A multi-permission adaptive automated testing process for circuit boards
By implementing an automated circuit board testing process with multi-permission adaptation, the problems of test threshold failure and unreasonable permission management in flexible production of automotive-grade circuit boards with multiple varieties and small batches have been solved. This process achieves dynamic correlation and optimization between test standards and actual conditions, reduces false test rate, and ensures a balance between quality and efficiency.
Patent Information
- Application Number
- CN202611242721.2
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Filing Date
- 2026-08-17
- Publication Date
- 2026-11-17
- Estimated Expiration
- 2046-08-17
AI Technical Summary
Existing automated testing processes suffer from quality and efficiency issues in scenarios involving flexible production of automotive-grade/high-reliability circuit boards in small batches and with frequent material batch changes, such as test threshold failure, high false test rate, and unreasonable access control.
The process of introducing automated circuit board testing with multi-permission adaptation is adopted. By acquiring operator identity authentication information and production mode recognition, combined with dynamic deviation judgment and differentiated threshold adaptation strategy, including physical state compensation, batch tolerance offset and cross-model inheritance and risk hedging, the dynamic correlation and optimization between test standards and actual physical state can be achieved.
This significantly reduced the false test rate during the initial stages of line/batch changes, ensured the parameter consistency of the automotive-grade quality system, achieved a dynamic balance between quality and efficiency under flexible production, and prevented interference from human factors and systematic omissions.
Smart Images

Figure CN122815145B_ABST
Abstract
Description
Technical Field
[0001] This invention belongs to the field of circuit board testing technology, specifically relating to an automated circuit board testing process with multi-permission adaptation. Background Technology
[0002] In the circuit board manufacturing process, automated testing (such as ICT online testing and FCT functional testing) is a core process for ensuring product quality. Existing automated testing processes typically use preset static test thresholds to judge the circuit boards and manage operator permissions through role-based access control. When an anomaly occurs during testing, the system usually locks the equipment directly, requiring an engineer with advanced privileges to log in to the system, directly modify the static threshold parameters in the test recipe, and then reissue it to the machine for execution.
[0003] However, in scenarios involving flexible production of automotive-grade / high-reliability circuit boards in small batches with frequent material batch changes, existing technologies suffer from the following serious drawbacks: First, frequent line changes and drastic variations in test probe wear conditions, coupled with inherent batch tolerance differences between different batches of components, cause fixed thresholds to easily fail at the initial stage of line change or batch change, resulting in numerous false tests. Second, existing access control is statically intercepted, leading engineers to frequently modify test formulas globally after line changes to quickly restore production. This not only undermines the stringent quality requirements of automotive-grade standards for parameter consistency but also lacks correlation with the current physical state of the probes, easily introducing systemic missed test risks. Finally, existing processes lack a permission and threshold linkage mechanism for the specific action of line / batch changes, resulting in a severe disconnect between test standards and the current actual physical and material conditions, failing to achieve a dynamic balance between quality and efficiency in flexible production. Based on these findings, a multi-permission adaptive automated testing process for circuit boards is proposed. Summary of the Invention
[0004] To address the aforementioned problems in the existing technology, this invention provides a multi-permission adaptive automated testing process for circuit boards.
[0005] The objective of this invention can be achieved through the following technical solutions: An automated testing process for circuit boards with multi-permission adaptation includes the following steps: S1: Obtain the identity authentication information of the current operator and match the permission level, and identify the current production mode. The production mode includes at least the normal production mode, the line changeover mode and the material batch switching mode. S2: Call the basic safety threshold to perform an automated initial test on the circuit board and obtain test data; if it is determined to be unqualified, calculate the dynamic deviation between the test data and the basic safety threshold. When the dynamic deviation is less than the preset edge threshold, mark it as an edge unqualified product and suspend the test. S3: Based on the current operator's permission level and the identified production mode, trigger the corresponding threshold adaptation and retesting strategy: When in line-switching mode and with operator privileges, the physical state compensation strategy is triggered, and the system automatically obtains probe state parameters to generate a state compensation threshold for retesting. When in material batch switching mode and with engineer privileges, the batch tolerance offset strategy is triggered, and the system allows the batch offset threshold to be called within the safety boundary to adjust the test window and retest. When in line-switching mode and with administrator privileges, cross-model inheritance and risk hedging strategies are triggered, similar model thresholds are called, and associated test items are forcibly added for retesting. S4: Automatically sort the circuit boards according to the retest results, record the test data and permission operation log, and dynamically converge the state compensation threshold or batch offset threshold based on the actual test data distribution after line change or batch change.
