Electronic component environment reliability experiment system

Abstract

The invention discloses an electronic component environment reliability experiment system, which relates to the field of electronic components and comprises a temperature control module, a humidity control module, a vibration module, an impact module, a master control module and a remote communication module, the master control module is integrated with a data receiving module, a data processing module, a reliability verification module, a process monitoring module and a result output module, the data receiving module receives data from the temperature control module, the humidity control module, the vibration module and the impact module, and the data processing module processes the received data. The process monitoring module monitors processed data in real time, the reliability verification module evaluates reliability performance of the electronic device in different environments, the remote communication module completes setting of environmental conditions and receiving of feasibility performance results, and the environment reliability experiment capability of the electronic device is improved.

Description

Technical Field

[0001] This invention relates to electronic components, and more specifically to an environmental reliability testing system for electronic components. Background Technology

[0002] Electronic components are a crucial part of electronic products, and their reliability determines the overall reliability and stability of the products. The reliability of electronic components in complex environments such as high temperature, alternating high and low temperatures, humidity, vibration, and shock is particularly important. Electronic components are the fundamental building blocks of modern electronic products, and their environmental reliability determines their reliability in complex environments. With the development of 5G communication, new energy vehicles, and industrial internet, electronic components face increasingly complex environmental applications. They must not only withstand high temperatures (≤300℃) but also cope with more complex combinations of humidity, vibration (2H-3000Hz), and shock. Furthermore, over 70% of electronic product system failures are caused by components failing to operate under harsh environmental conditions. Therefore, building precise, intelligent environmental reliability laboratories is an urgent need for enterprises and researchers.

[0003] With the continuous development of reliability engineering technology, environmental reliability testing systems are also constantly being improved and upgraded. Modern testing systems have higher testing accuracy and richer testing functions to meet the diverse needs of customers in environmental reliability testing. However, they still suffer from problems such as complex testing operations and high testing costs due to the analysis of test data. Based on the above description, the existing technological bottlenecks are analyzed as follows: 1. Insufficient ability to perform coupled testing of multiple environmental factors. Traditional experimental modules are mostly single modules, such as temperature, humidity, vibration, and shock modules, which cannot meet the requirements of multiple factors interacting in actual use. For example, current vibration modules are mostly single-frequency modules with a frequency <2000Hz and displacement accuracy (>51mm), which cannot meet the reliability testing requirements of avionics components under high-speed vibration environments; humidity control cannot be coupled with temperature control; and the coordination accuracy of humidity and temperature at high humidity is less than ±5%, which cannot meet the simulation of failure scenarios in coastal and humid and hot environments.

[0004] 2. Adaptability limitations of reliability assessment models Existing systems mostly use the basic Weibull distribution model for lifetime analysis, but this model has significant limitations: Parameter fixation: The shape parameter k and the scaling parameter λ of the traditional Weibull model depend on human experience and cannot adapt to the distribution of failure data under different environmental conditions (such as concentrated failure data under high temperature environment, while scattered failure data under vibration environment). Lack of multi-stress coupling analysis: The lack of mathematical description of the interaction of multiple factors such as temperature, humidity, and vibration leads to deviations between reliability assessment results and actual working conditions. According to industry test data, the life prediction error rate of traditional models in comprehensive stress tests exceeds 25%. Insufficient real-time performance: Relies on offline data processing, making it impossible to dynamically adjust test parameters during experiments and difficult to capture changes in the critical threshold of component failure. 3. Technical shortcomings in data processing and remote monitoring Multi-source heterogeneous data fusion is difficult: the output signal formats of temperature sensors, vibration sensors, shock sensors, etc. are not uniform (such as a mixture of analog and digital signals), and traditional data receiving modules lack compatible interface design, resulting in low data acquisition efficiency and easy data loss. Real-time monitoring and alarm lag: Existing process monitoring modules mostly use fixed threshold comparisons and do not dynamically adjust based on the real-time operating status of components. For example, when temperature and humidity change suddenly, the alarm response time exceeds 30 seconds, which cannot provide timely warning of potential failure risks. Weak remote control capabilities: Inconsistent communication protocols (such as a mix of Modbus and TCP / IP) lead to poor device interconnection compatibility and high latency in remote parameter setting (>100ms), making it difficult to meet the real-time requirements of multi-site collaborative experiments. 4. Insufficient integration and automation level of the experimental system. The existing experimental equipment has low precision in coordination and control between modules. For example, the start-stop synchronization accuracy of the temperature and vibration modules is less than 500ms, and the stress loading sequence is irregular. Data processing using manual control takes more than 24 hours from experimental data acquisition to the formation of a reliability report, which is difficult to keep up with the rapid iteration of product development. Summary of the Invention

[0005] To address the shortcomings of the aforementioned technologies, this invention discloses an environmental reliability testing system for electronic components. This system can coordinate and manage the entire testing process, as well as analyze and verify the reliability of the data. It is easy to use, highly automated, and efficient, effectively improving accuracy and reducing costs.

