Industrial circuit board operation abnormity monitoring system

Abstract

The invention, which relates to the industrial circuit board monitoring technology field, discloses an industrial circuit board operation abnormity monitoring system comprising a multi-sensor welding spot data acquisition module, a signal processing module, an element welding spot mechanical fatigue monitoring module, a welding spot fatigue fault grading early warning module, and an early warning and response module. Data fusion is carried out on the acoustic emission sensor, the strain gauge sensor and the ultrasonic sensor, the state of the welding spot is monitored synchronously from three dimensions of stress acoustic emission, mechanical deformation and internal defects, the strain gauge sensor captures tiny deformation around the welding spot, and the ultrasonic sensor detects internal potential defects. The acoustic emission sensor senses stress release signals, the three kinds of sensor data are subjected to deep fusion analysis through the D-S evidence theory, the problem of misjudgment caused by one-sided information is effectively avoided, stress waves during crack propagation are captured through the acoustic emission signals, and faults can be diagnosed in time in combination with deformation data of the strain gauge sensor.

Description

Technical Field

[0003] The present invention relates to the technical field of industrial circuit board monitoring, and particularly to an abnormal operation monitoring system for industrial circuit boards. Background Art

[0005] In the field of modern industrial automation, industrial circuit boards, as the core components for equipment control and signal transmission, their operational reliability directly affects production continuity and product quality. With the continuous improvement of circuit board integration, the number and density of solder joints have increased significantly. As the key parts connecting electronic components and substrates in circuit boards, solder joints are subjected to multiple effects such as mechanical stress, thermal stress, and environmental corrosion for a long time, and are extremely prone to failure problems such as fatigue, loosening, and cracking. Therefore, accurate monitoring of the operating state of solder joints has become the key to ensuring the stable operation of industrial systems.

[0006] However, the existing industrial circuit board solder joint monitoring technologies have limitations. In the monitoring of solder joint mechanical fatigue, for example, the solder joint random vibration fatigue life prediction method and system provided by the classification number CN105022860A for PCB do not consider multi-dimensional information such as mechanical deformation and stress release around the solder joints, resulting in one-sided monitoring results. Moreover, it is difficult to capture the tiny acoustic emission signals generated by early solder joint fatigue and unable to achieve early warning of faults. In terms of early warning of solder joint fatigue faults, a method, storage medium, and device for evaluating the fatigue degree of vehicle body solder joints provided by the classification number CN112685836A use a fixed damage threshold to divide the early warning levels, without considering the coupling effect of environmental parameters on the solder joint life, resulting in the disconnection between the maintenance strategy and the actual working conditions, lacking a hierarchical control mechanism, and unable to adopt differential response strategies for different degrees of fatigue states, which is extremely prone to false alarms or missed alarms, leading to waste of maintenance resources or expansion of faults.

[0007] In summary, the existing technologies have problems such as single monitoring dimension and lack of dynamic adaptation in early warning response, and it is difficult to meet the real-time and accurate assessment requirements for the health status of solder joints in industrial scenarios. There is an urgent need for a monitoring system that integrates multi-physical field sensing, dynamic weight data fusion, and hierarchical response strategies to break through the performance bottleneck of traditional methods. Summary of the Invention

[0009] The object of the present invention is to make up for the deficiencies of the prior art and provide an industrial circuit board operation anomaly monitoring system. It can synchronously monitor the solder joint state from three dimensions of stress acoustic emission, mechanical deformation, and internal defects by fusing data from acoustic emission sensors, strain gauge sensors, and ultrasonic sensors. The strain gauge sensor captures the subtle deformation around the solder joint, the ultrasonic sensor detects potential internal defects, and the acoustic emission sensor senses the stress release signal. The data of the three sensors are deeply fused and analyzed through the D-S evidence theory. Compared with single-sensor monitoring, this technology effectively avoids the misjudgment problem caused by one-sided information, and can capture the stress wave during crack propagation through the acoustic emission signal. Combining with the deformation data of the strain gauge sensor, it can timely and accurately diagnose faults, greatly improving the comprehensiveness and reliability of solder joint fault diagnosis.