[0006] As a further aspect of the present invention, the specific execution process of the physical state compensation strategy in step S3 is as follows: The system automatically collects the cumulative number of probe insertions, real-time contact resistance change rate, and ambient temperature and humidity data of the current test fixture; Substitute the collected physical state parameters into the preset contact impedance compensation model to calculate the threshold fine-tuning amount for the current probe wear state. The state compensation threshold is generated by combining the basic safety threshold with the threshold fine-tuning amount.
[0007] As a further aspect of the present invention, the specific execution process of the batch tolerance offset strategy in step S3 is as follows: The system obtains historical tolerance distribution data or first piece test data of the current new batch of materials and calculates the offset of characteristic parameters of the current batch of materials; Guided by the CPK distribution map displayed by the system, the engineer moves or scales the test window as a whole within the safety boundary defined by the basic safety threshold to generate the batch offset threshold. The system performs hard constraint verification on the batch offset threshold input by the engineer. If it exceeds the safety boundary, the application is rejected and an automotive-grade quality alarm is triggered.
[0008] As a further aspect of the present invention, in step S3, the specific execution process of the cross-model inheritance and risk hedging strategy is as follows: when a new product line lacks historical test data, the administrator calls the test threshold of the historical model with the highest similarity as the initial threshold; the system forcibly triggers the risk hedging mechanism: while inheriting the initial threshold, it automatically increases the insulation withstand voltage test time, increases the key signal sampling rate, or increases the number of extreme high and low temperature cycle tests of FCT in the test formula.
[0009] As a further aspect of the present invention, in step S4, the specific process of dynamic convergence is as follows: during the testing of the first N circuit boards after line or batch change, the system statistically analyzes the distribution characteristics of edge defective products in real time; if the edge defective products are concentrated on one side, the system automatically fine-tunes the state compensation threshold or batch offset threshold according to a preset step size until the number of edge defective products drops below the preset threshold, thereby completing the dynamic convergence of the test window, and solidifying the converged threshold as the formal test threshold for the batch or model.
[0010] As a further aspect of the present invention, after step S4, a dynamic evaluation step of permission credibility based on line-switching performance is also included: Track the frequency with which a specific engineer or administrator invokes dynamic thresholds after a line change or batch change, as well as the actual defect rate of the corresponding batch of circuit boards in subsequent processes; if the administrator's cross-model inheritance policy is frequently triggered after a line change, or the defect rate in the later stages exceeds the standard, the corresponding personnel's authority and credit score will be automatically reduced, and the batch tolerance offset safety boundary that can be adjusted during the next line change will be narrowed.
[0011] As a further aspect of the present invention, in step S2, the specific method for calculating the dynamic deviation is as follows: obtain the actual measured value of the test data, combine it with the nominal tolerance zone of the current material batch, calculate the absolute difference between the actual measured value and the tolerance zone boundary of the current batch; divide the absolute difference by the total width of the tolerance zone of the current batch to obtain the normalized dynamic deviation, which is used to eliminate the interference of inherent tolerance differences between different batches of materials on edge determination.
[0012] As a further aspect of the present invention, before step S1 and step S2, a process of issuing a line change recipe and hierarchical approval is included: when a line change mode is identified, the system automatically matches the corresponding approval flow according to the product level of the target circuit board.
[0013] As a further aspect of the present invention, step S3 also includes a dynamic permission downgrade mechanism in case of line change abnormality: In line change mode, if the number of consecutive edge defective products exceeds the preset continuous alarm threshold, the system automatically triggers permission downgrade; the current engineer's batch tolerance offset permission is revoked, the equipment is forcibly locked, and a high-level alarm is sent to request the administrator to intervene to perform physical calibration or in-depth investigation of the equipment.