[0006] To achieve the above-mentioned technical effects, the present invention adopts the following technical solution: An environmental reliability testing system for electronic components, comprising: The temperature control module evaluates the temperature tolerance and stability by simulating the operation of components under different temperature environments, including high temperature, low temperature, and temperature cycling; the high temperature is ≤300℃. The humidity control module simulates the effects of humidity in real-world environments to evaluate the moisture resistance of components in high-humidity environments, as well as their stability and tolerance under humid conditions. The vibration module simulates the stability and bearing capacity of components under vibration conditions, namely, frequency 2Hz-3000Hz, displacement ≤51mm, and load ≤3000Kg. The impact module assesses the reliability of components by simulating the impact conditions that components may encounter during transportation, installation, and use. The main control module coordinates and manages the entire experimental process, including data reception and processing, monitoring the experimental process, reliability verification, and result output. The main control module includes a data receiving module, a data processing module, a reliability verification module, a process monitoring module, and a result output module. The receiving module receives experimental data; the data processing module processes the received experimental data; the process monitoring module monitors the processed experimental data in real time; the reliability verification module analyzes the environmental reliability of electronic components; and the result output module transmits the analyzed results to a remote control center. The remote communication module uses standard communication protocols to establish a stable and high-speed data transmission channel between the experimental equipment and the remote control center, setting environmental parameters and threshold data, executing test programs, and receiving experimental results output by the main control module. The input terminals of the main control module are connected to the output terminals of the temperature control module, humidity control module, vibration module, and impact module, respectively; the output terminal of the remote communication module is connected to the input terminals of the temperature control module, humidity control module, vibration module, and impact module; and the main control module is connected to the remote control module.

[0007] As a further embodiment of the present invention, the temperature control module and the humidity control module are respectively composed of a temperature sensor and a humidity sensor, a control circuit and an actuator. The temperature sensor and the humidity sensor detect the ambient temperature and humidity. The control circuit receives the temperature and humidity signals, processes the data, compares the set value with the actual value, and outputs a control signal according to the comparison result. The actuator adjusts the temperature and humidity according to the signal output by the control circuit.

[0008] As a further embodiment of the present invention, the vibration module uses a vibration sensor to monitor the vibration parameters of the components in the vibration experiment, namely the vibration frequency and amplitude, and uses a displacement sensor to measure the vibration displacement data of the components during the vibration process.

[0009] As a further embodiment of the present invention, the impact module uses an impact sensor to measure the magnitude and duration of the impact force received by the component in the impact experiment, a strain sensor to measure the strain of the component when it is impacted, and an acceleration sensor to record the peak value and changes of the acceleration of the component when it is impacted in real time.

[0010] As a further embodiment of the present invention, the data receiving module includes a compatible data interface, an information encoding module, a receiving path planning module, and an information decoding module; the data processing module performs filtering, smoothing, and calibration processing on the data; the process monitoring module includes an information extraction module, a threshold comparison module, an alarm module, and a visualization display module; the reliability verification module performs lifetime data analysis using an improved Weibull distribution model, which comprises a parameter setting module, a Weibull ratio adjustment module, a Weibull regression module, a Weibull adaptive adjustment module, a reliability result detection module, and a reliability result output module.

[0011] As a further embodiment of the present invention, the compatible data interface uses a USB 3.2 interface to achieve adaptation between devices and instruments using different data interface standards and supports high-speed data transmission; the information encoding module is used to encode various information during the experiment, using the UTF-8 encoding standard to convert temperature, humidity, voltage, current, frequency, and acceleration data into a unified format and standard, and compressing the data to reduce storage space usage; the receiving path planning module plans the data receiving path according to the sensor's location, data type, and real-time requirements; the data decoding module decodes the data according to the encoding method and format, and then appends a checksum to the target data for verification.

[0012] As a further embodiment of the present invention, the information extraction module is used to extract information from the data processed by the data processing module, including voltage, current, temperature, humidity, frequency and acceleration data; the threshold comparison module is used to compare the extracted data with the normal operating range values ​​of the components to perform threshold comparison; the alarm module is used to issue an alarm when the received data exceeds the threshold; and the visualization display module is used to perform charting and curve processing on the data.