[0010] To solve the above technical problems, the present invention provides the following technical solution: an industrial circuit board operation anomaly monitoring system, which consists of: a multi-sensor solder joint data acquisition module, a signal processing module, a component solder joint mechanical fatigue monitoring module, a solder joint fatigue fault grading and early warning module, and an early warning and response module;

[0011] The multi-sensor solder joint data acquisition module respectively acquires the acoustic emission signal during the operation of the solder joint, the strain signal in the solder joint area, and the ultrasonic reflection echo signal of the internal structure of the solder joint;

[0012] The signal processing module receives the acquired original signal, removes interference through preprocessing operations, and extracts feature parameters related to solder joint fatigue from the preprocessed signal, including the amplitude, frequency, and energy of the acoustic emission signal, the change trend of the strain signal, and the defect characteristics reflected by the ultrasonic signal;

[0013] The multi-component solder joint mechanical fatigue monitoring module, based on the D-S evidence theory, takes the multi-source signals preprocessed by the signal processing module as evidence, and through calculating the basic probability assignment of the evidence, conducts fusion analysis on the acoustic emission signal, the strain signal, and the ultrasonic signal to obtain the solder joint health status evaluation value;

[0014] The solder joint fatigue fault grading and early warning module constructs a solder joint fatigue feature matrix based on the solder joint health status evaluation value of the multi-component solder joint mechanical fatigue monitoring module, and conducts solder joint fatigue state early warning by analyzing the change trend and abnormal degree of the feature parameters in the matrix;

[0015] The early warning and response module is used to trigger corresponding early warning strategies according to the results of the solder joint fatigue early warning state.

[0016] Furthermore, the multi-sensor solder joint data acquisition module includes:

[0017] Acoustic emission sensor: It is used to capture the transient elastic wave signals generated by the solder joints under stress and obtain the acoustic emission signals released by the solder joints due to internal structure changes and damages during operation;

[0018] Strain gauge sensor: It is used in the circuit board area of the solder joint. By monitoring the mechanical deformation of the solder joint area, it outputs strain signals reflecting the strain conditions around the solder joint;

[0019] Ultrasonic sensor: It emits ultrasonic pulses to the solder joint and receives the ultrasonic reflection echo signals generated by internal defects of the solder joint, so as to detect the defect conditions such as cracks and pores inside the solder joint;

[0020] The acoustic emission sensor, strain gauge sensor and ultrasonic sensor respectively collect acoustic emission signals, strain signals and ultrasonic reflection echo signals during the operation of the solder joint from three dimensions of stress acoustic emission, mechanical deformation and internal defects, and transmit the collected original signals to the signal processing module.

[0021] Furthermore, the signal processing module receives the acoustic emission signals, strain signals and ultrasonic reflection echo signals transmitted by the multi-sensor solder joint data acquisition module, and performs preprocessing and feature extraction:

[0022] For the acoustic emission signals, by filtering out the signals in the environmental noise frequency band, removing high-frequency and low-frequency interferences, normalizing the denoised signals, obtaining the time-frequency distribution of the signals, and extracting the maximum value, mean value of the signal amplitude, the central frequency and main frequency band width of the frequency distribution, and the characteristic parameters of the signal energy;

[0023] For the strain signals, use moving average filtering to smooth the signal curve to eliminate interferences, calculate the change slope of the strain signals per unit time, obtain the change trend of the strain signals, and extract the characteristic parameters of the maximum strain, minimum strain and change rate;

[0024] For the ultrasonic reflection echo signals, by measuring the time delay and amplitude of the echo signals, and combining with the ultrasonic propagation speed, extract the characteristic parameters of the peak value of the defect echo amplitude, echo time delay and defect size.

[0025] Furthermore, the component solder joint mechanical fatigue monitoring module extracts the corresponding characteristic parameter sets from the signal processing module, including the acoustic emission signal characteristic parameter set 、the strain signal characteristic parameter set 、the ultrasonic signal characteristic parameter set , in the D-S evidence theory, the degree of support of each sensor's characteristic parameter set for different health states of the solder joint is described by the basic probability assignment function BPA, that is, the BPA function for the acoustic emission sensor: , where represents the The support degree of class feature parameters for the health status of denotes the health status proposition of the solder joint, which includes i.e., the normal state , slight fatigue , moderate fatigue , severe fatigue . is the Euclidean distance between the acoustic emission feature parameter set and the standard feature vector corresponding to the proposition . The smaller the Euclidean distance, the closer the current acoustic emission signal feature is to the ideal feature corresponding to the proposition, that is, the higher the support degree of the signal for the proposition. is the weight coefficient of the acoustic emission sensor, which determines the importance of the sensor in the fusion and , where , , are the standard deviations of the historical data of the acoustic emission, strain, and ultrasonic sensors respectively. The standard deviation reflects the degree of data fluctuation. The sensor with small fluctuation is more stable and reliable, and is assigned a higher weight in the fusion. Similarly, the support degrees for different propositions are calculated respectively through and , and the three are fused into . The conflict coefficient . is the basic probability assignment value when the solder joint is in the state after fusion. In the formula, denotes the sum of the products of the support degrees for all combinations that satisfy , that is, the situation where all three sensors support the solder joint being in the state . The joint support degree is obtained by multiplying and summing their respective support degrees , , . denotes that the propositions supported by the three sensors are completely in conflict, that is, . The sum of the products of the support degrees of these conflict situations is calculated. The larger is used for normalization to eliminate the influence of conflicts and ensure that the sum of the after fusion is 1.

