A PCBA board test data intelligent management system and method

Abstract

The invention discloses an intelligent management system and method for PCBA board test data, and relates to the technical field of intelligent data management. The invention establishes a two-dimensional data association cluster, maps each state data change curve in the two-dimensional data association cluster, and divides the normal state data association cluster and the abnormal state data association cluster in the two-dimensional data association cluster, establishes a component test state evaluation network according to the distribution state of the structural components in the target PCBA board and the normal state data association cluster and the abnormal state data association cluster, and then simulates the abnormal association range of the component test state evaluation network according to the historical state change record, generates a component abnormal association path according to the simulation result, collects various real-time state data change curves of each structural component in the target PCBA board and inputs them into the component test state evaluation network, and then determines whether there is an abnormality in the real-time state data change curve.

Description

Technical Field

[0001] The present invention relates to the technical field of intelligent data management, and in particular to an intelligent management system and method for PCBA board test data. Background Art

[0002] With the rapid development of electronic products and the continuous advancement of technology, the importance of PCBA in modern electronic equipment has become increasingly prominent. The quality of PCBA directly affects the performance and reliability of the product, so it is particularly important to conduct systematic testing and management on it.

[0003] As the amount of test items and test data continues to increase, traditional management methods are unable to meet the needs of rapid response and real-time analysis. This limitation prevents the value of test data from being fully tapped, reducing the flexibility and efficiency of abnormal state detection during subsequent PCBA use.

[0004] Therefore, how to ensure that the value of test data is fully mined while improving the flexibility and efficiency of PCBA abnormal state detection based on test data analysis, to this end, an intelligent management system and method for PCBA board test data is provided. Summary of the invention

[0005] In order to solve the above technical problems, the object of the present invention is to provide an intelligent management system and method for PCBA board test data.

[0006] In order to achieve the above object, the present invention provides the following technical solutions:

[0007] A PCBA board test data intelligent management method comprises the following steps:

[0008] Step S1, setting several types of sensors for each structural component in the target PCBA board, and placing the target PCBA board in a periodic voltage environment, thereby collecting multiple state data change curves of each structural component, and integrating all state data change curves to generate a historical state change record;

[0009] Step S2, establishing a two-dimensional data association cluster, recording the first generated state data change curve as the first state data change curve, and then obtaining the association value and the association coefficient between the subsequently generated state data change curve and the first state data change curve, and then mapping each state data change curve into the two-dimensional data association cluster according to the association value and the association coefficient, and dividing the two-dimensional data association cluster into a normal state data association cluster and an abnormal state data association cluster;

[0010] Step S3, establishing a component test state evaluation network according to the distribution state of the structural components in the target PCBA board and the normal state data association cluster and the abnormal state data association cluster, and then simulating the abnormal association range of the component test state evaluation network according to the historical state change record, generating a number of abnormal association paths according to the simulation results, cross-overlapping the abnormal association paths to obtain component abnormal association paths, and mapping the component abnormal association paths to the component test state evaluation network;

[0011] Step S4: collect the real-time status data change curves of each structural component in the target PCBA board and input them into the component test status evaluation network to determine whether there is an abnormality in the real-time status data change curve. According to the judgment result, retrieve the component abnormality association path, and then obtain the structural component number that is expected to have abnormalities in the future and perform maintenance.

[0012] Furthermore, the acquisition process of the state data change curve includes:

[0013] The target PCBA board is placed in different periodic voltage environments, and each structural component in the target PCBA board is numbered, and several sensors are set for each structural component;

[0014] Setting up a cloud data management platform, and then setting up a data classification cycle and a data upload cycle for each sensor on the cloud data management platform, wherein the data classification cycle is composed of multiple data upload cycles;

[0015] Whenever the data upload cycle ends, each data sensor uploads the collected data to the cloud data management platform, and then the cloud data management platform generates the corresponding state data change curve based on the received data, and marks the structural component number, upload time and environmental voltage value;

[0016] When the data classification cycle ends, the cloud data management platform integrates the status data change curves generated by each target PCBA board in each data upload cycle to generate corresponding historical status change records.