[0014] The beneficial effects of this invention are as follows: By introducing a three-level judgment mechanism for edge-conforming products, edge-conforming products that are mistakenly judged as non-conforming under the traditional binary judgment are separated. Combined with subsequent differentiated retesting strategies, the false test rate in the initial stage of line / batch change is significantly reduced. It changes the traditional one-size-fits-all static access control mode. Different access levels trigger different threshold adjustment strategies under different production modes, ensuring both the consistency of parameters required by the automotive-grade quality system and providing reasonable operational space for different roles within their responsibilities. Through a physical state compensation strategy, the probe wear state is automatically associated; through a batch tolerance offset strategy, the inherent characteristics of the current batch of materials are associated; and through dynamic convergence, the test standard continuously approaches the optimal value, achieving a dynamic balance between quality and efficiency under flexible production. Attached Figure Description
[0015] 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.
[0016] Figure 1 This is a flowchart of the steps of the present invention. Detailed Implementation
[0017] The technical solutions in the embodiments of the present invention will be clearly and completely described below. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.
[0018] Please see Figure 1 An automated testing process for circuit boards with multi-permission adaptation includes the following steps: S1: Obtain the identity authentication information of the current operator and match the permission level, and identify the current production mode. The production mode includes at least the normal production mode, the line changeover mode and the material batch switching mode. S2: Call the basic safety threshold to perform an automated initial test on the circuit board and obtain test data; if it is determined to be unqualified, calculate the dynamic deviation between the test data and the basic safety threshold. When the dynamic deviation is less than the preset edge threshold, mark it as an edge unqualified product and suspend the test. S3: Based on the current operator's permission level and the identified production mode, trigger the corresponding threshold adaptation and retesting strategy: When in line-switching mode and with operator privileges, the physical state compensation strategy is triggered, and the system automatically obtains probe state parameters to generate a state compensation threshold for retesting. When in material batch switching mode and with engineer privileges, the batch tolerance offset strategy is triggered, and the system allows the batch offset threshold to be called within the safety boundary to adjust the test window and retest. When in line-switching mode and with administrator privileges, cross-model inheritance and risk hedging strategies are triggered, similar model thresholds are called, and associated test items are forcibly added for retesting. S4: Automatically sort the circuit boards according to the retest results, record the test data and permission operation log, and dynamically converge the state compensation threshold or batch offset threshold based on the actual test data distribution after line change or batch change.
[0019] Traditional solutions separate access control from threshold management, using the same static threshold adjustment method regardless of production mode or access level. This invention, in step S1, first performs dual identification: identifying both the operator's access level (e.g., operator, engineer, administrator) and the current production mode (e.g., normal production, line changeover, material batch switching). This dual identification provides a decision-making basis for subsequent differentiated strategy triggering.
[0020] Step S2 introduces an intermediate state determination mechanism for edge-defective products. Traditional methods use a binary determination (pass / fail), classifying a product as defective and locking the equipment once it exceeds a static threshold. This invention calculates the dynamic deviation between test data and a basic safety threshold, and sets an edge threshold, separately marking products with smaller deviations as edge-defective products. This three-level determination avoids indiscriminate misjudgment and provides accurate input for subsequent differentiated retesting.
[0021] Step S3 triggers different threshold adaptation strategies based on the combination of permissions and modes. Specifically, Operator + Line Switching Mode → Physical State Compensation Strategy: The system automatically generates compensation thresholds, which operators cannot modify. Engineer + Material Batch Switching Mode → Batch Tolerance Offset Strategy: Engineers adjust within safety boundaries. Administrator + Line Switching Mode → Cross-Model Inheritance and Risk Hedging Strategy: Historical thresholds are invoked and test items are forcibly added. This precise matching of permissions, modes, and strategies ensures flexibility in different scenarios while preventing quality risks caused by arbitrary threshold modifications through permission constraints.
[0022] By introducing a three-level judgment mechanism for edge-conforming products, edge-conforming products that are mistakenly judged as non-conforming under the traditional binary judgment method are separated. Combined with subsequent differentiated retesting strategies, the false test rate at the beginning of line or batch changes is significantly reduced. This changes the traditional one-size-fits-all static access control model. Different access levels trigger different threshold adjustment strategies under different production modes, ensuring the consistency of parameters required by the automotive-grade quality system while giving different roles reasonable operational space within their responsibilities. By automatically associating probe wear status through a physical state compensation strategy and associating inherent characteristics of the current batch of materials through a batch tolerance offset strategy, dynamic convergence continuously brings the test standard closer to the optimal value, achieving a dynamic balance between quality and efficiency under flexible production.
