A data processing method for PCB function test

Abstract

The present application relates to the technical field of data processing, and particularly relates to a data processing method for PCB function test, comprising: collecting the voltage of components and determining a data processing period; obtaining preliminary abnormal components according to voltage deviation; obtaining the mutation synchronism between components according to the voltage mutation time interval between components, determining the correlation probability between components according to the voltage difference between components and the mutation synchronism, and determining the associated components of the components; taking the proportion of the same components between the associated components as the consistency between the associated components, and grouping data based on the consistency between the associated components; and identifying functionally abnormal components according to the difference between the preliminary abnormal components and other components in the same group. The present application effectively improves the accuracy of PCB function test.

Description

Technical Field

[0001] This application relates to the field of data processing technology, specifically to a data processing method for PCB board functional testing. Background Technology

[0002] In modern electronic products, the PCB (Printed Circuit Board) is the fundamental platform that carries electronic components, playing a crucial role in connecting and supporting various components. With the increasing complexity and high-performance requirements of electronic devices, PCB quality inspection has become particularly important. Functional testing, as a key aspect of PCB quality control, aims to verify the functional performance of the circuit board under normal operating conditions. It simulates actual working environments to test the electrical characteristics, signal transmission quality, and operational status of various functional modules, thereby determining whether the PCB meets design requirements.

[0003] Traditional PCB testing methods primarily rely on visual inspection and X-ray testing to assess the appearance quality of circuit boards. These methods are ineffective at evaluating the functionality of PCBs in real-world operating environments. Therefore, current PCB functional testing involves collecting actual parameters of each component and analyzing the test data to determine their functionality. However, existing test data processing methods often involve reasoning and analysis based on multiple test data points for individual components. This results in complex experimental data, and the parameters of actual components are correlated. A functional malfunction in one component can lead to deviations in the parameters of other components, causing the test data for a single component to fail to directly reflect its functional deviation, thus leading to inaccurate functional test results. Summary of the Invention

[0004] This invention provides a data processing method for PCB board functional testing to solve existing problems.

[0005] The data processing method for PCB board functional testing of the present invention adopts the following technical solution:

[0006] One embodiment of the present invention provides a data processing method for PCB board functional testing, the method comprising the following steps:

[0007] The voltage of the components is collected, the voltage data is segmented, and multiple data processing cycles are obtained.

[0008] Based on the deviation between the actual voltage and the rated voltage, preliminary abnormal components are identified;

[0009] Within a data processing cycle, the correlation probability between components is determined based on the voltage difference and the synchronization of sudden changes between components; wherein the voltage difference between components is negatively correlated with the correlation probability between components, and the synchronization of sudden changes between components is positively correlated with the correlation probability between components; the synchronization of sudden changes between components is negatively correlated with the minimum time interval of sudden voltage changes between components; and the associated components are determined based on the correlation probability between components.

[0010] Based on the voltage change and voltage mutation difference between data processing cycles, the probability of changes in the associated components of the component is obtained. The probability of changes in the associated components of the component is positively correlated with the voltage change and voltage mutation difference between data processing cycles. When the probability of changes in the associated components is greater than a preset threshold, the associated components of the component are recalculated. When the probability of changes in the associated components is less than or equal to the preset threshold, the associated components of the component remain unchanged.

[0011] The proportion of identical components among related components is used as the consistency between related components.

[0012] Data is grouped based on the consistency between related components; within the same group, functionally abnormal components are identified based on the differences between the initially abnormal components and other components.

[0013] Preferably, the specific steps for obtaining preliminary abnormal components based on the deviation between the actual voltage and the rated voltage are as follows:

[0014] (1) For a component, calculate the absolute value of the voltage difference between the actual voltage and the rated voltage. And calculate the absolute value of the difference between the actual voltage at each moment and the actual voltage at the previous moment. ;

[0015] (2) The absolute value of the voltage difference at any given time The absolute value of the difference between the voltage and the voltage The sum of these values ​​represents the initial anomaly probability of a component at any given time. ;

[0016] (3) For each type of component, select the maximum value of the initial abnormality probability at each time step; preset an abnormality threshold, and determine the components whose maximum value of the initial abnormality is greater than the abnormality threshold as initial abnormal components.