[0013] As a further embodiment of the present invention, the reliability verification module performs data analysis using an improved Weibull distribution model; the parameter setting module is used to set constants in the improved Weibull distribution model; the Weibull proportional adjustment module is used to adjust the proportional parameters in the improved Weibull distribution model; the Weibull regression module is used to analyze the failure modes of electronic components under specific environmental conditions; the Weibull adaptive adjustment module is used to adaptively adjust experimental conditions or parameters based on the failure data of electronic components under specific environments and the Weibull distribution characteristics; the reliability result detection module is used to evaluate the performance and reliability level under specific environmental conditions based on the improved Weibull distribution model; the reliability result output module is used to output the reliability level result obtained by the reliability result detection module; the probability density of the improved Weibull stepwise method is: In formula (1), x is a random variable. 'k' is the scale parameter, and 'k' is the shape parameter. When the scale parameter is the same, the larger the shape parameter k, the more concentrated the failure data of the sample; conversely, the smaller the shape parameter k, the larger the area where the failure data falls, and the more dispersed the sample failure data. The steps of the Weibull analysis method are as follows: Determine sample size and time: Determine the sample size and testing time based on the actual usage environment of the product, potential problems, and expected lifespan; Collect fault data: Test the product according to the set sample size and test time, and record the fault data; Plotting the failure rate curve: Plot the product's failure rate curve based on the failure data; Fitting the failure rate curve: The failure rate curve is fitted using the Weibull distribution function to obtain the Weibull distribution parameters; Analyze the results and assess reliability: The reliability of the product is assessed by analyzing the fitting results. The reliability function for electronic components is: ;In formula (2), This represents the number of failures by time t. Indicates the total number of test products. This represents the output characteristic value of the Weibull distribution function, where M represents the type of experiment. Indicates from unit time to Total number of products within the time range This represents the number of products at time i. The time interval is represented by i, which represents the unit of time from the start of the experiment. The failure rate is the probability that a product will fail within a unit of time after time t. The failure probability function is: ;In formula (3), This represents the number of products with normal performance at time t. The number of failed products within the monitoring time interval. The output characteristic value represents the failure probability function, and W represents a constant value under different environmental conditions. This represents the number of products that failed instantaneously at time t; The failure rate of electronic components follows an exponential distribution, and the relationship between reliability and failure rate is as follows: ;In formula (4), Indicates failure rate. 'b' represents a constant value. As a further technical solution of the present invention, the parameter setting module is controlled by an ADE7758 chip to set the name, type, default value, allowed value range, or control instructions of the Weibull distribution model parameters; The Weibull scaling module adjusts the proportions of the Weibull distribution model parameters through parameter estimation and optimization algorithms. The Weibull regression module fits the Weibull distribution parameters by maximizing the likelihood function; The Weibull adaptive adjustment module changes the model structure by adjusting model weights, changing parameter update frequency, or responding to specific events. The reliability result detection is achieved by filtering data parameter information through failure time, environmental conditions, life test, sequential test, and environmental stress. The reliability result output module outputs data information by removing outliers, filling in missing data, converting data formats, and calculating the mean, median, or variance.

[0014] The beneficial effects of this invention are as follows: Unlike conventional technologies, this invention discloses an environmental reliability testing system for electronic components. Received data is sent to a data processing module, which filters, smooths, and calibrates the data. The processed data is then sent to a process monitoring module, which presents the data in an intuitive manner. The intuitive data obtained from the process monitoring module is then verified by a reliability verification module, and the results are output to a result output module. This system enables automated testing based on pre-set environmental conditions and component operating thresholds. Furthermore, it allows for reliability analysis of the obtained experimental data using an improved Weibull distribution model, reducing testing complexity, lowering labor costs, and improving the reliability of analysis results.

[0015] The innovative aspects of this invention can also be: Multi-module collaborative control: The main control module enables precise coupled loading of stresses such as temperature (≤300℃), humidity, vibration (2Hz-3000Hz, displacement ≤51mm), and impact to simulate the complex environment under real working conditions. Improved Weibull distribution model: Introduces mechanisms such as adaptive parameter adjustment and multi-stress regression analysis to solve the evaluation accuracy problem of traditional models in scenarios with data dispersion and multi-factor coupling, reducing the lifetime prediction error rate to within 8%; Intelligent data processing architecture: Through a compatible data interface (USB3.2), real-time threshold comparison algorithm and visualization monitoring module, it achieves high-precision real-time monitoring with a 40% improvement in data acquisition efficiency and an alarm response time of less than 10 seconds; Remote collaboration capability: Based on standard communication protocols, a high-speed data channel is built to support remote parameter settings (latency < 50ms) and real-time synchronization of experimental results, meeting the collaboration needs in distributed R&D scenarios. Attached Figure Description

[0016] 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, wherein: Figure 1 This is a schematic diagram of the overall architecture of an environmental reliability testing system for electronic components according to the present invention. Figure 2 This is a schematic diagram of the overall control module of an environmental reliability testing system for electronic components according to the present invention. Figure 3 This is a schematic diagram of the data receiving module of an environmental reliability testing system for electronic components according to the present invention. Figure 4 This is a schematic diagram of the monitoring process module of an environmental reliability testing system for electronic components according to the present invention. Figure 5 This is a flowchart of the Weibull analysis method for an environmental reliability testing system for electronic components according to the present invention. Detailed Implementation

[0017] The preferred embodiments of the present invention will be described below with reference to the accompanying drawings. It should be understood that the embodiments described herein are for illustration and explanation only and are not intended to limit the present invention.