[0023] As a further aspect of the present invention, the specific execution process of the physical state compensation strategy in step S3 is as follows: The system automatically collects the cumulative number of probe insertions, real-time contact resistance change rate, and ambient temperature and humidity data of the current test fixture; Substitute the collected physical state parameters into the preset contact impedance compensation model to calculate the threshold fine-tuning amount for the current probe wear state. The state compensation threshold is generated by combining the basic safety threshold with the threshold fine-tuning amount.
[0024] In flexible manufacturing scenarios, frequent line changes lead to drastic changes in the wear state of test probes. During testing, the probes repeatedly contact the circuit board contacts, causing wear on the internal springs, contact tips, and probe tubes, resulting in changes in contact resistance. Factors such as contact pressure and probe geometry also significantly affect contact resistance. In traditional solutions, engineers often ignore the current physical state of the probes and directly modify the global threshold, leading to a disconnect between threshold adjustments and actual conditions.
[0025] The physical state compensation strategy of this invention achieves objectivity and automation of threshold adjustment through the automatic acquisition and modeling calculation of multi-dimensional physical parameters. Specifically, the system collects three key physical parameters in real time: the cumulative number of probe punctures, reflecting the cumulative degree of mechanical wear; the real-time contact resistance change rate, reflecting the current electrical contact quality; and environmental temperature and humidity data, reflecting the impact of environmental factors on contact impedance. These three parameters are substituted into a preset contact impedance compensation model, which is established based on historical experimental data and can quantitatively calculate the threshold fine-tuning amount required under the current physical state. Finally, the basic safety threshold and the fine-tuning amount are superimposed to generate the state compensation threshold.
[0026] Crucially, under this strategy, operators can only trigger retest commands and have no right to manually input or modify the specific value of the status compensation threshold. This design fundamentally prevents human factors from interfering with the objectivity of the threshold, ensuring the scientific nature and traceability of threshold adjustments.
[0027] As a further aspect of the present invention, the specific execution process of the batch tolerance offset strategy in step S3 is as follows: The system obtains historical tolerance distribution data or first piece test data of the current new batch of materials and calculates the offset of characteristic parameters of the current batch of materials; Guided by the CPK distribution map displayed by the system, the engineer moves or scales the test window as a whole within the safety boundary defined by the basic safety threshold to generate the batch offset threshold. The system performs hard constraint verification on the batch offset threshold input by the engineer. If it exceeds the safety boundary, the application is rejected and an automotive-grade quality alarm is triggered.
[0028] In scenarios involving frequent batch changes in materials, components from different batches exhibit inherent batch tolerance differences. HMLV production often requires sourcing components from multiple suppliers, resulting in variations in electrical characteristics such as capacitance tolerance and resistance values between different batches. In traditional solutions, engineers, in an effort to quickly resume production, often relax global thresholds. While this may temporarily pass the tests for the current batch, it could lead to systemic missed tests due to excessively low testing standards.
[0029] The batch tolerance offset strategy of this invention, while granting engineers adjustment authority, ensures the safety and scientific nature of the adjustments through multiple constraints. First, the system acquires historical tolerance distribution data or first-piece test data for the current new batch of materials and calculates the characteristic parameter offset, providing a data foundation for engineers' decisions. Second, the system guides engineers by displaying a CPK (Process Capability Index) distribution chart. CPK is a measure of a process's ability to consistently produce products that meet specifications. Through the visual guidance of the CPK distribution chart, engineers can intuitively judge the relationship between the distribution characteristics of the current batch of materials and the specification limits, thereby making scientific adjustment decisions.
[0030] Guided by the CPK distribution chart, engineers shift or scale the test window within the safety boundaries defined by the basic safety threshold. Shifting addresses shifts in the batch mean, while scaling addresses changes in batch dispersion. Finally, the system performs a hard constraint check on the batch offset threshold input by the engineer: if it exceeds the safety boundary, the application is rejected and an automotive-grade quality alarm is triggered. This hard constraint mechanism ensures that any threshold adjustment will not exceed the minimum quality baseline for automotive-grade products.