[0017] Preferably, the specific steps for determining the correlation probability between components based on voltage differences and abrupt change synchronization are as follows:

[0018] The synchronization of sudden voltage changes between components is obtained based on the minimum time interval between sudden voltage changes between components.

[0019] The ratio of the synchronization of sudden changes between components to the voltage difference between components is used as the correlation probability between components.

[0020] Preferably, the specific steps for obtaining the synchronization of voltage changes between components based on the minimum time interval between voltage changes between components are as follows:

[0021] Within a data processing cycle, determine all abrupt changes of the components;

[0022] Calculate the minimum time interval between all abrupt change times of one component and the abrupt change times of another component, and obtain the average minimum time interval of all abrupt change times;

[0023] The reciprocal of the mean of the minimum time interval is used as the synchronization of sudden changes between components.

[0024] Preferably, the specific steps for determining the associated components based on the association probability between components are as follows:

[0025] Within a data processing cycle, determine the association probability of a component with all other components; select the N components with the highest association probabilities as associated components of a component; N is a preset value.

[0026] Preferably, the specific steps for obtaining the probability of changes in associated components based on voltage changes and voltage abrupt changes between data processing cycles include the following:

[0027] (1) Within one data processing cycle, calculate the average voltage of each component; then calculate the absolute value of the difference between the average voltages of the same component in different data processing cycles; finally, calculate the average of the absolute values ​​of the differences of all components as the voltage change between data processing cycles. ;

[0028] (2) Within a data processing cycle, determine the number of times each component experiences voltage fluctuations; then calculate the difference in the number of times the same component experiences voltage fluctuations in different data processing cycles; finally, calculate the average of the differences in the number of times the voltage fluctuations occur for all components, as the difference in voltage fluctuations between data processing cycles. ;

[0029] (3) The voltage change between the data processing cycles The difference in voltage fluctuations between the data processing cycles. The product of these factors represents the probability of changes in the associated components.

[0030] Preferably, the specific steps for determining the consistency between related components by the proportion of identical components among related components are as follows:

[0031] In the pair of corresponding related components, the number of identical components is determined, and then the ratio of the number of identical components to the number of related components is taken as the consistency between the related components; wherein the number of related components is a preset value.

[0032] Preferably, the specific steps for grouping data based on the consistency between associated components are as follows:

[0033] (1) Treat the data corresponding to each component in a data processing cycle as a single data, wherein a single data includes a component, the component’s associated components, the component, and the test data corresponding to the component’s associated components.

[0034] (2) Determine the consistency of the related components between two individual data points;

[0035] (3) Based on the consistency of the associated components of the two individual data corresponding to each other, all individual data are clustered to obtain multiple clusters; one cluster contains multiple individual data, so one cluster is recorded as a group of data.

[0036] Preferably, the identification of functionally abnormal components within the same group, based on the differences between the initially abnormal components and other components, includes the following specific steps:

[0037] (1) Take the individual data corresponding to the time when the initial abnormality probability of a component is at its maximum value as a feature data of a component; then determine the data group in which the feature data is located;

[0038] (2) In the determined data grouping, first calculate the average voltage of each component at multiple test times corresponding to the group; then calculate the absolute value of the difference between the average voltage of the initially abnormal component and other components, as the difference between the initially abnormal component and other components.

[0039] (3) Select functionally abnormal components based on the differences between the preliminary abnormal components and other components.

[0040] Preferably, the specific steps for screening out functionally abnormal components based on the differences between the initially abnormal components and other components are as follows:

[0041] A preset functional abnormality threshold is set, and components that are initially abnormal and whose difference from other components is greater than the preset functional abnormality threshold are identified as functional abnormal components.