[0018] like Figures 1-5 As shown, an environmental reliability testing system for electronic components includes: The temperature control module evaluates the temperature tolerance and stability by simulating the operation of components under different temperature environments, including high temperature, low temperature, and temperature cycling; the high temperature is ≤300℃. The humidity control module simulates the effects of humidity in real-world environments to evaluate the moisture resistance of components in high-humidity environments, as well as their stability and tolerance under humid conditions. The vibration module simulates the stability and bearing capacity of components under vibration conditions, namely, frequency 2Hz-3000Hz, displacement ≤51mm, and load ≤3000Kg. The impact module assesses the reliability of components by simulating the impact conditions that components may encounter during transportation, installation, and use. The main control module coordinates and manages the entire experimental process, including data reception and processing, monitoring the experimental process, reliability verification, and result output. The main control module includes a data receiving module, a data processing module, a reliability verification module, a process monitoring module, and a result output module. The receiving module receives experimental data; the data processing module processes the received experimental data; the process monitoring module monitors the processed experimental data in real time; the reliability verification module analyzes the environmental reliability of electronic components; and the result output module transmits the analyzed results to a remote control center. The remote communication module uses standard communication protocols to establish a stable and high-speed data transmission channel between the experimental equipment and the remote control center, setting environmental parameters and threshold data, executing test programs, and receiving experimental results output by the main control module. The input terminals of the main control module are connected to the output terminals of the temperature control module, humidity control module, vibration module, and impact module, respectively; the output terminal of the remote communication module is connected to the input terminals of the temperature control module, humidity control module, vibration module, and impact module; and the main control module is connected to the remote control module.

[0019] In specific applications, MEMS sensors (integrating triaxial acceleration, temperature and humidity, strain gauges, and barometric pressure sensing) and MEMS sensor arrays (integrating triaxial acceleration, temperature, strain gauges, and barometric pressure sensing) are used to simultaneously acquire data (±0.1μs) and sense force at multiple scales in time and space. FBG fiber optic gratings are used to collect micro-strain data ≤1μm / m from the pins of the components, capturing vibration / impact stress.

[0020] In practical application examples, historical experimental data is used to train GMM to obtain the real-time threshold (μ±3σ) of each sensor data in real time; the 3σ criterion and Dixon criterion are introduced to automatically eliminate transient signals, such as the electric flash at the moment the amplitude module turns on and off.

[0021] In a specific embodiment, under high temperature and high humidity conditions, the rate of increase in temperature and humidity is dynamically adjusted (gradient ≤ 0.5℃ / min) based on the failure critical point predicted by the Weibull model. In a specific embodiment, during vibration testing, the frequency components of the input signal are analyzed in real time based on Fourier transform, and the frequency sweep mode (linear frequency sweep / logarithmic frequency sweep) of the vibration module is automatically switched.

[0022] In a specific embodiment, an LSTM is embedded into the reliability verification section. The LSTM receives time-series inputs such as the number of temperature cycles, vibration accumulation time, and humidity exposure time, and outputs a probability input of the remaining lifespan of the components. An attention mechanism is added to the LSTM to pay attention to critical stress inputs (e.g., the duration of temperature exposure has a 65% weighting on the impact of solder joint aging).

[0023] In real-world applications, a three-dimensional thermal-structural model is constructed using FEA, with the measured temperature field (±1℃) and vibration acceleration spectrum (VGG) as inputs to track processes such as solder joint fatigue cracking and chip temperature changes in real time. The optimal test environment (vibration acceleration amplitude margin magnified by 15%) is simulated and generated by a digital twin.

[0024] In this example, a simulation environment is built using Unity3D, and pre-experiments are conducted via a remotely controlled virtual experimental platform. For instance, virtual components are dragged into a temperature and humidity chamber to simulate the moisture absorption and expansion of materials; vibration table parameters are changed to display stress cloud diagrams of the virtual component pins (display accuracy 0.01 MPa). The central control unit calls edge computing libraries, such as NVIDIA Jetson AGX, and uses a lightweight neural network (Mobile NetV3) for local real-time compression: real-time compression of vibration data (compression ratio 10:1, distortion <5%); and a CNN is used to accurately capture impact signal events and distinguish between transportation impact events and accidental touch events.

[0025] In a specific embodiment, the key experimental data (curves, failure timestamp values, and Weibull model parameter values) are calculated using the SHA-256 algorithm, and the data is stored on the blockchain using a consortium blockchain (Hyperledger Fabric). Each block contains the device ID, experiment number, timestamp, and the hash value of the previous block, and cannot be altered.

[0026] In a specific embodiment, a Bayesian network is introduced into the improved Weibull model to establish a conditional probability table for the "temperature-humidity-vibration" stress factor. For example: In the example, under high humidity, the temperature-affected welding ignition rate coefficient increases by 2.3 times; under high frequency amplitude, the impact power equivalent path is reduced by 40%. For the demand for small test pieces of new components, a failure mode similarity space is constructed based on Siamese Network and a 5-shot learning algorithm is adopted. Only 5 failed test pieces are needed to learn the early failure (temperature anomaly correlation of gallium nitride power block) characteristics of the new device.

[0027] In this example, the vibration module uses magnetic levitation contactless drive (displacement control with an accuracy of ±0.01mm) and air-floating vibration isolation base (vibration isolation rate >95%@10Hz) to prevent friction and vibration transmission distortion caused by traditional mechanical vibration tables.

[0028] In practical applications, a semiconductor cooling chip energy recovery circuit is introduced into the temperature module to convert the heat generated during cooling into electrical energy (with an efficiency of no less than 15%); a servo motor power feedback is introduced into the vibration module to feed braking energy back to the system battery. In practical applications, innovative directions include... Figure 1 As shown.