[0017] Furthermore, the process of obtaining the correlation value and the correlation coefficient includes:

[0018] Establish a two-dimensional data association cluster, and generate all state data change curves for each structural component in the first data induction cycle, which are recorded as the first state data change curve;

[0019] A two-dimensional coordinate system is established, and several time nodes are divided on the two-dimensional coordinate system according to the time length of the data upload cycle. When the data upload cycle in the second data induction cycle ends, the cloud data management platform records the state data change curves generated by each sensor as the second state data change curve, and the state data change curves uploaded subsequently are named in the same way;

[0020] The distance values ​​and correlation coefficients between the coordinate points of the first state data change curve and the second state data change curve, the first state data change curve and the third state data change curve, ..., the first state data change curve and the j-th state data change curve at each time node are obtained in sequence, and the distance values ​​between the coordinate points at each time node are counted and recorded as the correlation values ​​α between the corresponding state change curves, where j is a natural number greater than 1.

[0021] Furthermore, the process of mapping each state data change curve to a two-dimensional data association cluster according to the association value and the association coefficient includes:

[0022] When the first state data change curve of each type of data is generated, a data node is set at the center of the two-dimensional data association cluster, which is recorded as the central data node;

[0023] When the cloud data management platform generates the correlation coefficient β and the correlation value α between each subsequently generated state data change curve and the first state data change curve, each data node is mapped to both sides of the central data node according to the size relationship between the correlation coefficient β and 0, and the distance value between each data node and the central data node is set according to the correlation value α.

[0024] Furthermore, the process of establishing the abnormal state data association cluster and the normal state data association cluster includes:

[0025] Set an association distance threshold, record each data node as a central data node in turn, and determine whether the distance between each central data node and the data nodes on the left and right sides is less than or equal to the association distance threshold;

[0026] According to the judgment result, each central data node and its corresponding associated data node are integrated to establish a temporary data associated cluster, and a fusion distance threshold is set. If the distance between the central data nodes of two temporary data associated clusters is less than or equal to the fusion distance threshold, the pair of temporary data associated clusters are merged into a new temporary data associated cluster, and the data node closest to the center position is selected as the new central data node, otherwise no operation is performed;

[0027] Repeat the above temporary data association cluster fusion operation until each newly generated temporary data association cluster cannot be fused, and then record the current temporary data association cluster as the state data association cluster;

[0028] At the end of each data upload cycle, the total number of data nodes in each state data association cluster is counted, and then the state data association cluster with the largest number of data nodes is recorded as the normal state data association cluster, and the remaining state data association clusters are recorded as abnormal state data association clusters.

[0029] Furthermore, the process of establishing the component test status evaluation network includes:

[0030] Establish n component nodes, mark each component node with the corresponding structural component number, and connect each component node according to the distribution of each structural component on the target PCBA board. n represents the total number of structural components.

[0031] According to the number of data types in the status data change curve, a corresponding data status node is set for each component node, and according to the number of abnormal status data association clusters and normal status data association clusters of various types of status data change curves, multiple abnormal data nodes and one normal data node are established, and each abnormal status data association cluster and normal status data association cluster are input into the abnormal data node and the normal data node in turn, so as to obtain a component test status evaluation network.

[0032] Further, the process of establishing an abnormal association path based on the component test status evaluation network includes:

[0033] Set a simulation timeline, set several event time nodes on the simulation timeline, and extract the state data change curve generated by each data upload cycle from each historical state change record, and then judge in sequence according to the time sequence whether each state data change curve of each structural component at the same event time node is within the abnormal state data association cluster;

[0034] If it is determined that there is a state data change curve in the abnormal state data associated cluster, then further determine whether at the previous event time node, the component node connected to the component node corresponding to the state data change curve also has a state data change curve in the abnormal state data associated cluster, until it is determined that there is no state data change curve;

[0035] The abnormal data node judged to exist last time is recorded as the preliminary start abnormal data node, and the abnormal data nodes of the judgment process are connected in sequence until the abnormal data node whose state data change curve is currently judged to exist is in the abnormal state data association cluster, thereby obtaining the abnormal association path;

[0036] If it is determined that there is no state data change curve in the abnormal state data associated cluster, no operation is performed.