[0026] Furthermore, after fusing the feature parameters, the component solder joint mechanical fatigue monitoring module calculates the solder joint health status evaluation value to judge the fatigue state of the solder joint, where i is the serial number of the health status level, is the basic probability assignment value corresponding to the state, Represents the basic probability assignment value of the v-th health state proposition.

[0027] Furthermore, the solder joint fatigue fault classification and early warning module constructs a solder joint fatigue feature matrix based on the characteristic parameter sets A, B, and C of the component solder joint mechanical fatigue monitoring module , analyzes the abnormal degree of the characteristic parameters in the matrix, and for each parameter , , calculates the normal working range of the characteristic parameters in the matrix and the over-limit ratio , where is the current value of the characteristic parameter in the matrix, and are the mean and standard deviation of the characteristic parameter in the matrix. Using this normal working range and the solder joint health state to output by the component solder joint mechanical fatigue monitoring module and the health state evaluation value S, four-level early warnings are set.

[0028] Furthermore, the four-level early warnings of the solder joint fatigue fault classification and early warning module are as follows:

[0029] Normal state, trigger condition: when the health state evaluation value meets and all parameters of the characteristic matrix are within the normal range;

[0030] Slight fatigue early warning, trigger condition: when the health state evaluation value meets , and at least one parameter in the characteristic matrix exceeds its normal range but the over-limit ratio does not reach 15%;

[0031] Moderate fatigue early warning, trigger condition: and at least one parameter in the characteristic matrix exceeds 15% of its normal range but does not reach 30% and triggers when either condition is met;

[0032] Severe fatigue early warning, trigger condition: and at least one parameter in the characteristic matrix exceeds of its normal range and triggers when either condition is met.

[0033] Furthermore, the early warning strategy of the early warning and response module includes:

[0034] When there is no warning in the normal state, maintain the regular monitoring frequency and record the normal operation data;

[0035] When a slight fatigue warning occurs, the system automatically generates a warning prompt message, including the solder joint position, the current evaluation value S, the name of the over-limit parameter, and the over-limit ratio, and pushes it to the designated maintenance personnel through the system message;

[0036] When a moderate fatigue warning occurs, in addition to the warning prompt message sent by the system, the system automatically increases the monitoring frequency of the solder joint, calls historical data for comparison, generates a prediction report on the fatigue development trend of the solder joint, and performs audible and visual warning flashes;

[0037] When a severe fatigue warning occurs, send a warning prompt message with an emergency label to the equipment responsible person, safety management personnel, and production supervisor, automatically cut off the power supply of the equipment where the circuit board is located, switch to the standby circuit, and generate a detailed fault diagnosis report, including the complete data of the solder joint fatigue characteristic matrix, the analysis of the fault development trend, maintenance suggestions, and the estimated maintenance time.

[0038] Compared with the prior art, the industrial circuit board operation anomaly monitoring system has the following beneficial effects:

[0039] First, the present invention synchronously monitors the solder joint state from three dimensions of stress acoustic emission, mechanical deformation, and internal defects by fusing the data of acoustic emission sensors, strain gauge sensors, and ultrasonic sensors. The strain gauge sensor captures the subtle deformation around the solder joint, the ultrasonic sensor detects potential internal defects, and the acoustic emission sensor senses the stress release signal. The data of the three sensors are deeply fused and analyzed through the D-S evidence theory. Compared with single-sensor monitoring, this technology effectively avoids the misjudgment problem caused by one-sided information. By capturing the stress wave during crack propagation through the acoustic emission signal and combining the deformation data of the strain gauge sensor, the fault can be diagnosed timely and accurately, greatly improving the comprehensiveness and reliability of solder joint fault diagnosis.

[0040] Second, the solder joint fatigue fault grading and warning mechanism constructed by the present invention divides the fatigue state into three levels: slight, moderate, and severe based on the parameter change trend and anomaly degree of the solder joint fatigue characteristic matrix, and formulates different warning strategies accordingly. When slight fatigue occurs, prompt information is sent in time to remind the staff to pay attention. When moderate fatigue occurs, encrypted monitoring is started and maintenance suggestions are provided. When severe fatigue occurs, the equipment is automatically linked to stop, preventing the deterioration of the fault. This refined control mode changes the traditional fixed alarm method and realizes the transformation from fault monitoring to active defense.