[0029] Table 1 Technical Table of this Application

[0030] By incorporating the above innovative technological factors, a new generation of environmental reliability testing platform for electronic components is formed, integrating "reliable perception, intelligent control, data analysis, and trustworthy interaction." This platform completely breaks through the bottlenecks of traditional equipment, such as multi-stress coupling, complex failure prediction, and remote collaboration, providing a brand-new technical support for the research and development and quality management of future high-end electronic components.

[0031] Other embodiments of this application are described below.

[0032] As a further embodiment of the present invention, the temperature control module and the humidity control module are respectively composed of a temperature sensor and a humidity sensor, a control circuit and an actuator. The temperature sensor and the humidity sensor detect the ambient temperature and humidity. The control circuit receives the temperature and humidity signals, processes the data, compares the set value with the actual value, and outputs a control signal according to the comparison result. The actuator adjusts the temperature and humidity according to the signal output by the control circuit.

[0033] In the example above, a distributed fiber Bragg grating sensor network is installed in the temperature control module. Utilizing fiber wavelength encoding, it achieves multi-point temperature measurement (±0.1℃) and humidity measurement (integrated humidity probe, such as a polyimide film). Time-division multiplexing is used to access the temperature control circuit, enabling ultra-long (10km signal transmission) signal transmission without attenuation. The FBG sensor exhibits 3 times higher reliability and 10 times better electromagnetic interference resistance in high-temperature (≤300℃) environments, eliminating the need for a temperature adjustment circuit and saving on hardware design.

[0034] In practical applications, a self-developed Kalman filter is embedded into the control scheme. The temperature and humidity sensor signals are compared with the online estimated drift of cooling power, heating power, and humidifier oil spray volume (temperature drift ≤ 0.01℃, humidity drift ≤ 0.5%RH) using recursive least squares. If the sensor deviation exceeds the threshold, such as ±0.5℃ or ±3%RH, self-calibration is performed. Zeroing is performed using a high-precision constant temperature oil bath or salt solution. Compared with periodic calibration, the temperature and humidity accuracy is improved to ±0.2℃ and ±2%RH, and the calibration frequency is reduced by 70%, saving labor.

[0035] In a specific example, for high temperature and high humidity conditions, adjusting the cooling and dehumidification power allocation ratio (gradient ≤ 0.5℃ / min or 5%RH / min) at the failure threshold predicted by the Weibull model improves the temperature and humidity coupling control accuracy by 40%. In this specific example, a fuzzy PID controller is added, automatically selecting the appropriate operating mode as the temperature rises (rapid temperature rise control at low temperatures; PID control maintaining accuracy at high temperatures). Compared to traditional PID controllers, testing time is reduced by 25%, power consumption by 20%, and overshoot is reduced by 60% when temperature and humidity change. A semiconductor cooling module (TEC) and a heating module (TEG) are added to the temperature control system. Utilizing the Peltier and Seebeck effects, energy can be transferred between the two. If the TEC is cooling, the TEG generates electricity from waste heat (conversion efficiency ≥15%), which is then transferred to the system battery via DC-DC converter. If the TEC is cooling, the TEG is heating, and the energy transfer process is reversed, achieving energy savings. Approximately 120Wh of energy can be saved in 8 hours, the total system power consumption is reduced by 18%, and the size of the heat dissipation module is reduced by 30%.

[0036] In specific application cases, a three-dimensional temperature and humidity model is established based on CFD. Inputting measured temperature and humidity data (accuracy ±0.1℃ / ±1%) and wind speed (0.01m / s), the model displays the temperature gradient (accuracy ±0.3℃) and humidity gradient (accuracy ±3%) within the chamber in real time. Utilizing a digital twin, the optimal sensor distribution under extreme conditions (e.g., humidity sensors deployed in stagnant areas) is predicted, forecasting the range of temperature and humidity fluctuations and overcoming the limitation of single-sensor point measurements not being able to characterize location distribution. Temperature and humidity uniformity is improved to ±0.5℃ and ±5%RH, and experimental results are repeatable in 35% of cases.

[0037] In practical applications, an edge computer (NVIDIA Jetson Nano) is embedded in the control circuit, using a lightweight network (modelNetV3) to operate locally. Real-time compression of temperature and humidity waveform data (compression ratio 10:1, distortion <5%), reducing 90% of invalid data transmission; Using a convolutional neural network (CNN), abnormal temperatures such as sudden increases in humidity before condensation and aging of heating wires can be detected, with an alarm time of less than 2 seconds.

[0038] Data such as setpoints, calibration records, temperature and humidity curves, and failure timestamps are SHA-256 encoded and stored on the Union Chain (e.g., Hyperledger Fabric). Each block includes a device ID, timestamp, hash of the previous block, and digital signature, ensuring data immutability.

[0039] Application value: It supports the requirement of ISO17025 laboratory accreditation for traceability data, and increases the judicial verification of disputed data from 2 laboratories to cross-laboratory sharing and comparison of experimental data.