[0037] Furthermore, the process of establishing the component exception association path according to the exception association path includes:

[0038] Determine whether the structural component numbers and data types corresponding to the preliminary starting abnormal data nodes of each abnormal associated path generated corresponding to the historical state change record are the same. If they are the same, overlap the preliminary starting abnormal data nodes of the corresponding abnormal associated paths, and again determine whether the structural component numbers and data types corresponding to the abnormal data nodes connected to the preliminary starting abnormal data nodes are the same. Based on the judgment result, overlap the two abnormal data nodes again until the abnormal data nodes in the corresponding abnormal associated paths cannot overlap, thereby obtaining several component abnormal associated paths, and mapping all component abnormal associated path components to the test state evaluation network of the corresponding environmental voltage value.

[0039] An intelligent management system for PCBA board test data, comprising a component data acquisition module, a state analysis module, an evaluation network establishment module and an anomaly detection module;

[0040] The component data acquisition module is used to collect multiple state data change curves of each structural component and integrate all state data change curves to generate historical state change records.

[0041] The state analysis module is used to establish a two-dimensional data association cluster, record the first generated state data change curve as the first state data change curve, and then obtain the association value and association coefficient of the subsequently generated state data change curve and the first state data change curve, and then map each state data change curve to the two-dimensional data association cluster according to the association value and the association coefficient;

[0042] The evaluation network establishment module is used to establish a component test state evaluation network according to the distribution state of the structural components in the target PCBA board and the normal state data association cluster and the abnormal state data association cluster, simulate the abnormal association range of the component test state evaluation network through the historical state change record, generate the component abnormal association path according to the simulation result, and map the component abnormal association path to the component test state evaluation network;

[0043] The anomaly detection module is used to input the real-time status data change curves of each structural component into the component test status evaluation network, determine whether there is an anomaly in the real-time status data change curve, retrieve the component anomaly association path according to the judgment result, and then obtain the structural component number that is expected to have an abnormality in the future and perform maintenance.

[0044] Compared with the prior art, the present invention has the following beneficial effects:

[0045] The present invention establishes a component test state evaluation network according to the distribution state of structural components in a target PCBA board and a normal state data association cluster and an abnormal state data association cluster, and then simulates the abnormal association range of the component test state evaluation network according to the historical state change record, generates a component abnormal association path according to the simulation result, collects various real-time state data change curves of each structural component in the target PCBA board and inputs them into the component test state evaluation network, judges whether there is an abnormality in the real-time state data change curve, retrieves the component abnormal association path according to the judgment result, and then obtains the structural component number that is expected to have an abnormality in the future and performs maintenance, thereby ensuring that the value of the test data is fully mined while improving the flexibility and efficiency of detecting the abnormal state of the PCBA according to the test data analysis. BRIEF DESCRIPTION OF THE DRAWINGS

[0046] In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, the drawings required for use in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments recorded in the present invention.

[0047] Figure 1 The figure is a flow chart of the method of the present invention. DETAILED DESCRIPTION

[0048] To make the purpose, technical solution and advantages of the present invention clearer, the technical solution of the present invention will be described in detail below. Obviously, the described embodiments are only part of the embodiments of the present invention, rather than all the embodiments. Based on the embodiments of the present invention, all other implementation methods obtained by ordinary technicians in this field without making creative work are based on the scope of protection of the present invention. Example 1

[0049] like Figure 1 As shown, a PCBA board test data intelligent management method includes the following steps:

[0050] Step S1, setting several types of sensors for each structural component in the target PCBA board, and placing the target PCBA board in a periodic voltage environment, thereby collecting multiple state data change curves of each structural component, and integrating all state data change curves to generate a historical state change record;

[0051] Step S2, establishing a two-dimensional data association cluster, recording the first generated state data change curve as the first state data change curve, and then obtaining the association value and the association coefficient between the subsequently generated state data change curve and the first state data change curve, and then mapping each state data change curve into the two-dimensional data association cluster according to the association value and the association coefficient, and dividing the two-dimensional data association cluster into a normal state data association cluster and an abnormal state data association cluster;

