an industrial computer for test qualification certification to obtain a first test qualified target PCBA product set, including: extracting a first test signal set from the plurality of test signal sets; In the data processing unit of the industrial computer, iteratively associate and identify the first current long-time monitoring parameter sequence, the first voltage long-time monitoring parameter sequence, the first power long-time monitoring parameter sequence and the first frequency long-time monitoring parameter sequence of the first test signal set to determine the first iteration-associated current monitoring parameter, the first iteration-associated voltage monitoring parameter, the first iteration-associated power monitoring parameter and the first iteration-associated frequency monitoring parameter; Perform deviation analysis on the first iteration-associated current monitoring parameter, the first iteration-associated voltage monitoring parameter, the first iteration-associated power monitoring parameter and the first iteration-associated frequency monitoring parameter from the corresponding standard values ​​to obtain a first iteration-associated current monitoring parameter deviation, a first iteration-associated voltage monitoring parameter deviation, a first iteration-associated power monitoring parameter deviation and a first iteration-associated frequency monitoring parameter deviation; Performing a weighted analysis on the first iteration-associated current monitoring parameter deviation, the first iteration-associated voltage monitoring parameter deviation, the first iteration-associated power monitoring parameter deviation, and the first iteration-associated frequency monitoring parameter deviation to obtain a first comprehensive deviation; Determine whether the first comprehensive deviation meets the preset requirement, and if so, the test result of the first target PCBA product at the workstation corresponding to the first test signal set is qualified; The data processing unit is used to perform test qualification authentication on the multiple test signal sets, and multiple target PCBA products that have passed the test are added into a first test qualified target PCBA product set.

4. The multi-task parallel testing method for a multi-station FCT testing system as claimed in claim 3, characterized in that: In the data processing unit of the industrial computer, iterative association identification is performed on a first current long-time monitoring parameter sequence, a first voltage long-time monitoring parameter sequence, a first power long-time monitoring parameter sequence, and a first frequency long-time monitoring parameter sequence of the first test signal set to determine a first iteration-associated current monitoring parameter, a first iteration-associated voltage monitoring parameter, a first iteration-associated power monitoring parameter, and a first iteration-associated frequency monitoring parameter, including: Extract the first current long time-series monitoring parameter at the first position in the first current long time-series monitoring parameter sequence and the second current long time-series monitoring parameter to perform iterative correlation analysis to obtain a first iterative current monitoring parameter; Iteratively associate and identify the first iterative current monitoring parameter and the first current long-time monitoring parameter located at the third position to obtain a second iterative current monitoring parameter; After multiple iterations, until the last first current long-time monitoring parameter in the first current long-time monitoring parameter sequence completes iteration, a first iteration associated current monitoring parameter is obtained; Iteratively associate and identify the first voltage long-time monitoring parameter sequence, the first power long-time monitoring parameter sequence, and the first frequency long-time monitoring parameter sequence to obtain the first iteratively associated voltage monitoring parameters, the first iteratively associated power monitoring parameters, and the first iteratively associated frequency monitoring parameters.

5. The multi-task parallel testing method for a multi-station FCT testing system as claimed in claim 4, characterized in that: include: Perform parameter component similarity identification on the first current long-time monitoring parameter at the first position and the second current long-time monitoring parameter at the second position to obtain a first iterative current parameter component similarity set; Normalizing the first iterative current parameter component similarity set to obtain a first iterative coefficient set; A graph convolution network is used to perform convolution calculation on the first iteration coefficient set and the second-bit first current long-time monitoring parameter to obtain a first iteration current monitoring parameter.

6. The multi-task parallel testing method for a multi-station FCT testing system according to claim 1, characterized in that: Traversing the first test qualified target PCBA product set to collect load working signals, using an industrial computer to perform feature recognition on multiple load working signal sets, and obtaining multiple load working signal feature sets, including: Traversing the multiple load working signal sets to perform parameter mean calculation to obtain multiple load working signal parameter mean sets; Traversing the plurality of load working signal sets to perform parameter fluctuation variance calculation to obtain a plurality of load working signal fluctuation variance sets; The feature recognition network layer is used to perform feature recognition on the multiple load working signal parameter mean value sets and the multiple load working signal fluctuation variance sets to obtain the multiple load working signal feature sets.

7. The multi-task parallel testing method for a multi-station FCT testing system as claimed in claim 6, characterized in that: include: Acquire a plurality of sample load working signal parameter mean value sets and a plurality of sample load working signal fluctuation variance sets as training data sets; The training data set is used to perform supervised training on a framework built based on a convolutional neural network until convergence, thereby obtaining a trained feature recognition network layer.

8. The multi-task parallel testing method for a multi-station FCT testing system according to claim 1, characterized in that: Load multiple target PCBA products to the corresponding stations on the needle bed, and get the start signal after closing the cover, including: After loading is completed, the visual inspection module is used to collect product images to obtain the target loading image; Using a workstation positioning identifier to position and identify the target loading image to obtain a positioning and identification result; When the positioning identification result is that the positioning identification is passed, a start signal is obtained after the cover is closed.

Images