[0040] In practical applications, augmented reality simulations based on HoloLens2 allow remote experts to view the temperature and humidity environment (thermal maps) within a laboratory and make adjustments using a virtual control panel. For example: Gesture operation to adjust temperature and humidity settings (accuracy ±0.1℃ / ±1% RH); Voice commands trigger historical temperature and humidity curve comparison and analysis (supports data backtracking within 30 days).

[0041] Digital threading technology is used to track the entire experimental process of temperature and humidity control strategy design, parameter setting and data acquisition. The correspondence between the temperature and humidity control strategy design, parameter setting and experimental results is recorded digitally (as shown in the database, for example, the temperature and humidity curve corresponding to a batch of capacitor failure data).

[0042] Application value: 60% improvement in experimental data search rate, increased cross-project temperature and humidity control strategy reuse rate from 30% to 85%, and rapid generation of ISO compliant experimental reports. Experimental results are shown in Table 2.

[0043] Table 2 Experimental Schematic Table

[0044] As a further embodiment of the present invention, the vibration module uses a vibration sensor to monitor the vibration parameters of the components in the vibration experiment, namely the vibration frequency and amplitude, and uses a displacement sensor to measure the vibration displacement data of the components during the vibration process.

[0045] As a further embodiment of the present invention, the impact module uses an impact sensor to measure the magnitude and duration of the impact force received by the component in the impact experiment, a strain sensor to measure the strain of the component when it is impacted, and an acceleration sensor to record the peak value and changes of the acceleration of the component when it is impacted in real time.

[0046] In a specific embodiment, the vibration frequency domain features (frequency / amplitude) and time domain displacement trajectory are acquired simultaneously to construct a full-cycle kinematic profile of the vibration. As a further embodiment of the present invention, the data receiving module includes a compatible data interface, an information encoding module, a receiving path planning module, and an information decoding module; the data processing module performs filtering, smoothing, and calibration processing on the data; the process monitoring module includes an information extraction module, a threshold comparison module, an alarm module, and a visualization display module; the reliability verification module performs lifetime data analysis using an improved Weibull distribution model, which comprises a parameter setting module, a Weibull ratio adjustment module, a Weibull regression module, a Weibull adaptive adjustment module, a reliability result detection module, and a reliability result output module.

[0047] As a further embodiment of the present invention, the compatible data interface uses a USB 3.2 interface to achieve adaptation between devices and instruments using different data interface standards and supports high-speed data transmission; the information encoding module is used to encode various information during the experiment, using the UTF-8 encoding standard to convert temperature, humidity, voltage, current, frequency, and acceleration data into a unified format and standard, and compressing the data to reduce storage space usage; the receiving path planning module plans the data receiving path according to the sensor's location, data type, and real-time requirements; the data decoding module decodes the data according to the encoding method and format, and then appends a checksum to the target data for verification.

[0048] As a further embodiment of the present invention, the information extraction module is used to extract information from the data processed by the data processing module, including voltage, current, temperature, humidity, frequency and acceleration data; the threshold comparison module is used to compare the extracted data with the normal operating range values ​​of the components to perform threshold comparison; the alarm module is used to issue an alarm when the received data exceeds the threshold; and the visualization display module is used to perform charting and curve processing on the data.

[0049] As a further embodiment of the present invention, the reliability verification module performs data analysis using an improved Weibull distribution model; the parameter setting module is used to set constants in the improved Weibull distribution model; the Weibull proportional adjustment module is used to adjust the proportional parameters in the improved Weibull distribution model; the Weibull regression module is used to analyze the failure modes of electronic components under specific environmental conditions; the Weibull adaptive adjustment module is used to adaptively adjust experimental conditions or parameters based on the failure data of electronic components under specific environments and the Weibull distribution characteristics; the reliability result detection module is used to evaluate the performance and reliability level under specific environmental conditions based on the improved Weibull distribution model; the reliability result output module is used to output the reliability level result obtained by the reliability result detection module; the probability density of the improved Weibull stepwise method is: In formula (1), x is a random variable. 'k' is the scale parameter, and 'k' is the shape parameter. When the scale parameter is the same, the larger the shape parameter k, the more concentrated the failure data of the sample; conversely, the smaller the shape parameter k, the larger the area where the failure data falls, and the more dispersed the sample failure data. The steps of the Weibull analysis method are as follows: Determine sample size and time: Determine the sample size and testing time based on the actual usage environment of the product, potential problems, and expected lifespan; Collect fault data: Test the product according to the set sample size and test time, and record the fault data; Plotting the failure rate curve: Plot the product's failure rate curve based on the failure data; Fitting the failure rate curve: The failure rate curve is fitted using the Weibull distribution function to obtain the Weibull distribution parameters; Analyzing the results and determining reliability: The reliability of the product is determined by analyzing the fitting results. A specific example utilizes an improved Weber distributed model, incorporating a closed-loop feedback control system to establish a dynamically adjustable verifiable reliability test. The test parameters included: temperature (65°C to -65°C); humidity (10%RH, 100%); vibration (5~2000Hz); and electrical stress: voltage ±15%.