[0052] Step S3, establishing a component test state evaluation network according to the distribution state of the structural components in the target PCBA board and the normal state data association cluster and the abnormal state data association cluster, and then simulating the abnormal association range of the component test state evaluation network according to the historical state change record, generating a number of abnormal association paths according to the simulation results, cross-overlapping the abnormal association paths to obtain component abnormal association paths, and mapping the component abnormal association paths to the component test state evaluation network;

[0053] Step S4: collect the real-time status data change curves of each structural component in the target PCBA board and input them into the component test status evaluation network to determine whether there is an abnormality in the real-time status data change curve. According to the judgment result, retrieve the component abnormality association path, and then obtain the structural component number that is expected to have abnormalities in the future and perform maintenance. Example 2

[0054] This embodiment is a further limitation of Embodiment 1, and Step S1 is implemented by the following process:

[0055] Place N target PCBA boards of the same type in different periodic voltage environments, and set numbers a1, a2, a3, ..., a for each structural component in the target PCBA board. n , a number of sensors are set for each structural component, where N and n are natural numbers greater than 0;

[0056] Sensors on the structural components, such as temperature sensors, voltage sensors, power sensors, etc.;

[0057] The voltage value of the periodic voltage environment changes in large cycles and small cycles, wherein the large cycle is composed of k small cycles, and the small cycles of each large cycle are the same, and the voltage in each small cycle changes periodically, wherein k is a natural number greater than 0;

[0058] A cloud data management platform is set, and the cloud data management platform sets a data classification cycle and a data upload cycle for each sensor, wherein the time length of the data classification cycle is equal to the time length of a large cycle of the periodic voltage environment, and the time length of the data upload cycle is equal to the time length of a small cycle of the periodic voltage environment;

[0059] Whenever the data upload cycle ends, each data sensor uploads the collected data to the cloud data management platform, and then the cloud data management platform generates the corresponding state data change curve based on the received data, and marks the structural component number, upload time and environmental voltage value;

[0060] When the data classification cycle ends, the cloud data management platform integrates the status data change curves generated by each target PCBA board in each data upload cycle to generate corresponding historical status change records. Example 3

[0061] This embodiment is a further limitation of Embodiment 1, and Step S2 is implemented by the following process:

[0062] The cloud data management platform establishes a number of two-dimensional data association clusters according to the number of state data change curves of each structural component, and records each state data change curve of each structural component in the historical state change record generated in the first data induction cycle as a first state data change curve;

[0063] A two-dimensional coordinate system is established, and several time nodes are divided on the two-dimensional coordinate system according to the time length of the data upload cycle. When the data upload cycle in the second data induction cycle ends, the cloud data management platform records the state data change curves generated by each sensor as the second state data change curve, and the state data change curves uploaded subsequently are named in the same way;

[0064] Then, the second state data change curve and the first state data change curve with the structural component number, the environmental voltage value and the same data type are mapped on a two-dimensional coordinate system;

[0065] The distance values ​​between the coordinate points of the first state data change curve and the second state data change curve, the first state data change curve and the third state data change curve, ..., the first state data change curve and the j-th state data change curve at each time node are obtained in sequence, and the distance values ​​between the coordinate points at each time node are counted and recorded as the correlation value α between the corresponding state change curves, wherein the calculation formula of the correlation value α is: ;

[0066] in represents the correlation value α between the first state data change curve and the j-th state data change curve, , , as well as Respectively represent the coordinate values ​​of the first state data change curve and the j-th state data change curve at the m-th time node, where M and j are natural numbers greater than 1;

[0067] At the same time, the correlation coefficient β between the first state data change curve and the second state data change curve, the first state data change curve and the third state data change curve, ..., the first state data change curve and the j-th state data change curve is obtained, wherein the calculation formula of the correlation coefficient β is:

[0068] in represents the correlation coefficient β between the first state data change curve and the j-th state data change curve, and They represent the curve average values ​​of the first state data change curve and the j-th state data change curve, respectively, wherein the calculation formula of the curve average value is: y m Indicates the horizontal coordinate value of the coordinate point of the state data change curve at the mth time node.