[0041] Other advantages, objectives, and features of the present invention will be described to some extent in the subsequent specification, and to some extent, will be obvious to those skilled in the art based on the study of the following text, or can be taught from the practice of the present invention. Brief Description of the Drawings

[0043] In order to more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the following will briefly introduce the drawings required in the description of the embodiments or the prior art. Obviously, the drawings in the following description are only some embodiments of the present invention. For those of ordinary skill in the art, without creative efforts, other drawings can also be obtained based on these drawings.

[0044] Figure 1 It is an operation flowchart of an industrial circuit board operation anomaly monitoring system;

[0045] Figure 2 It is a schematic diagram of the system composition of an industrial circuit board operation anomaly monitoring system. Detailed Embodiments

[0047] In order to further elaborate on the technical means and effects adopted by the present invention to achieve the predetermined invention purpose, the following will, in conjunction with the drawings and preferred embodiments, detail the specific embodiments, structures, features and their effects of the present invention as follows.

[0048] Embodiment 1

[0049] This embodiment provides a specific implementation manner of an industrial circuit board operation anomaly monitoring system, as Figure 2 shown. Through a multi-sensor solder joint data acquisition module, a signal processing module, a component solder joint mechanical fatigue monitoring module, a solder joint fatigue fault grading and early warning module, and an early warning and response module, precise monitoring and active prevention and control of the fatigue state of the solder joints of the industrial circuit board are realized. The system constructs a multi-source data acquisition system based on three sensors of acoustic emission, strain, and ultrasonic, uses the D-S evidence theory to complete multi-modal signal fusion analysis, realizes the quantitative evaluation of the fatigue state through the feature matrix and the grading threshold, and formulates differentiated response strategies for different early warning levels, solving the problems of single dimension and lagging early warning of traditional monitoring technologies, and significantly improving the operation reliability and maintenance efficiency of the industrial circuit board.

[0050] The described multi-sensor solder joint data acquisition module is used for the core task of synchronously obtaining solder joint operation status information from three dimensions: stress acoustic emission, mechanical deformation, and internal defects. This module consists of an acoustic emission sensor, a strain gauge sensor, and an ultrasonic sensor. The three are deployed in a spatially distributed manner to form an all-round monitoring network for the solder joint. Among them, the acoustic emission sensor is deployed on the surface of the circuit board near the solder joint. Its core function is to capture the transient elastic wave signals released due to internal structural damage (such as the initiation and propagation of microcracks) under stress in the solder joint. When the chemical bonds between the molecules of the solder joint material break or the crystal structure slips, elastic waves with an extremely short time (microsecond level) will be generated. Such signals carry key information such as the location, type, and severity of the damage. The sensor converts mechanical vibration into an electrical signal through the piezoelectric effect. Its frequency response range covers 10 kHz to 1 MHz, and it can effectively sense the weak acoustic emission signals generated by early fatigue; the strain gauge sensor adopts a foil structure and is pasted on the circuit board substrate around the solder joint through a special adhesive to form real-time monitoring of the mechanical deformation in the solder joint area. When the solder joint deforms under mechanical stress (such as vibration, thermal expansion), the resistance value of the strain gauge will change linearly with the expansion and contraction of the substrate. The resistance change is converted into a voltage signal output through a Wheatstone bridge circuit. This sensor can capture the subtle deformation of the solder joint and provide dynamic deformation data support for fatigue monitoring; the ultrasonic sensor consists of a transmitting probe and a receiving probe, which are coupled to the surface of the solder joint to achieve internal defect detection through the pulse reflection method. The transmitting probe emits high-frequency ultrasonic pulses (frequency range 1 MHz - 10 MHz) to the solder joint. When the sound wave encounters defects such as cracks and pores, echo signals will be generated. The receiving probe captures the echo signals and converts them into electrical signals. By analyzing the time delay and amplitude of the echo, the location, size, and nature of the defect can be inferred. The time delay of the echo corresponds to the depth of the defect, and the amplitude reflects the reflection area of the defect. The three types of sensors ensure the acquisition of multi-physical field data of the solder joint within the same time window through a synchronous acquisition trigger mechanism. The original signals are transmitted to the signal processing module through shielded cables, and differential amplification technology is used during the transmission process to suppress common-mode interference and ensure signal integrity.