[0042] The beneficial effects of the technical solution of this invention are as follows: Functional testing of the PCB board is performed directly using a single parameter, reducing the types of data analyzed and thus reducing computational complexity. Simultaneously, by establishing parameter relationships between PCB components, associated components are identified. These associated components are updated to reflect changes in parameters during the testing process, ensuring stable parameter relationships between all components and their associated components. Furthermore, the test data is grouped based on the consistent relationships between associated components, ensuring comparative analysis of test data within the same group. Through comparison of data within the same group, components exhibiting functional abnormalities can be accurately identified. Attached Figure Description

[0043] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0044] Figure 1 This is a flowchart illustrating the steps of a data processing method for PCB board functional testing according to the present invention. Detailed Implementation

[0045] To further illustrate the technical means and effects adopted by the present invention to achieve its intended purpose, the following, in conjunction with the accompanying drawings and preferred embodiments, details the specific implementation, structure, features, and effects of a data processing method for PCB board functional testing proposed according to the present invention. In the following description, different "one embodiment" or "another embodiment" do not necessarily refer to the same embodiment. Furthermore, specific features, structures, or characteristics in one or more embodiments can be combined in any suitable form.

[0046] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention pertains.

[0047] The following description, in conjunction with the accompanying drawings, details a specific scheme for a data processing method for PCB board functional testing provided by the present invention.

[0048] Please see Figure 1The diagram illustrates a flowchart of a data processing method for PCB board functional testing according to an embodiment of the present invention. The method includes the following steps:

[0049] S001: Collect dynamic voltage data of components and determine the data processing cycle.

[0050] In PCB board functional testing, component functionality is the primary focus. The operating status of components inevitably affects their voltage; therefore, voltage data collection is the first step in the testing process. Since PCB board functional testing is a dynamic process, voltage data acquisition is also dynamic.

[0051] The equipment and parameters required for dynamic voltage acquisition are as follows:

[0052] (1) Digital oscilloscope: used for dynamic voltage acquisition, capable of recording the change of transformer voltage over time.

[0053] Bandwidth: ≥100MHz, capable of capturing high-frequency ripple.

[0054] Sampling rate: ≥1GS / s, ensuring no distortion of signals below 500kHz.

[0055] Vertical resolution: ≥8 bits, quantization error less than 0.4%.

[0056] Probe: 10:1, passive probe.

[0057] (2) Data acquisition card: multi-channel synchronous acquisition, suitable for voltage acquisition of batch components, i.e. suitable for mass production testing.

[0058] Number of channels: 16 or more, capable of simultaneously acquiring the voltage of multiple components.

[0059] Sampling rate: greater than or equal to 10kHz, capable of accommodating both static and slow dynamic voltages.

[0060] Accuracy: ±0.1%, meeting the functional testing requirements of components.

[0061] Input range: ±10V, covering common voltages on PCB boards.

[0062] (3) Test fixture / probe station: stably connects the PCB test points and the acquisition equipment to ensure reliable connection between the acquisition equipment and the component pins / test points and reduce contact interference.

[0063] Probe diameter: 0.3-0.8mm, matching the size of PCB test points.

[0064] Contact resistance: <10mΩ, to avoid voltage measurement errors caused by contact resistance.

[0065] Probe material: beryllium copper, wear-resistant, ensuring stable contact during multiple tests.

[0066] During PCB board functional testing, a complete data acquisition process is required to ensure the integrity and accuracy of the acquired data, thereby guaranteeing the reliability of the functional test. Therefore, the specific implementation process for data acquisition in PCB board functional testing is as follows:

[0067] (1) Preliminary preparation

[0068] Determine the target components: List the components and pins whose voltages need to be measured.

[0069] Locating test points: Based on the PCB layout diagram, determine the test points corresponding to the component pins and mark their positions to avoid not being able to find the test points during testing.

[0070] Create a data collection list: record data according to "component reference number + pin function + expected voltage range".

[0071] (2) Equipment calibration

[0072] Oscilloscope calibration: Use an oscilloscope to calibrate the signal source, check the amplitude error and timing error measured by the oscilloscope, and ensure that the dynamic voltage waveform is free from distortion.

[0073] Test fixture inspection: Use a multimeter to measure the probe contact resistance to avoid low voltage measurement values ​​due to poor contact.

[0074] (3) Set up the data collection environment

[0075] Power supply settings: Power on the PCB board to ensure stable input voltage and avoid voltage distortion of components due to abnormal power supply.

[0076] Reduce electromagnetic interference:

[0077] The test bench is covered with grounded copper foil, and the equipment casing is grounded to avoid electrostatic interference;

[0078] Shielded cables with a length of ≤1m are used for signal acquisition to reduce high-frequency noise coupling.