[0069] Furthermore, whenever the cloud data management platform generates a state data change curve, a data node is set in each two-dimensional data association cluster, and when the first state data change curve of each type of data is generated, a data node is set at the center of the two-dimensional data association cluster, recorded as the central data node, and the first state data change curve is stored in the data node;

[0070] When the cloud data management platform subsequently generates a second state data change curve, a third state data change curve, ..., a j-th state data change curve, according to the correlation coefficient β and the correlation value α between each subsequently generated state data change curve and the first state data change curve, and according to the magnitude relationship between the correlation coefficient β and 0, each data node is mapped to both sides of the central data node, and the distance value between each data node and the central data node is set according to the correlation value α;

[0071] If the correlation coefficient β is less than 0, the data node is mapped to the left of the central data node;

[0072] If the correlation coefficient β is equal to 0, the data node is mapped to coincide with the central data node;

[0073] If the correlation coefficient β is greater than 0, the data node is mapped to the right of the central data node;

[0074] Set an association distance threshold, and then record each data node as a central data node in turn, and determine whether the distance between each central data node and the data nodes on the left and right sides is less than or equal to the association distance threshold;

[0075] If the distance is greater than the associated distance threshold, no operation will be performed. Otherwise, the data node is recorded as the associated data node of the corresponding central data node, and then each central data node and its corresponding associated data node are integrated to establish a temporary data association cluster.

[0076] A fusion distance threshold is set, which is less than half of the association distance threshold. If the distance between the central data nodes of two temporary data association clusters is less than or equal to the fusion distance threshold, the pair of temporary data association clusters are merged into a new temporary data association cluster, and the data node closest to the center position is selected as the new central data node, otherwise no operation is performed;

[0077] Repeat the above temporary data association cluster fusion operation until each newly generated temporary data association cluster cannot be fused, and then record the current temporary data association cluster as the state data association cluster;

[0078] At the end of each data upload cycle, the total number of data nodes in each state data association cluster is counted, and then the state data association cluster with the largest number of data nodes is recorded as the normal state data association cluster, and the remaining state data association clusters are recorded as abnormal state data association clusters;

[0079] The normal state data association clusters of different data upload cycles are overlapped and mapped, and the fusion distance threshold is used as the radius. Then, the location with the largest data node density is selected from each overlapping mapped normal state data association cluster to establish a new normal state data association cluster, and the process of establishing a new normal state data association cluster is used to establish a new abnormal state data association cluster. Example 4

[0080] This embodiment is a further limitation of Embodiment 1, and step S3 is implemented by the following process:

[0081] According to the number of structural components in the target PCBA board, n component nodes are established, and the corresponding structural component number is marked on each component node, and then each component node is connected according to the distribution status of each structural component on the target PCBA board;

[0082] A corresponding data status node is set for each component node according to the number of data types of the status data change curve, and a plurality of abnormal data nodes and one normal data node are established according to the number of abnormal status data association clusters and normal status data association clusters of various types of status data change curves. Each abnormal status data association cluster and normal status data association cluster are input into the abnormal data node and the normal data node in sequence, so as to obtain a component test status evaluation network under various voltage environments, and annotate the component test status evaluation network with the environmental voltage value.

[0083] Furthermore, according to each data upload cycle, the historical state change record is generated to simulate the abnormal correlation range of the component test state evaluation network. The abnormal correlation range simulation process includes:

[0084] Set a simulation timeline, set several event time nodes on the simulation timeline, and extract the state data change curve generated in each data upload cycle from each historical state change record, and then determine in sequence according to the time sequence which abnormal state data association cluster and normal state data association cluster each state data change curve of each structural component belongs to at the same event time node;

[0085] If it is determined that there is a state data change curve in the abnormal state data associated cluster, it is further determined whether, at the previous event time node, the component node connected to the component node corresponding to the state data change curve also has a state data change curve in the abnormal state data associated cluster;