[0051] The signal processing module performs noise reduction, normalization, and feature parameter extraction on the original signals transmitted by multiple sensors, providing a high-quality data basis for subsequent fusion analysis. This module designs different processing flows according to the characteristics of different types of signals: for acoustic emission signals, first filter out the environmental noise frequency band through a band-pass filter, and further suppress high-frequency white noise and low-frequency drift interference using the wavelet noise reduction algorithm. The denoised signal is normalized to map the amplitude to the interval [0, 1], and the time-frequency distribution characteristics are obtained. On this basis, amplitude features are extracted: the maximum value reflects the energy release intensity of sudden damage events, the mean value represents the overall energy level of the signal, frequency features: the center frequency reflects the main energy distribution frequency band of the signal, the main frequency band width reflects the complexity of the frequency components, energy features: the energy value is calculated by integrating the square of the signal amplitude, which is used to measure the energy accumulation effect during the damage process; the strain signal processing flow takes smoothing filtering and trend analysis as the core, uses moving average filtering to eliminate the randomness in the signal, and obtains the deformation trend by calculating the difference in strain values per unit time (i.e., the change slope). The characteristic parameters include extreme value features: the maximum and minimum values define the strain fluctuation range, reflecting the stress cycle amplitude borne by the solder joint, rate features: the change rate (strain increment per unit time) is used to judge the speed of deformation development, and the sudden change rate may indicate a rapid damage process; the ultrasonic reflection echo signal processing focuses on defect feature extraction, calculates the defect position through echo positioning, and extracts amplitude features: the ratio of the peak value of the defect echo amplitude to the echo amplitude in the defect-free area, which is used to evaluate the relative size of the defect, time delay features: the difference between the echo time delay and the propagation time of the sound wave in the defect-free material, corresponding to the buried depth of the defect, size features: estimate the height or length of the defect through the change in echo phase; after all the characteristic parameters are standardized, a feature vector with a unified dimension is formed, laying a foundation for multi-source data fusion.

[0052] The described component solder joint mechanical fatigue monitoring module realizes the fusion reasoning of multi-source feature data based on the D-S evidence theory, and quantitatively evaluates the health state of the solder joint. The processing flow of this module includes basic probability assignment (BPA) calculation, evidence fusion, and generation of the health state evaluation value. First, for the three types of characteristic parameter sets of acoustic emission (A), strain (B), and ultrasound (C), basic probability assignment functions are respectively constructed. In this embodiment, taking the acoustic emission sensor as an example, its BPA function is defined as: , where represents the health state proposition of the solder joint corresponding to normal, slight fatigue, moderate fatigue, and severe fatigue respectively), is the Euclidean distance between the current acoustic emission feature vector and the corresponding standard feature vector. The smaller the Euclidean distance, the closer the current signal feature is to the ideal mode of this health state, and the higher the support degree for . The weight coefficient Determined by the standard deviation of the sensor historical data: Among them, , , are the standard deviations of the historical characteristic parameters of three types of sensors respectively, reflecting the stability of the data. The smaller the standard deviation, the higher the reliability of the sensor, and the greater the weight assigned in the fusion. Similarly, calculate the BPA functions and . Next, multi-evidence fusion is achieved through the D-S combination rule, and the fused basic probability assignment value is: , where the conflict coefficient characterizes the conflict degree between sensor evidences: When , the evidences are completely consistent; The larger it is, the more significant the evidence conflict. The denominator is used for normalization to ensure that the sum of the probability assignment values after fusion is 1. Finally, the fatigue degree of the solder joint is quantified through the health state evaluation value : . This formula obtains a continuous evaluation value between 1 and 4 by weighted averaging the basic probability assignment values of each health state. The closer it is to 1, the healthier the solder joint state; the closer it is to 4, the more serious the fatigue degree.

[0053] Based on the fused health state evaluation value and the characteristic parameter matrix, the solder joint fatigue fault grading and early warning module constructs a four-level early warning mechanism to achieve dynamic grading control of the fatigue state. This module includes characteristic matrix construction, abnormal degree analysis, and early warning rule definition. Specifically: By integrating three types of characteristic parameter sets (acoustic emission , strain , ultrasonic ), construct the solder joint fatigue characteristic matrix: . Each element ( ) in the matrix corresponds to a characteristic parameter. Its normal working range is determined through statistical analysis as , where is the parameter mean, is the standard deviation. The normal range is set based on the 3σ principle. Exceeding this range is regarded as abnormal, and the abnormal degree is quantified through the overrun ratio : reflects the relative degree of the current parameter value deviating from the normal range. The larger the value, the more significant the abnormality. Based on the health state evaluation value and the overrun ratio , define the four-level early warning rules, that is, the normal state ( ): When And all characteristic parameters are within the normal range, indicating that the solder joint is in a healthy operating state; minor fatigue warning : When and at least one parameter exceeds the limit ratio , it indicates that early fatigue signs begin to appear in the solder joint; moderate fatigue warning( ): When or at least one parameter exceeds the limit ratio , it indicates that the fatigue degree has developed significantly; severe fatigue warning : When or at least one parameter exceeds the limit ratio , it indicates that the solder joint is about to fail, and emergency measures need to be taken immediately. The warning rule ensures that a significant anomaly in a single dimension can trigger the corresponding level of warning, avoiding risks caused by missed judgments of a single indicator.