[0079] When acquiring dynamic voltage, a short grounding spring is used to ground the oscilloscope probe instead of a long grounding wire to avoid introducing high-frequency noise.

[0080] (4) Dynamic voltage acquisition

[0081] Using a digital oscilloscope, measure the absolute value of the voltage under steady-state conditions and record the voltage data at each sampling moment.

[0082] During PCB board testing, the data collected is easily affected by the environment and operation, leading to data deviation. To ensure the accuracy of the collected data, the following precautions should be taken:

[0083] (1) Anti-interference processing

[0084] During dynamic acquisition, the oscilloscope probe should be kept away from high-frequency components on the PCB to avoid false ripples caused by electromagnetic coupling;

[0085] When measuring low voltage, use the "millivolt" setting on the multimeter, not the "volt" setting.

[0086] (2) Synchronization guarantee

[0087] When acquiring data from multiple channels, use a data acquisition card or oscilloscope to trigger multiple channels synchronously to ensure consistent timestamps and keep the error <1ms.

[0088] (3) Probe contact reliability

[0089] When the test point is oxidized, first gently wipe the test point with fine sandpaper, and then contact it with the probe to avoid excessive contact resistance, which would cause the voltage measurement value to be too low.

[0090] For SMD components without exposed pins, use a miniature probe to directly contact the pads at both ends of the component.

[0091] Based on the above steps, data acquisition is performed. Since the voltage data acquisition has a high-frequency characteristic, it is impossible to process high-frequency sampling data in actual PCB test data processing. Therefore, data segmentation is required in the timing sequence, and then the data within a segment is processed uniformly. Thus, data segmentation in the timing sequence determines a data processing cycle. The time interval between the timing segments is set to 2 seconds, meaning the length of one data processing cycle is 2 seconds. The entire test process involves acquiring voltage data for a total of 2 minutes.

[0092] S002: Based on the deviation of the component voltage, calculate the initial probability of component abnormality and identify the components with initial abnormality.

[0093] When components perform their functions, depending on their specific roles, different components will exhibit different behaviors in different data sets when malfunctioning. For example, passive components often show parameter deviations, active components often show signal and control failures, and power components often show energy conversion abnormalities. However, for components to operate, they must have voltage characteristics; the voltage across different components can indicate their functional status.

[0094] Therefore, the first step is to identify component malfunctions based on the voltage across the components. Generally, the greater the deviation between the component's voltage and its rated voltage, the higher the probability of malfunction, and the greater the voltage fluctuation, the higher the probability of functional abnormality.

[0095] As one embodiment, the process of determining the preliminary anomaly probability of any component is as follows:

[0096] (1) For a component, calculate the absolute value of the voltage difference between its actual voltage and its rated voltage at each moment. .

[0097] (2) Calculate the absolute value of the difference between the actual voltage of a component at each moment and the actual voltage at the previous moment. .

[0098] (3) Determine the initial anomaly probability of a component at any given time. for:

[0099]

[0100] in It represents the absolute value of the voltage difference between the actual voltage and the rated voltage of a component at any given time. This represents the absolute value of the difference between the actual voltage of a component at each moment and the actual voltage at the previous moment. This represents the normalization function.

[0101] Furthermore, each component has a preliminary probability of failure at every moment. Therefore, a preliminary anomaly analysis needs to be performed based on the preliminary anomaly probabilities at all moments. It should be noted that after a component malfunctions, the resulting voltage anomaly is mainly manifested in instantaneous voltage changes. Therefore, the preliminary anomaly analysis of a component's function needs to be based on the preliminary anomaly probability at a single moment.

[0102] As one example, the specific process for identifying preliminary abnormal components is as follows:

[0103] (1) For each component, select the maximum value of the initial abnormality degree from the initial abnormality probability at each time step.

[0104] (2) Set an abnormal threshold. Components whose maximum value of the initial abnormality level is greater than the abnormal threshold are identified as initially abnormal components. In this embodiment, the abnormal threshold is set to 0.7.

[0105] As another embodiment, the process of identifying preliminary abnormal components is as follows:

[0106] For a given component, calculate the absolute value of the voltage difference between its actual voltage and its rated voltage at each instant. The maximum absolute value of the voltage difference of a component at all sampling times is determined; then, the abnormal threshold is directly set to 1, and components whose maximum absolute value is greater than the abnormal threshold are identified as preliminary abnormal components.