[0031] As a further aspect of the present invention, in step S3, the specific execution process of the cross-model inheritance and risk hedging strategy is as follows: when a new product line lacks historical test data, the administrator calls the test threshold of the historical model with the highest similarity as the initial threshold; the system forcibly triggers the risk hedging mechanism: while inheriting the initial threshold, it automatically increases the insulation withstand voltage test time, increases the key signal sampling rate, or increases the number of extreme high and low temperature cycle tests of FCT in the test formula, thereby increasing the test depth and dimension to hedge against the risk of missed tests that may be caused by directly inheriting the historical threshold.
[0032] In HMLV flexible manufacturing scenarios, companies often face the challenge of introducing entirely new product models—models that may have never been produced before and lack any historical testing data. In traditional solutions, engineers often find themselves at a loss when faced with a completely new product, resorting to roughly setting thresholds based on experience or simply applying thresholds from similar products. This crude approach carries extremely high risks in automotive-grade products.
[0033] The cross-model inheritance and risk hedging strategy of this invention specifically addresses the dilemma of zero historical data. When the system recognizes that the current product is being replaced on a new production line and the user has administrator privileges, it first calls the test threshold of the historical model with the highest similarity as the initial threshold. This provides a reasonable starting point for the new product and avoids blindly setting thresholds from scratch. However, simply inheriting historical thresholds carries the risk of missed tests; that is, although two models may be similar, their electrical characteristics and testing focuses may differ.
[0034] To address this, the system forcibly triggers a risk hedging mechanism: while inheriting the initial threshold, it automatically increases the depth and dimensions of testing, including but not limited to increasing the insulation withstand voltage test time, increasing the sampling rate of key signals, and increasing the number of extreme high and low temperature cycle tests in the FCT. These additional test items act as a safety net for the inherited threshold; even if the inherited threshold is not precise enough, increasing the depth and dimensions of testing can effectively capture potential defects, thereby hedging against the risk of missed tests.
[0035] As a further aspect of the present invention, in step S4, the specific process of dynamic convergence is as follows: during the testing of the first N circuit boards after line or batch change, the system statistically analyzes the distribution characteristics of edge defective products in real time; if the edge defective products are concentrated on one side, the system automatically fine-tunes the state compensation threshold or batch offset threshold according to a preset step size until the number of edge defective products drops below the preset threshold, thereby completing the dynamic convergence of the test window, and solidifying the converged threshold as the formal test threshold for the batch or model.
[0036] During the initial stages of line or batch changes, due to the combined effects of factors such as changes in probe wear and batch differences in materials, even with physical condition compensation or batch tolerance offsets, test thresholds may still deviate. This manifests as edge-conforming products being concentrated on one side of the acceptable boundary. In traditional solutions, engineers need to repeatedly adjust the thresholds manually and retest, which is extremely inefficient and relies heavily on personal experience.
[0037] The dynamic convergence mechanism of this invention achieves automatic optimization of the test threshold. The specific process consists of three stages: Stage 1: Statistics. During the testing of the first N circuit boards after a line change or batch change, the system statistically analyzes the distribution characteristics of edge defective products in real time to determine whether they are concentrated on the upper or lower limit side. Stage 2: Fine-tuning. If edge defective products are concentrated on one side, it indicates that there is a systematic shift in the current threshold. The system automatically fine-tunes the threshold in the opposite direction by a preset step size. Each adjustment is followed by retesting subsequent circuit boards until the number of edge defective products drops below the preset threshold. Stage 3: Consolidation. The converged threshold is consolidated as the official test threshold for this batch / model.
[0038] As a further aspect of the present invention, after step S4, a dynamic evaluation step of the authority credit rating based on line change performance is also included: tracking the frequency of a specific engineer or administrator calling the dynamic threshold after a line change or batch change, and the actual defect rate of the corresponding batch of circuit boards in subsequent processes; if the administrator's cross-model inheritance strategy is frequently triggered after a line change, or the defect rate in the later stage exceeds the standard, the authority credit rating of the corresponding personnel is automatically reduced, and the batch tolerance offset safety boundary that can be adjusted during the next line change is narrowed.
[0039] In the traditional RBAC (Restricted Access Control) permission management model, permission granting is static; once a certain level of permission is obtained, it remains unchanged for the validity period. This static model fails to reflect the actual performance of the permission user. An engineer who frequently abuses permissions, causing excessive back-end defect rates, can still enjoy the same permissions as a high-performing engineer.