[0050] This transforms traditional static Weibull analysis into a dynamic testing ecosystem of "sensing-failure response-model evolution," enabling a leap from "passive statistics" to "active prediction" in electronic component reliability testing and verification. It is important to note that for sample sizes <20, the Bootstrap resampling algorithm should be used to ensure stable parameter estimation for small samples.

[0051] The test environment and controlled objects are shown in Table 3.

[0052] Table 3. Experimental Schematic Table

[0053] Sample selection: A certain type of power module (n=50, of which the small sample group n=15 and the regular group n=35) Failure criteria: Leakage current > 100μA or voltage gain decrease > 15%. Data collection: Failure time (accurate to 0.1 hours). The environmental stress combination at failure (temperature / humidity / vibration frequency / voltage fluctuation). Failure modes (solder joint cracking / dielectric breakdown / semiconductor failure) are grouped as shown in Table 4.

[0054] Table 4 Experimental Data

[0055] The experimental data for the standard sample group (n=35) are shown in Table 5.

[0056] Table 5 Experimental Data

[0057] Experiments revealed that the improved model can be corrected for environmental stress (e.g., temperature acceleration factor AF = 1.35). The parameters are closer to actual working conditions, and the KS test pass rate is improved by 17%. The small sample group (n=15) is shown in Table 6.

[0058] Table 6. Schematic diagram of test records

[0059] Experiments revealed that the Bootstrap algorithm reduces the standard deviation of parameters in small samples by 60% and increases the equivalent sample size by 87%, thus resolving the estimation bias problem of traditional methods when n < 20. Reliability metrics are compared in Table 7.

[0060] Table 7. Typical stress scenarios (25℃ / 50% RH / 50Hz)

[0061] Table 8. Ultimate stress scenarios (125℃ / 85% RH / 1500Hz)

[0062] Stress adaptive adjustment: If the failure time of 3 consecutive samples is <λ / 2, the amplitude is automatically reduced by 10% (to prevent excessive acceleration of failure), the test time is increased by 15% but the data integrity rate is 30%; Model update: For every 5 missing data points, the parameters are re-estimated to new parameters. The update rate is k=2 / hour to ensure the model's applicability to time-varying conditions (R²≥0.95).

[0063] When Boot resampling is enabled with n=15, the standard deviation of the estimated parameters is only 0.4 times that of the traditional parameter estimates, which effectively solves the small sample problem in the early reliability verification of new devices. The equivalent sample size was increased from 40% to 85% to utilize small sample data, and the parameter estimation error rate was reduced from 25% to 7%. The improved model calculates the stress weight in the environment in real time (temperature affects 65% in high humidity and high temperature scenarios), and the accuracy error rate of predicting failure time under extreme conditions was reduced from 32% to 8%.

[0064] Closed-loop feedback control achieves a closed-loop balance of "stress loading - failure response - model correction", shortening the test cycle by 20%, increasing the inspection rate of key failure modes of weld joints and thermal fatigue by 45%, increasing the KS detection pass rate from 78% to 92%, improving the accuracy of MTBF calculation by 15.7%, and achieving a 92% agreement rate between 10,000-hour reliability prediction and actual test results.

[0065] The "Environmental Awareness - Model Evolution" testing and verification ecosystem has been established, elevating reliability verification from "post-event verification" to "pre-event prediction," while simultaneously providing early warnings of failure time up to 48 hours in advance (e.g., k < 1.2). This experiment verifies that an improved Weibull distribution model combined with closed-loop feedback control enhances the accuracy and efficiency of reliability verification for electronic components in complex environments, providing a reference method for evaluating the "environmental adaptability" of high-end electronic components. The reliability function of the electronic component is: ;In formula (2), This represents the number of failures by time t. Indicates the total number of test products. This represents the output characteristic value of the Weibull distribution function, where M represents the type of experiment. Indicates from unit time to Total number of products within the time range This represents the number of products at time i. The time interval is represented by i, which represents the unit of time from the start of the experiment. The failure rate is the probability that a product will fail within a unit of time after time t. The failure probability function is: ;In formula (3), This represents the number of products with normal performance at time t. The number of failed products within the monitoring time interval. The output characteristic value represents the failure probability function, and W represents a constant value under different environmental conditions. This represents the number of products that failed instantaneously at time t; The failure rate of electronic components follows an exponential distribution, and the relationship between reliability and failure rate is as follows: ;In formula (4), Indicates failure rate. 'b' represents a constant value. Multi-stage aging tests were conducted to simulate different stress levels that the product might encounter throughout its entire lifespan. Accelerated life testing was used to expedite the product failure process and collect sufficient failure data in a shorter time. Ten experimental samples were subjected to accelerated life tests under different environmental conditions, and the failure data are shown in Table 9.

[0066] Table 9 Failure Data

[0067] As a further embodiment of the present invention, the data receiving module receives data via a high-speed data transmission interface and protocol; the data processing module performs filtering, smoothing, and calibration processing on the data; the process monitoring module monitors environmental parameters, experimental equipment status, and experimental process; and uses graphical interface technology to visually present the processed data in the form of curves and save the data to a database or file. Simultaneously, it issues alarms for abnormal values. The reliability verification module analyzes and verifies the curves generated by the process monitoring module using Welb distribution; and the result output module outputs experimental conclusions based on the analysis results.