[0107] S003: Based on the voltage differences and abrupt change synchronization between components, determine the correlation probability between components, identify the associated components, and determine whether the associated components have changed.

[0108] In PCB boards, the parameters of components are interconnected. When one component malfunctions, it can easily trigger parameter abnormalities in other components through circuit connections and energy transfer. This chain reaction is particularly common in complex circuits. Essentially, the normal connection relationships between components in the circuit are broken, forcing related components to operate in non-design states, leading to parameter deviations or even damage.

[0109] It is known that during actual PCB board functional testing, a deviation in the parameters of a single component may cause a similar deviation in other components. Therefore, a deviation in the parameters of a single component cannot directly indicate a functional malfunction. It is necessary to consider the interrelationships between components and the relationship between parameter changes of multiple components to determine the component with the final functional malfunction.

[0110] Therefore, before identifying the malfunctioning components, it is first necessary to identify other components with which a single component has parameter correlations, referred to as the associated components. The greater the voltage difference between components, the less correlated their parameters are. Furthermore, the parameter correlation between components is also manifested in parameter changes, such as sudden voltage changes. When the voltage of one component changes, the voltage of its associated parameters may also change suddenly, and the timing of these changes is close, meaning that the components exhibit synchronous changes.

[0111] It should be noted that as testing progresses, the performance of components may change, and therefore the parameter relationships with other components will also change. In other words, the associated components of a component may change as testing continues. Therefore, the associated components of a component can only be reflected within the same data processing cycle. Thus, it is necessary to determine the associated components of a component separately in each data processing cycle.

[0112] As one embodiment, the specific process for determining the associated components of a component is as follows:

[0113] (1) Within a data processing cycle, calculate the average voltage of each component at all times; and calculate the absolute value of the difference between the average voltages of any two components, which is called the voltage difference between the components, denoted as . .

[0114] (2) In the voltage sequence of each component, calculate the change of each voltage in time. Then, the voltage with a change greater than the mutation threshold is determined as the mutation voltage. At the same time, the time of the mutation voltage is recorded as the mutation time. In this embodiment, the mutation threshold is set to 2V.

[0115] (3) Within one data processing cycle, calculate the interval between abrupt changes between each pair of components. For each pair of components, calculate the minimum interval between all abrupt changes in one component and the minimum interval between all abrupt changes in the other component. Then, average the minimum intervals between all abrupt changes between each pair of components, and denote it as . .

[0116] This allows for synchronization of sudden changes between components. for:

[0117]

[0118] In the formula, This represents the average of all abrupt changes between any two components.

[0119] (4) Determine the correlation probability between components based on the voltage differences and the synchronization of sudden changes among components. for:

[0120]

[0121] In the formula, This indicates the synchronization of sudden changes between two components. This indicates the voltage difference between two components.

[0122] (5) Within a data processing cycle, a component corresponds to an association probability with all other components. At this time, for a component, the N components with the highest association probability are selected as its associated components. In this embodiment, N is set to 4.

[0123] Furthermore, in the functional processing of PCB boards, to ensure the completeness and accuracy of functional testing, the test duration generally includes multiple data processing cycles. Therefore, it is necessary to determine the associated components in each data processing cycle, which leads to a relatively large amount of computation. However, the states of the components do not all change in different data processing cycles, so the parameter relationships between the components do not all change. Therefore, the relationships between components may not exist in different data processing cycles.

[0124] Therefore, in actual data processing, it is necessary to determine whether the related components of a component have changed based on the parameter changes between data processing cycles. Specifically, the smaller the voltage change between data processing cycles, the lower the probability of changes in the related components; conversely, the greater the voltage fluctuation difference between data processing cycles, the higher the probability of changes in the related components.

[0125] As one example, the specific process for determining whether the associated components of a component have changed is as follows:

[0126] (1) Let the data processing period corresponding to the determination process of the current associated component be the known data period. Calculate the average voltage of each component in the known data processing period and the current data processing period respectively. Then calculate the absolute value of the difference between the average voltage of the same component in the known data processing period and the current data processing period. Finally, calculate the average of the absolute values ​​of the differences between the average voltages of all components as the voltage change between data processing periods. .