[0054] The warning and response module triggers differentiated disposal strategies according to the warning level, including the normal state (no warning): The system maintains the conventional monitoring frequency (such as collecting data once a minute), automatically records the operating parameters of the solder joint as historical baseline data, and dynamically updates the mean and standard deviation of each characteristic parameter through long-term data accumulation to achieve adaptive adjustment of the normal working range and improve the robustness of the monitoring system to environmental changes; minor fatigue warning: The system immediately generates a warning prompt message, the content of which includes the solder joint position, the current evaluation value, the name of the parameter exceeding the limit, and the exceeding limit ratio. The information is pushed to the terminal device of the designated maintenance personnel. After receiving the prompt, the maintenance personnel can call the historical data curve of the solder joint through the system, analyze the fatigue development trend, and formulate a preventive maintenance plan; moderate fatigue warning: In addition to performing all operations of the minor fatigue warning, the system automatically increases the monitoring frequency of the solder joint to once per second and starts the historical data comparison analysis, matches the current characteristic parameter sequence with the historical fatigue sample library, predicts the fatigue development trend in the next period of time, generates a trend report including the remaining life estimate. At the same time, the monitoring interface triggers an audible and visual warning to remind the on-site personnel to pay attention. In addition, the system pushes a work order containing maintenance suggestions to the maintenance department, and the suggestion content is generated based on the similar failure case library; severe fatigue warning: The system enters the emergency response mode. First, it automatically cuts off the main power supply of the device where the circuit board is located through hard-wired control, and at the same time triggers the standby circuit switching mechanism to ensure that the production process is not interrupted. Subsequently, it sends warning messages with emergency marks to the device responsible person, safety management personnel, and production supervisor. The information includes the complete data of the solder joint fatigue characteristic matrix, the analysis of the fault development trend, and the detailed maintenance plan. At the same time, the system automatically generates a standardized fault diagnosis report, which can be directly used for maintenance record filing and subsequent fault tracing.

[0055] In summary, through the full - process design of multi - sensor collaborative perception, multi - modal signal intelligent processing, D - S evidence theory fusion analysis, and hierarchical early - warning response, this embodiment constructs a complete industrial circuit board solder joint fatigue monitoring system. The multi - sensor module realizes comprehensive monitoring from multiple dimensions of stress, deformation, and defects. The signal processing module improves the data availability through feature extraction. The mechanical fatigue monitoring module solves the one - sidedness problem of a single sensor by using dynamic weight fusion. The hierarchical early - warning module accurately depicts the fatigue state through quantization rules, and the response module further converts the monitoring results into actual prevention and control capabilities.

[0056] Embodiment 2

[0057] As Figure 1 shown, this embodiment provides a process for monitoring the abnormal operation of solder joints on an industrial circuit board through an industrial circuit board operation abnormal monitoring system. The specific steps of this process are as follows:

[0058] Step 1: Synchronous acquisition of multi - sensor data

[0059] Deploy acoustic emission sensors to capture transient elastic wave signals generated by the stress release of solder joints in real - time;

[0060] Install strain - gauge sensors in the solder joint area to continuously monitor mechanical deformation data;

[0061] Emit pulses to the solder joints through ultrasonic probes and receive the echo signals reflected by internal defects;

[0062] Synchronously transmit the three types of original signals to the signal processing module;

[0063] Step 2: Pre - processing of multi - modal signals

[0064] Filter out high - frequency environmental noise and extract the peak amplitude and main frequency band energy distribution of the signals;

[0065] Use moving average filtering to eliminate temperature drift and calculate the change rate of deformation gradient per unit time;

[0066] Separate the circuit board substrate noise from the defect reflection wave and measure the echo propagation delay and amplitude attenuation rate;

[0067] Step 3: Fusion evaluation of the health status of solder joints

[0068] Construct an acoustic emission feature set A, a strain feature set B, and an ultrasonic feature set C;

[0069] Calculate the support probabilities of each sensor for four health statuses (normal / slight / moderate / severe);

[0070] Dynamically weight and fuse multi - source evidence (assign higher weights to sensors with more stable historical data);

[0071] Output the solder joint health assessment value S

[0072] Step 4: Hierarchical early warning for the fatigue feature matrix

[0073] Construct a solder joint fatigue feature matrix;