[0086] If it does not exist, the abnormal data node corresponding to the abnormal state data association cluster is recorded as the preliminary starting abnormal data node, otherwise it is further determined that the component node associated with the component node connected to the component node corresponding to the state data change curve also has a state data change curve in the abnormal state data association cluster at the previous time node, until it is determined that it does not exist;

[0087] Starting from the abnormal data node where the last judgment of the existence of the state data change curve is in the abnormal state data association cluster as the preliminary starting abnormal data node, the abnormal data nodes where the judgment of the existence of the state data change curve is in the abnormal state data association cluster are connected in sequence until the abnormal data node where the current judgment of the existence of the state data change curve is in the abnormal state data association cluster is obtained, thereby obtaining the abnormal association path;

[0088] If it is determined that there is no state data change curve in the abnormal state data associated cluster, no operation is performed;

[0089] Determine whether the structural component numbers and data types corresponding to the preliminary starting abnormal data nodes of each abnormal associated path generated corresponding to the historical state change record are the same. If they are the same, overlap the preliminary starting abnormal data nodes of the corresponding abnormal associated paths, and again determine whether the structural component numbers and data types corresponding to the abnormal data nodes connected to the preliminary starting abnormal data nodes are the same. Based on the judgment result, overlap the two abnormal data nodes again until the abnormal data nodes in the corresponding abnormal associated paths cannot overlap, thereby obtaining several component abnormal associated paths, and mapping all component abnormal associated path components to the test state evaluation network of the corresponding environmental voltage value. Example 5

[0090] This embodiment is a further limitation of Embodiment 1, and step S4 is implemented by the following process:

[0091] Collect the real-time status data change curves of each structural component in the target PCBA board, and call the corresponding test status evaluation network according to the environmental voltage value of the target PCBA board;

[0092] Through the test status evaluation network, it is determined whether there are any abnormalities in the real-time status data change curves. If there are any abnormalities, the corresponding component abnormality association path is retrieved according to the abnormal data node corresponding to the real-time status data change curve;

[0093] The numbers of structural components that are expected to have subsequent abnormalities are obtained according to the component abnormality association path, and then component abnormality warning prompts are generated and sent to the staff. The staff maintains the corresponding structural components according to the component abnormality warning prompts.

[0094] The present invention also discloses an intelligent management system for intelligent management of PCBA board test data, including a component data acquisition module, a state analysis module, an evaluation network establishment module and an abnormality detection module;

[0095] The component data acquisition module is used to collect multiple state data change curves of each structural component and integrate all state data change curves to generate historical state change records.

[0096] The state analysis module is used to establish a two-dimensional data association cluster, record the first generated state data change curve as the first state data change curve, and then obtain the association value and association coefficient of the subsequently generated state data change curve and the first state data change curve, and then map each state data change curve to the two-dimensional data association cluster according to the association value and the association coefficient;

[0097] The evaluation network establishment module is used to establish a component test state evaluation network according to the distribution state of the structural components in the target PCBA board and the normal state data association cluster and the abnormal state data association cluster, simulate the abnormal association range of the component test state evaluation network through the historical state change record, generate the component abnormal association path according to the simulation result, and map the component abnormal association path to the component test state evaluation network;

[0098] The anomaly detection module is used to input the real-time status data change curves of each structural component into the component test status evaluation network, determine whether there is an anomaly in the real-time status data change curve, retrieve the component anomaly association path according to the judgment result, and then obtain the structural component number that is expected to have an abnormality in the future and perform maintenance.

[0099] The above embodiments are only used to illustrate the technical method of the present invention rather than to limit it. Although the present invention has been described in detail with reference to the preferred embodiments, those skilled in the art should understand that the technical method of the present invention may be modified or replaced by equivalents without departing from the spirit and scope of the technical method of the present invention.