[0040] This invention introduces a performance-based dynamic credit rating mechanism on the basis of the traditional RBAC model. The system continuously tracks two types of key indicators: first, the frequency and method of specific personnel calling dynamic thresholds after line / batch changes, whether the administrator's cross-model inheritance strategy is frequently triggered, and whether large-scale batch tolerance offsets are frequently performed; second, the actual defect rate of the corresponding batch of circuit boards in subsequent processes, such as aging tests, final inspection, and customer service.
[0041] If a staff member's line change operations frequently trigger high-privilege policies, it indicates that their regular privileges are insufficient to handle line change requirements, or that the subsequent defect rate of the batch they are responsible for exceeds the standard. The system automatically lowers their privilege credibility. The direct consequence of lower credibility is that it reduces the safe boundary for adjusting batch tolerance deviations during the next line change, i.e., it imposes stricter adjustment restrictions, forcing them to be more cautious in making threshold adjustments. This mechanism creates positive incentives and negative constraints.
[0042] As a further aspect of the present invention, in step S2, the specific method for calculating the dynamic deviation is as follows: obtain the actual measured value of the test data, combine it with the nominal tolerance zone of the current material batch, calculate the absolute difference between the actual measured value and the tolerance zone boundary of the current batch; divide the absolute difference by the total width of the tolerance zone of the current batch to obtain the normalized dynamic deviation, which is used to eliminate the interference of inherent tolerance differences between different batches of materials on edge determination.
[0043] In traditional methods, non-conformance is typically determined using absolute thresholds; measurements exceeding fixed upper and lower limits are considered non-conforming. However, different batches of materials have different nominal tolerance zones. Batch A may have poor component consistency, while batch B may have good component consistency. If a uniform absolute deviation standard is used, edge products from batches with wide tolerance zones may be misjudged as severely non-conforming, while edge products from batches with narrow tolerance zones may be missed as conforming.
[0044] The dynamic deviation calculation method of this invention eliminates the interference of inherent batch tolerance differences through normalization. Specifically: First, the actual measured value of the test data is obtained. Combined with the nominal tolerance band of the current material batch, the absolute difference between the actual measured value and the tolerance band boundary of the current batch is calculated. This difference reflects the degree to which the measured value overflows the acceptable boundary. Then, this absolute difference is divided by the total width of the tolerance band of the current batch to obtain the normalized dynamic deviation. The ingenuity of this normalization process lies in its conversion of the deviation measurement from an absolute scale to a relative scale. For batches with wider tolerance bands, the normalized deviation is smaller; for batches with narrower tolerance bands, the normalized deviation is larger. In this way, the setting of the edge threshold has universality across batches.
[0045] As a further aspect of the present invention, before step S1 and step S2, a process of issuing a line change recipe and hierarchical approval is included: when a line change mode is identified, the system automatically matches the corresponding approval flow according to the product level of the target circuit board.
[0046] In automotive-grade PCBA production, any change to the test formula can affect product quality. Traditionally, engineers can modify thresholds and directly issue them to the testing equipment, lacking effective approval control and posing serious quality risks. This invention's tiered approval mechanism implements differentiated control based on product grade. When the system identifies a line changeover mode, it first automatically matches the corresponding approval flow based on the target circuit board's product grade (automotive, industrial, consumer, etc.). For automotive-grade products, after a test engineer modifies the basic safety threshold, it must undergo multi-level electronic signature approval from high-authority administrators before being issued to automated testing equipment. This dual-signature requirement ensures that any threshold modification undergoes independent review from both quality and technical dimensions—the quality director checks for compliance with the quality system, and the system administrator checks for technical feasibility. Simultaneously, the electronic signature approval process automatically generates a complete approval log, meeting the traceability requirements of the automotive-grade quality system.
[0047] As a further aspect of the present invention, step S3 also includes a dynamic permission downgrade mechanism in case of line change abnormality: In line change mode, if the number of consecutive edge defective products exceeds the preset continuous alarm threshold, the system automatically triggers permission downgrade; the current engineer's batch tolerance offset permission is revoked, the equipment is forcibly locked, and a high-level alarm is sent to request the administrator to intervene to perform physical calibration or in-depth investigation of the equipment.