[0074] Calculate the overlimit ratio of each characteristic parameter deviating from the normal range;

[0075] Mark the positions and deviation degrees of the overlimit parameters

[0076] Dynamically trigger four-level early warnings: normal, mild fatigue early warning, moderate fatigue early warning, and severe fatigue early warning;

[0077] Step 5: Execute the hierarchical response strategy

[0078] Normal state: Maintain the conventional sampling frequency;

[0079] Mild fatigue: Automatically generate an early warning work order (including solder joint coordinates, S value, and list of overlimit parameters), and push the message to the mobile terminal of the designated maintenance personnel;

[0080] Moderate fatigue: Increase the sampling frequency, generate a fatigue development trend prediction curve, and trigger the audible and visual alarm on the device panel;

[0081] Severe fatigue: Send a red emergency alarm to the three-level responsible persons (equipment / safety / production supervisor), cut off the power supply of the target circuit board, switch to the redundant backup circuit, and generate a diagnostic report containing the repair plan.

[0082] The above are only the preferred embodiments of the present invention, and do not impose any form of limitation on the present invention. Although the present invention has been disclosed above with the preferred embodiments, it is not intended to limit the present invention. Any person skilled in the art can make some changes or modifications to the above-disclosed technical content to obtain equivalent embodiments with equivalent changes, but as long as the technical content of the present invention is not departed from, any brief modifications, equivalent changes, and modifications made to the above embodiments based on the technical essence of the present invention still fall within the scope of the technical solution of the present invention.

Claims

1. An industrial circuit board operation anomaly monitoring system, characterized in that, The system consists of: a multi-sensor solder joint data acquisition module, a signal processing module, a component solder joint mechanical fatigue monitoring module, a solder joint fatigue fault grading and early warning module, and an early warning and response module; The multi-sensor solder joint data acquisition module respectively acquires acoustic emission signals during the operation of the solder joint, strain signals in the solder joint area, and ultrasonic reflection echo signals of the internal structure of the solder joint; The signal processing module receives the acquired original signals, removes interference through preprocessing operations, and extracts feature parameters related to solder joint fatigue from the preprocessed signals, including the amplitude, frequency, and energy of the acoustic emission signals, the change trend of the strain signals, and the defect features reflected by the ultrasonic signals; The multi-component solder joint mechanical fatigue monitoring module is based on the D-S evidence theory. Taking the multi-source signals preprocessed by the signal processing module as evidence, through calculating the basic probability assignment of the evidence, it conducts fusion analysis on the acoustic emission signals, strain signals, and ultrasonic signals to obtain an evaluation value of the solder joint health state; The solder joint fatigue fault grading and early warning module constructs a solder joint fatigue feature matrix based on the evaluation value of the solder joint health state of the multi-component solder joint mechanical fatigue monitoring module, and conducts early warning of the solder joint fatigue state by analyzing the change trend and abnormal degree of the feature parameters in the matrix; The early warning and response module is used to trigger corresponding early warning strategies according to the results of the solder joint fatigue early warning state.

2. The industrial circuit board operation anomaly monitoring system according to claim 1, wherein, The multi-sensor solder joint data acquisition module includes: An acoustic emission sensor: used to capture the transient elastic wave signals generated by the solder joint under stress, and obtain the acoustic emission signals released during the operation of the solder joint due to internal structure changes and damage; A strain gauge sensor: used in the circuit board area of the solder joint, by monitoring the mechanical deformation of the solder joint area, to output strain signals reflecting the strain conditions around the solder joint; An ultrasonic sensor: emits ultrasonic pulses to the solder joint and receives the ultrasonic reflection echo signals generated by internal defects of the solder joint, so as to detect the defect conditions such as cracks and pores inside the solder joint; The acoustic emission sensor, strain gauge sensor, and ultrasonic sensor respectively synchronously acquire acoustic emission signals, strain signals, and ultrasonic reflection echo signals during the operation of the solder joint from the three dimensions of stress acoustic emission, mechanical deformation, and internal defects, and transmit the acquired original signals to the signal processing module.

3. The industrial circuit board operation anomaly monitoring system according to claim 1, characterized in that The signal processing module receives the acoustic emission signals, strain signals, and ultrasonic reflection echo signals transmitted by the multi-sensor solder joint data acquisition module, and conducts preprocessing and feature extraction: For the acoustic emission signals, by filtering out the signal in the environmental noise frequency band, removing high-frequency and low-frequency interference, normalizing the denoised signal, obtaining the time-frequency distribution of the signal, and extracting the maximum value, mean value of the signal amplitude, the central frequency and main frequency band width of the frequency distribution, and the feature parameters of the signal energy; For the strain signals, use moving average filtering to smooth the signal curve to eliminate interference, calculate the change slope of the strain signal per unit time to obtain the change trend of the strain signal, and extract the feature parameters of the maximum strain, minimum strain, and change rate; For the ultrasonic reflected echo signal, by measuring the time delay and amplitude of the echo signal and combining with the ultrasonic propagation speed, the characteristic parameters such as the peak value of the defect echo amplitude, the echo time delay, and the defect size are extracted.