Claims

1. A PCBA board test data intelligent management method, characterized in that: The following steps are involved: Step S1, setting several types of sensors for each structural component in the target PCBA board, and placing the target PCBA board in a periodic voltage environment, thereby collecting multiple state data change curves of each structural component, and integrating all state data change curves to generate a historical state change record; Step S2, establishing a two-dimensional data association cluster, recording the first generated state data change curve as the first state data change curve, and then obtaining the association value and the association coefficient between the subsequently generated state data change curve and the first state data change curve, and then mapping each state data change curve into the two-dimensional data association cluster according to the association value and the association coefficient, and dividing the two-dimensional data association cluster into a normal state data association cluster and an abnormal state data association cluster; Step S3, establishing a component test state evaluation network according to the distribution state of the structural components in the target PCBA board and the normal state data association cluster and the abnormal state data association cluster, and then simulating the abnormal association range of the component test state evaluation network according to the historical state change record, generating a number of abnormal association paths according to the simulation results, cross-overlapping the abnormal association paths to obtain component abnormal association paths, and mapping the component abnormal association paths to the component test state evaluation network; Step S4: collect the real-time status data change curves of each structural component in the target PCBA board and input them into the component test status evaluation network to determine whether there is an abnormality in the real-time status data change curve. According to the judgment result, retrieve the component abnormality association path, and then obtain the structural component number that is expected to have abnormalities in the future and perform maintenance.

2. A PCBA board test data intelligent management method according to claim 1, characterized in that: The acquisition process of the state data change curve includes: The target PCBA board is placed in different periodic voltage environments, and each structural component in the target PCBA board is numbered, and several sensors are set for each structural component; Setting up a cloud data management platform, and then setting up a data classification cycle for each sensor on the cloud data management platform, wherein the data classification cycle is composed of multiple data upload cycles; Whenever the data upload cycle ends, each data sensor uploads the collected data to the cloud data management platform, and then the cloud data management platform generates the corresponding state data change curve based on the received data, and marks the structural component number, upload time and environmental voltage value; When the data classification cycle ends, the cloud data management platform integrates the status data change curves generated by each data upload cycle to generate a historical status change record.

3. A PCBA board test data intelligent management method according to claim 2, characterized in that: The process of obtaining the correlation value and the correlation coefficient includes: A two-dimensional data association cluster is established, and the state data change curves generated by each structural component in the first data induction cycle are all recorded as the first state data change curves; A two-dimensional coordinate system is established, and several time nodes are divided on the two-dimensional coordinate system according to the time length of the data upload cycle. When the data upload cycle in the second data induction cycle ends, the cloud data management platform records the state data change curves generated by each sensor as the second state data change curve, and the state data change curves uploaded subsequently are named in the same way; The distance values ​​and correlation coefficients between the coordinate points of the first state data change curve and the second state data change curve, the first state data change curve and the third state data change curve, ..., the first state data change curve and the j-th state data change curve at each time node are obtained in sequence, and the distance values ​​between the coordinate points at each time node are counted and recorded as the correlation values ​​α between the corresponding state change curves, where j is a natural number greater than 1.

4. A PCBA board test data intelligent management method according to claim 3, characterized in that: The process of mapping each state data change curve to a two-dimensional data association cluster according to the association value and the association coefficient includes: When the first state data change curve of each type of data is generated, a data node is set at the center of the two-dimensional data association cluster, which is recorded as the central data node; When the cloud data management platform generates the correlation coefficient β and the correlation value α between each subsequently generated state data change curve and the first state data change curve, each data node is mapped to both sides of the central data node according to the size relationship between the correlation coefficient β and 0, and the distance value between each data node and the central data node is set according to the correlation value α.

5. A PCBA board test data intelligent management method according to claim 4, characterized in that: The process of establishing the abnormal state data association cluster and the normal state data association cluster includes: Set an association distance threshold, record each data node as a central data node in turn, and determine whether the distance between each central data node and the data nodes on the left and right sides is less than or equal to the association distance threshold; According to the judgment result, each central data node and its corresponding associated data node are integrated to establish a temporary data associated cluster, and a fusion distance threshold is set. If the distance between the central data nodes of two temporary data associated clusters is less than or equal to the fusion distance threshold, the corresponding two temporary data associated clusters are merged into a new temporary data associated cluster, and the data node closest to the center position is selected as the new central data node, otherwise no operation is performed; Repeat the temporary data association cluster fusion operation until each newly generated temporary data association cluster cannot be fused, and then record the current temporary data association cluster as the state data association cluster; At the end of each data upload cycle, the total number of data nodes in each state data association cluster is counted, and then the state data association cluster with the largest number of data nodes is recorded as the normal state data association cluster, and the remaining state data association clusters are recorded as abnormal state data association clusters.