[0048] Even with physical condition compensation strategies, a large number of marginally defective products may still occur in line-change mode under certain circumstances. This could indicate a systemic problem, such as improper installation of test fixtures, severely worn probes requiring replacement, or significant deviations in the new batch of materials. In traditional solutions, engineers often continuously adjust the thresholds to try and pass the test in such situations, which may mask the true systemic problem.
[0049] The dynamic permission degradation mechanism of this invention is automatically triggered when the number of consecutive edge-defective products exceeds a preset threshold. The principle behind this mechanism is that the continuous occurrence of a large number of edge-defective products is itself an abnormal signal, indicating that the current threshold adjustment strategy is no longer effective and that there may be a deeper problem. At this time, the system automatically revokes the current engineer's batch tolerance offset permission, forcibly locks the equipment, and sends a high-level alarm requesting administrator intervention.
[0050] After administrator intervention, physical calibration or in-depth troubleshooting of the equipment is performed—this could be due to probe replacement, recalibration of test fixtures, or batch defects in the materials themselves. Only after these systemic issues have been ruled out can the equipment be unlocked and production resume.
[0051] The automated testing process in the above embodiments relies on an automated testing control system, which includes: The pattern recognition and perception layer is used to identify line change / batch change patterns and collect probe status, environmental parameters, and material batch information; it is responsible for real-time sensing of changes in the production environment. It continuously monitors the production line status to identify line change / batch change patterns, while simultaneously collecting three types of key sensing data—probe status, environmental parameters such as temperature and humidity, and material batch information. This sensing data provides real-time and accurate input for upper-level decision-making.
[0052] The multi-dimensional threshold decision layer, comprising a contact impedance compensation module, a batch tolerance analysis module, and a cross-model inheritance hedging module, is used to execute adaptation strategies and dynamic convergence corresponding to each permission level. This is the system's brain, containing three core functional modules. The contact impedance compensation module executes a physical state compensation strategy based on probe state data provided by the perception layer; the batch tolerance analysis module executes a batch tolerance offset strategy based on material batch information; and the cross-model inheritance hedging module executes cross-model inheritance and risk hedging strategies based on a historical database. These three modules work collaboratively to cover threshold adaptation requirements in different scenarios.
[0053] The access control and credit layer is used for identity authentication, joint matching of production mode and permissions, and dynamic evaluation of permission creditworthiness based on line-change performance. This is the system's policy layer, responsible for three key functions: identity authentication, joint matching of production mode and permissions, and dynamic evaluation of permission creditworthiness based on line-change performance. This layer ensures that the entire system's operation is carried out under strict access control.
[0054] The execution and sorting layer is used to perform initial testing, retesting, dynamic parameter adjustment, and automatic circuit board sorting. This is the system's execution layer, responsible for translating the decisions from the decision-making layer into actual actions. This includes performing initial and retesting, dynamically adjusting test parameters based on decision results, and automatically sorting circuit boards based on test results. This layer achieves a complete closed loop from decision-making to action.
[0055] It will be apparent to those skilled in the art that the present invention is not limited to the details of the exemplary embodiments described above, and that the invention can be implemented in other specific forms without departing from the spirit or essential characteristics of the invention. Therefore, the embodiments should be considered in all respects as exemplary and non-limiting, and the scope of the invention is defined by the appended claims rather than the foregoing description. Thus, it is intended that all variations falling within the meaning and scope of equivalents of the claims be included within the present invention.
Claims
1. A method for automated testing of circuit boards with multi-permission adaptation, characterized in that: Includes the following steps: S1: Obtain the identity authentication information of the current operator and match the permission level, and identify the current production mode. The production mode includes at least the normal production mode, the line changeover mode and the material batch switching mode. S2: Call the basic safety threshold to perform an automated initial test on the circuit board and obtain test data; if it is determined to be unqualified, calculate the dynamic deviation between the test data and the basic safety threshold. When the dynamic deviation is less than the preset edge threshold, mark it as an edge unqualified product and suspend the test. S3: Based on the current operator's permission level and the identified production mode, trigger the corresponding threshold adaptation and retesting strategy: When in line-switching mode and with operator privileges, the physical state compensation strategy is triggered, and the system automatically obtains probe state parameters to generate a state compensation threshold for retesting. When in material batch switching mode and with engineer privileges, the batch tolerance offset strategy is triggered, and the system allows the batch offset threshold to be called within the safety boundary to adjust the test window and retest. When in line-change mode and with administrator privileges, a cross-model inheritance and risk hedging strategy is triggered. This involves calling the threshold for similar models and forcibly adding associated test items for retesting. The specific execution process of the cross-model inheritance and risk hedging strategy is as follows: When a new product line lacks historical test data, the administrator calls the test threshold of the historical model with the highest similarity as the initial threshold; the system forcibly triggers the risk hedging mechanism: while inheriting the initial threshold, it automatically adds insulation withstand voltage test time, increases the sampling rate of key signals, or increases the number of extreme high and low temperature cycle tests for FCT in the test formula. S4: Automatically sort the circuit boards according to the retest results, record the test data and permission operation log, and dynamically converge the state compensation threshold or batch offset threshold based on the actual test data distribution after line change or batch change.