[0127] (2) Determine the number of voltage surges for each component in the known data processing cycle and the current data processing cycle; then calculate the difference in the number of voltage surges for the same component in the known data processing cycle and the current data processing cycle; finally, calculate the average of the differences in the number of voltage surges for all components as the difference in voltage surges between data processing cycles. .

[0128] (3) Based on the voltage change and voltage change difference between data cycles, determine the probability that the associated components of the current data processing cycle will change. for:

[0129]

[0130] In the formula, This indicates the voltage change between data processing cycles. This indicates the difference in voltage fluctuations between data processing cycles. This represents the normalization function.

[0131] (4) Set an update threshold. When the probability P2 is greater than the update threshold, the associated components of the component change. At this time, it is necessary to recalculate the associated components of each component.

[0132] S004: Determine the consistency between related components based on the proportion of identical components among related components.

[0133] The functional testing process of a PCB board involves parameter relationships between different components, and the aforementioned process has already identified the associated components.

[0134] Functional testing of electronic components often requires comparing the parameter relationships between different components to determine their functionality. Therefore, components need to be grouped, and parameters compared within each group. However, in actual functional testing, the influence of related components must be considered. Therefore, component grouping needs to be implemented based on the consistency between related components.

[0135] Given that one component corresponds to multiple associated components, the consistency of the multiple associated components of one component is first reflected in the proportion of the number of the same components. The higher the proportion of the number of the same components among the associated components of two components, the higher the consistency between the associated components.

[0136] As one example, the calculation process for consistency between related components is as follows:

[0137] First, identify the common components in each pair of related components, and then determine the percentage of common components in the related components. Where N represents the number of associated components, and m represents the number of identical components among the associated components; at this time, the proportion of the number is... The consistency S between related components, which are considered as components, is represented as:

[0138]

[0139] In the formula, N represents the number of associated components, and m represents the number of identical components among the associated components.

[0140] As another embodiment, considering the association probability of the associated components themselves, the higher the association probability of the same component among the associated components corresponding to two components, the higher the consistency between the associated components. The smaller the difference in the association probability of the same component, the higher the consistency between the associated components. The calculation process for the consistency between the associated components of corresponding components is as follows:

[0141] (1) Within the same data processing cycle, identify the common components in the associated components of each pair of components, and obtain the percentage of common components in the associated components. , where N represents the number of associated components, and m represents the number of identical components among the associated components.

[0142] (2) Among the paired components corresponding to each other, calculate the absolute value of the difference in the probability of association between the same components, denoted as . ; Calculate the sum of the association probabilities of identical components, denoted as . ;

[0143] (3) Determine the consistency between related components of two components within the same data processing cycle. for:

[0144]

[0145] In the formula, This represents the percentage of identical components among related components, where N represents the total number of related components, and m represents the percentage of identical components among the related components. This represents the absolute value of the difference in the association probability of identical components among related components. It represents the sum of the association probabilities of identical components among related components.

[0146] In some other embodiments, considering the consistency differences between associated components of components with the same data processing cycle and components with different data processing cycles, the greater the cycle distance between the data processing cycles corresponding to the associated components of two components, the smaller the consistency between the associated components. The process of achieving consistency between associated components of two components is as follows:

[0147] (1) Within the same data processing cycle, identify the common components in the associated components of each pair of components, and determine the proportion of the common components in the associated components. The consistency S between related components within the same data processing cycle is defined as N, where N represents the number of related components and m represents the number of identical components among the related components.

[0148] (2) Using the calculation method of consistency between related components of two components within the same data processing cycle, the preliminary consistency between related components of two components in different data processing cycles is determined.

[0149] (3) Determine the cycle distance between two data processing cycles. It should be noted that the cycle distance between two adjacent data processing cycles is 1.

[0150] (4) Determine the consistency between the related components of two components between different data processing cycles. for:

[0151]

[0152] In the formula, This indicates the initial consistency between related components of two components in different data processing cycles. This indicates the cycle distance between two data processing cycles.

[0153] Based on the above steps, the consistency between related components in all test data is determined.

[0154] S005: Group data based on the consistency between related components and identify malfunctioning components within each group.