6. A PCBA board test data intelligent management method according to claim 5, characterized in that: The process of establishing the component test status evaluation network includes: Establish n component nodes, mark each component node with the corresponding structural component number, and connect each component node according to the distribution of each structural component on the target PCBA board. n represents the total number of structural components. According to the number of data types in the status data change curve, a corresponding data status node is set for each component node, and according to the number of abnormal status data association clusters and normal status data association clusters of various types of status data change curves, multiple abnormal data nodes and one normal data node are established, and each abnormal status data association cluster and normal status data association cluster are input into the abnormal data node and the normal data node in turn, so as to obtain a component test status evaluation network.

7. A PCBA board test data intelligent management method according to claim 6, characterized in that: The process of establishing an abnormal correlation path based on the component test status evaluation network includes: Set a simulation timeline, set several event time nodes on the simulation timeline, and extract the state data change curve generated by each data upload cycle from each historical state change record, and then judge in sequence according to the time sequence whether each state data change curve of each structural component at the same event time node is within the abnormal state data association cluster; If it is determined that there is a state data change curve in the abnormal state data associated cluster, then further determine whether at the previous event time node, the component node connected to the component node corresponding to the state data change curve also has a state data change curve in the abnormal state data associated cluster, until it is determined that there is no state data change curve; The abnormal data node judged to exist last time is recorded as the preliminary start abnormal data node, and the abnormal data nodes of the judgment process are connected in sequence until the abnormal data node whose state data change curve is currently judged to exist is in the abnormal state data association cluster, thereby obtaining the abnormal association path; If it is determined that there is no state data change curve in the abnormal state data associated cluster, no operation is performed.

8. A PCBA board test data intelligent management method according to claim 7, characterized in that: The process of establishing a component exception association path according to the exception association path includes: Determine whether the structural component numbers and data types corresponding to the preliminary starting abnormal data nodes of each abnormal associated path generated corresponding to the historical state change record are the same. If they are the same, overlap the preliminary starting abnormal data nodes of the corresponding abnormal associated paths, and again determine whether the structural component numbers and data types corresponding to the abnormal data nodes connected to the preliminary starting abnormal data nodes are the same. Based on the judgment result, overlap the two abnormal data nodes again until the abnormal data nodes in the corresponding abnormal associated paths cannot overlap, thereby obtaining several component abnormal associated paths, and mapping all component abnormal associated path components to the test state evaluation network of the corresponding environmental voltage value.

9. An intelligent management system for PCBA board test data, used to implement an intelligent management method for PCBA board test data according to any one of claims 1 to 8, characterized in that: It includes component data acquisition module, status analysis module, evaluation network establishment module and anomaly detection module; The component data acquisition module is used to collect multiple state data change curves of each structural component, and integrate all state data change curves to generate historical state change records; The state analysis module is used to establish a two-dimensional data association cluster, record the first generated state data change curve as the first state data change curve, and then obtain the association value and association coefficient of the subsequently generated state data change curve and the first state data change curve, and then map each state data change curve to the two-dimensional data association cluster according to the association value and the association coefficient; The evaluation network establishment module is used to establish a component test state evaluation network according to the distribution state of the structural components in the target PCBA board and the normal state data association cluster and the abnormal state data association cluster, simulate the abnormal association range of the component test state evaluation network through the historical state change record, generate the component abnormal association path according to the simulation result, and map the component abnormal association path to the component test state evaluation network; The anomaly detection module is used to input the real-time status data change curves of each structural component into the component test status evaluation network, determine whether there is an anomaly in the real-time status data change curve, retrieve the component anomaly association path according to the judgment result, and then obtain the structural component number that is expected to have an abnormality in the future and perform maintenance.

Images