2. The method for automated circuit board testing with multi-permission adaptation according to claim 1, characterized in that: In step S3, the specific execution process of the physical state compensation strategy is as follows: The system automatically collects the cumulative number of probe insertions, real-time contact resistance change rate, and ambient temperature and humidity data of the current test fixture; Substitute the collected physical state parameters into the preset contact impedance compensation model to calculate the threshold fine-tuning amount for the current probe wear state. The state compensation threshold is generated by combining the basic safety threshold with the threshold fine-tuning amount.
3. The method for automated circuit board testing with multi-permission adaptation according to claim 2, characterized in that: In step S3, the specific execution process of the batch tolerance offset strategy is as follows: The system obtains historical tolerance distribution data or first piece test data of the current new batch of materials and calculates the offset of characteristic parameters of the current batch of materials; Guided by the CPK distribution map displayed by the system, the engineer moves or scales the test window as a whole within the safety boundary defined by the basic safety threshold to generate the batch offset threshold. The system performs hard constraint verification on the batch offset threshold input by the engineer. If it exceeds the safety boundary, the application is rejected and an automotive-grade quality alarm is triggered.
4. The method for automated circuit board testing with multi-permission adaptation according to claim 3, characterized in that: In step S4, the specific process of dynamic convergence is as follows: during the testing of the first N circuit boards after line or batch change, the system statistically analyzes the distribution characteristics of edge defective products in real time; if the edge defective products are concentrated on one side, the system automatically fine-tunes the state compensation threshold or batch offset threshold by a preset step size until the number of edge defective products drops below the preset threshold, thus completing the dynamic convergence of the test window, and solidifying the converged threshold as the formal test threshold for the batch or model.
5. The method for automated circuit board testing with multi-permission adaptation according to claim 4, characterized in that: Following step S4, a dynamic assessment step of permission credibility based on line-switching performance is also included: Track the frequency with which a specific engineer or administrator invokes dynamic thresholds after a line change or batch change, as well as the actual defect rate of the corresponding batch of circuit boards in subsequent processes; if the administrator's cross-model inheritance policy is frequently triggered after a line change, or the defect rate in the later stages exceeds the standard, the corresponding personnel's authority and credit score will be automatically reduced, and the batch tolerance offset safety boundary that can be adjusted during the next line change will be narrowed.
6. The method for automated circuit board testing with multi-permission adaptation according to claim 5, characterized in that: In step S2, the specific method for calculating the dynamic deviation is as follows: obtain the actual measured value of the test data, combine it with the nominal tolerance zone of the current material batch, and calculate the absolute difference between the actual measured value and the tolerance zone boundary of the current batch; divide the absolute difference by the total width of the tolerance zone of the current batch to obtain the normalized dynamic deviation, which is used to eliminate the interference of inherent tolerance differences between different batches of materials on edge determination.
7. The method for automated circuit board testing with multi-permission adaptation according to claim 6, characterized in that: Between steps S1 and S2, there is also a process for issuing and approving line change recipes: when a line change mode is identified, the system automatically matches the corresponding approval flow based on the product level of the target circuit board.
8. The method for automated circuit board testing with multi-permission adaptation according to claim 7, characterized in that: Step S3 also includes a dynamic permission downgrade mechanism in case of line switching anomalies: In line switching mode, if the number of consecutive edge defective products exceeds the preset continuous alarm threshold, the system automatically triggers permission downgrade.
Citation Information
Patent Citations
Chip packaging test system and method
CN118782487A
Non-defective product production real-time monitoring system of vehicle gauge equipment
CN120909258A