[0155] Based on the above steps, the consistency of the associated components corresponding to the individual components is determined. At this point, during PCB component functional testing, the function of a single component may be reflected in the parameters of multiple components, i.e., in the associated components of that single component. Therefore, in actual functional test data analysis, it is necessary to compare multiple sets of data to identify components with abnormal functions.

[0156] Given that there are parameter correlations between known components and their corresponding related components, the actual functional testing of components requires further comparison of the parameter performance of the related components to ultimately achieve functional testing. Therefore, for the obtained component functional test data, it is first necessary to group the data according to the consistency of the component's related components.

[0157] As one example, the process of grouping component test data based on the consistency of associated components is as follows:

[0158] (1) Within a data processing cycle, each component corresponds to a single data, which includes a component, the component’s associated components, the component and the test data corresponding to the component’s associated components; the consistency between the two pairs of single data corresponds to the two pairs of associated components.

[0159] (2) Obtain the consistency of the associated components between each pair of individual data.

[0160] (3) Based on the consistency of the associated components, all individual data are clustered to obtain multiple clusters. One cluster contains multiple individual data, so one cluster is recorded as a group of data. In this embodiment, the clustering method directly adopts the mean shift algorithm.

[0161] Based on the above steps, group the individual data from all the data corresponding to the entire testing process.

[0162] Furthermore, given that a single data point contains a component with a preliminary probability of being abnormal, and that a preliminary abnormal component has been identified among all components, after grouping the individual data points, it is necessary to determine the functionally abnormal component based on the abnormal relationships between other data points within the same group.

[0163] As one embodiment, the specific process for identifying malfunctioning components is as follows:

[0164] (1) For a component with a preliminary abnormality, the time corresponding to the maximum value of its preliminary abnormality probability is determined, that is, the data processing cycle corresponding to the maximum value of the preliminary abnormality probability is determined, thereby determining the individual data corresponding to the maximum value of the preliminary abnormality probability and the group in which the individual data is located.

[0165] (2) In the grouping, a component includes multiple test times, so first calculate the average voltage of each component at the multiple test times corresponding to the group. Then calculate the deviation of the average voltage of the initially abnormal component from that of other components. .

[0166] (3) Set a functional abnormality threshold, based on the deviation of the average voltage value. When the value exceeds the aforementioned functional abnormality threshold, the component is determined to be a functionally abnormal component; otherwise, it is a functionally normal component.

[0167] Thus, by grouping data, the functional testing of components is achieved.

[0168] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A data processing method for PCB board functional testing, characterized in that, The method includes the following steps: The voltage of the components is collected, the voltage data is segmented, and multiple data processing cycles are obtained. Based on the deviation between the actual voltage and the rated voltage, preliminary abnormal components are identified; Within a data processing cycle, the correlation probability between components is determined based on the voltage difference and the synchronization of sudden changes between components; wherein the voltage difference between components is negatively correlated with the correlation probability between components, and the synchronization of sudden changes between components is positively correlated with the correlation probability between components; the synchronization of sudden changes between components is negatively correlated with the minimum time interval of sudden voltage changes between components; and the associated components are determined based on the correlation probability between components. Based on the voltage change and voltage mutation difference between data processing cycles, the probability of changes in the associated components of the component is obtained. The probability of changes in the associated components of the component is positively correlated with the voltage change and voltage mutation difference between data processing cycles. When the probability of changes in the associated components is greater than a preset threshold, the associated components of the component are recalculated. When the probability of changes in the associated components is less than or equal to the preset threshold, the associated components of the component remain unchanged. The proportion of identical components among related components is used as the consistency between related components. Data is grouped based on the consistency between related components; within the same group, functionally abnormal components are identified based on the differences between the initially abnormal components and other components.

2. The data processing method for PCB board functional testing according to claim 1, characterized in that, The specific steps for obtaining preliminary abnormal components based on the deviation between the actual voltage and the rated voltage are as follows: (1) For a component, calculate the absolute value eU of the voltage difference between the actual voltage and the rated voltage; and calculate the absolute value dU of the difference between the actual voltage at each moment and the actual voltage at the previous moment. (2) The sum of the absolute value of the voltage difference eU at any time and the absolute value of the difference between the voltages dU is taken as the initial abnormal probability P of a component at any time. (3) For each type of component, select the maximum value of the initial abnormality probability at each time step; preset an abnormality threshold, and determine the component whose maximum value of the initial abnormality is greater than the abnormality threshold as the initial abnormal component.