4. An industrial circuit board operation anomaly monitoring system according to claim 1, characterized in that, The mechanical fatigue monitoring module of the component solder joint extracts the corresponding characteristic parameter sets from the signal processing module, including the acoustic emission signal characteristic parameter set , the strain signal characteristic parameter set , the ultrasonic signal characteristic parameter set . In the D-S evidence theory, the degree of support of each sensor's characteristic parameter set for different health states of the solder joint is described by the basic probability assignment function BPA. That is, for the BPA function of the acoustic emission sensor: , where represents the degree of support of the th type of characteristic parameter for the health state , represents the proposition of the health state of the solder joint, , which includes , that is, the normal state , slight fatigue , moderate fatigue , severe fatigue . is the Euclidean distance between the acoustic emission characteristic parameter set and the standard characteristic vector corresponding to the proposition . is the weight coefficient of the acoustic emission sensor and , where , , are the standard deviations of the historical data of the acoustic emission, strain, and ultrasonic sensors respectively. Similarly, the degrees of support for different propositions are calculated through and respectively, and the three are fused into , the conflict coefficient . is the basic probability assignment value when the solder joint is in the state after fusion.

5. An industrial circuit board operation anomaly monitoring system according to claim 4, characterized in that, After fusing the characteristic parameters, the mechanical fatigue monitoring module of the component solder joint calculates the evaluation value of the solder joint health state , and judges the fatigue state of the solder joint, where i is the serial number of the health state level is the basic probability assignment value corresponding to the state represents the basic probability assignment value of the v-th health state proposition 6. The industrial circuit board operation anomaly monitoring system according to claim 4, wherein The solder joint fatigue fault classification and early warning module constructs a solder joint fatigue feature matrix based on the characteristic parameter sets A, B, and C of the component solder joint mechanical fatigue monitoring module , analyzes the abnormal degree of the characteristic parameters in the matrix, and for each parameter , , calculates the normal operating range of the characteristic parameters in the matrix and the over-limit ratio , where is the current value of the characteristic parameter in the matrix, and are the mean and standard deviation of the characteristic parameter in the matrix. Using this normal operating range and the solder joint health status to output by the component solder joint mechanical fatigue monitoring module and the health status evaluation value S, four-level early warnings are set.

7. An industrial circuit board operation anomaly monitoring system according to claim 6, characterized in that, The four-level early warning of the solder joint fatigue fault grading early warning module is as follows: Normal state, triggering condition: when the health status evaluation value meets and the feature matrix all parameters are within the normal range; Minor fatigue warning, trigger condition: when the health status assessment value meets , and at least one parameter in the feature matrix exceeds its normal range but the overlimit ratio does not reach 15%; Moderate fatigue warning, triggering conditions: and the feature matrix When at least one of the parameters exceeds 15% but does not reach 30% of its normal range, it is triggered when one of them is satisfied; Severe fatigue warning, trigger conditions: and the feature matrix When at least one parameter in it exceeds its normal range It is triggered when one of them is satisfied.

8. An abnormal operation monitoring system for an industrial circuit board according to claim 1, characterized in that, The early warning strategies of the early warning and response module include: When there is no early warning in the normal state, maintain the conventional monitoring frequency and record the normal operation data; When there is a slight fatigue early warning, the system automatically generates an early warning prompt message, including the solder joint position, the current evaluation value S, the name of the over-limit parameter, and the over-limit ratio, and pushes it to the designated maintenance personnel through the system message; When there is a moderate fatigue early warning, in addition to the early warning prompt message sent by the system, the monitoring frequency of this solder joint is automatically increased, historical data is called for comparison, a prediction report on the fatigue development trend of this solder joint is generated, and an audible and visual early warning flash is carried out; When there is a severe fatigue early warning, an early warning prompt message with an emergency label is sent to the equipment responsible person, the safety management personnel, and the production supervisor, the power supply of the equipment where this circuit board is located is automatically cut off, and the standby circuit is switched to, and a detailed fault diagnosis report is generated, including the complete data of the solder joint fatigue characteristic matrix, the analysis of the fault development trend, the maintenance suggestions, and the estimated maintenance time.

Images