3. The data processing method for PCB board functional testing according to claim 1, characterized in that, The specific steps for determining the correlation probability between components based on voltage differences and abrupt change synchronization are as follows: The synchronization of sudden voltage changes between components is obtained based on the minimum time interval between sudden voltage changes between components. The ratio of the synchronization of sudden changes between components to the voltage difference between components is used as the correlation probability between components.

4. The data processing method for PCB board functional testing according to claim 3, characterized in that, The specific steps for obtaining the synchronization of voltage changes between components based on the minimum time interval between voltage changes between components are as follows: Within a data processing cycle, determine all abrupt changes of the components; Calculate the minimum time interval between all abrupt change times of one component and the abrupt change times of another component, and obtain the average minimum time interval of all abrupt change times; The reciprocal of the mean of the minimum time interval is used as the synchronization of sudden changes between components.

5. The data processing method for PCB board functional testing according to claim 1, characterized in that, The specific steps for determining the associated components based on the association probability between components are as follows: Within a data processing cycle, determine the association probability of a component with all other components; select the N components with the highest association probabilities as associated components of a component; N is a preset value.

6. The data processing method for PCB board functional testing according to claim 1, characterized in that, The specific steps for obtaining the probability of changes in associated components based on voltage changes and voltage abrupt differences between data processing cycles are as follows: (1) Calculate the average voltage of each component within a data processing cycle; then calculate the absolute value of the difference between the average voltage of the same component in different data processing cycles; finally calculate the average of the absolute values ​​of the differences of all components as the voltage change X(U) between data processing cycles. (2) Within a data processing cycle, determine the number of times each component experiences voltage mutation; then calculate the difference in the number of times the same component experiences voltage mutation in different data processing cycles; finally, calculate the average of the differences in the number of times the voltage mutation occurs for all components, as the voltage mutation difference ΔM between data processing cycles. (3) The product of the voltage change X(U) between the data processing cycles and the voltage change ΔM between the data processing cycles is used as the probability of the related components of the component changing.

7. The data processing method for PCB board functional testing according to claim 1, characterized in that, The specific steps for determining the consistency between related components by the proportion of identical components among related components are as follows: In the pair of corresponding related components, the number of identical components is determined, and then the ratio of the number of identical components to the number of related components is taken as the consistency between the related components; wherein the number of related components is a preset value.

8. The data processing method for PCB board functional testing according to claim 1, characterized in that, The specific steps for grouping data based on the consistency between related components are as follows: (1) Treat the data corresponding to each component in a data processing cycle as a single data, wherein a single data includes a component, the component’s associated components, the component, and the test data corresponding to the component’s associated components. (2) Determine the consistency of the related components between two individual data points; (3) Based on the consistency of the associated components of the two components corresponding to each pair of individual data, all individual data are clustered to obtain multiple clusters; A cluster contains multiple individual data points, so a cluster is referred to as a set of data.

9. The data processing method for PCB board functional testing according to claim 8, characterized in that, Within the same group, the identification of functionally abnormal components based on the differences between the initially abnormal components and other components includes the following specific steps: (1) Take the individual data point corresponding to the time when the initial abnormality probability of a preliminary abnormal component is maximum as a feature data point of a preliminary abnormal component; then determine the data group in which the feature data point is located. (2) In the determined data grouping, first calculate the average voltage of each component at multiple test times corresponding to the group; Then, the absolute value of the difference between the average voltage of the initially abnormal component and other components is calculated as the difference between the initially abnormal component and other components; (3) Select functionally abnormal components based on the differences between the preliminary abnormal components and other components.

10. The data processing method for PCB board functional testing according to claim 9, characterized in that, The specific steps for screening out functionally abnormal components based on the differences between the preliminary abnormal components and other components are as follows: A preset functional abnormality threshold is set, and components that are initially abnormal and whose difference from other components is greater than the preset functional abnormality threshold are identified as functional abnormal components.

Images