[0068] As a further embodiment of the present invention, the process monitoring module operates as follows: Data processed by the data processing module is sent to the information extraction module; an initial curve is drawn using the matplotlib plot function; data is periodically acquired via a timer; and when new real-time data arrives, the data points on the curve and the coordinate axis range of the curve are updated. Data detected under different environments is compared with thresholds set according to the performance standards of electronic components and experimental requirements. When the detected feature value exceeds the preset threshold, the early warning module triggers an early warning mechanism and generates corresponding early warning information.

[0069] The parameter setting module is controlled by the ADE7758 chip to set the name, type, default value, allowed value range or control command of the Weibull distribution model parameters; The Weibull scaling module adjusts the proportions of the Weibull distribution model parameters through parameter estimation and optimization algorithms. The Weibull regression module fits the Weibull distribution parameters by maximizing the likelihood function; The Weibull adaptive adjustment module changes the model structure by adjusting model weights, changing parameter update frequency, or responding to specific events. The reliability result detection is achieved by filtering data parameter information through failure time, environmental conditions, life test, sequential test, and environmental stress. The reliability output module outputs reliability results by discarding outliers, adding missing values, converting to a data format, and outputting the data using the mean, median, or variance.

[0070] In specific embodiments, the Weibull regression module generally refers to a tool module in statistical analysis software or reliability engineering software used to perform Weibull regression analysis. Weibull regression is a statistical method used to analyze the parameters of a certain Weibull distribution to obtain the reliability of a certain product. A Weibull regression module may include the following parts: Data import and preprocessing: Supports the import of lifetime data and preprocesses the data, including removing outliers and handling missing data.

[0071] The Weibull regression algorithm module includes the Weibull regression analysis algorithm, which may estimate the maximum value of the parameters (β) and proportionality parameter (η) of the maximum likelihood function and the possible (location parameter (λ or θ)) through maximum likelihood estimation. The algorithm may be based on different statistical methods, such as maximum likelihood estimation (MLE), Bayesian estimation, etc. Parameter estimation and optimization may estimate the parameters of the optimal Weibull distribution-fitted model and may provide an optimization module to improve the goodness of fit. Results output and interpretation include the estimated parameters of the Weibull distribution, goodness of fit (e.g., χ² statistic, etc.), and possible graphical output, including Weibull distribution density curves, cumulative distribution curves, etc. Model validation and evaluation may employ methods to verify the accuracy and robustness of the Weibull regression model, including cross-validation, residual analysis, etc.

[0072] The Weibull adaptive adjustment module can acquire real-time data from the actual operating environment of the product (such as real-time running time, real-time temperature, and load) to influence product reliability. Parameter estimation and optimization employ statistical algorithms. Maximum likelihood estimation (MLE) and particle swarm optimization (PSO) are used to automatically estimate the Weibull distribution (β), proportional (η) parameters, and other parameters. Parameters are adjusted based on loss functions such as goodness-of-fit tests or prediction errors, continuously updating the fit. The model is adjusted using newly acquired data to adapt to possible environmental changes or aging. Adaptability may include dynamically adjusting model weights, adjusting parameter update rates, and responding to events (such as failures or maintenance) to modify the model.

[0073] While specific embodiments of the present invention have been described above, those skilled in the art should understand that these specific embodiments are merely illustrative. Those skilled in the art can omit, substitute, and modify the details of the above methods and systems in various ways without departing from the principles and essence of the present invention. For example, combining the above method steps to perform substantially the same function and achieve substantially the same result according to substantially the same method falls within the scope of the present invention. Therefore, the scope of the present invention is defined only by the appended claims.

Claims

1. An environmental reliability testing system for electronic components, characterized in that: include: The temperature control module evaluates the temperature tolerance and stability by simulating the operation of components under different temperature environments, including high temperature, low temperature, and temperature cycling; the high temperature is ≤300℃. The humidity control module simulates the effects of humidity in real-world environments to evaluate the moisture resistance of components in high-humidity environments, as well as their stability and tolerance under humid conditions. The vibration module simulates the stability and bearing capacity of components under vibration conditions, namely, frequency 2Hz-3000Hz, displacement ≤51mm, and load ≤3000Kg. The impact module assesses the reliability of components by simulating the impact conditions that components may encounter during transportation, installation, and use. The overall control module coordinates and manages the entire experimental process, including data reception and processing, monitoring the experimental process, reliability verification, and result output. The overall control module includes a data receiving module, a data processing module, a reliability verification module, a process monitoring module, and a result output module. The receiving module receives experimental data; the data processing module processes the received experimental data; the process monitoring module monitors the processed experimental data in real time; the reliability verification module analyzes the environmental reliability of electronic components; and the result output module transmits the analyzed results to a remote control center. The reliability verification module uses an improved Weibull distribution model for lifetime data analysis. This improved Weibull distribution model includes a parameter setting module, a Weibull ratio adjustment module, a Weibull regression module, a Weibull adaptive adjustment module, a reliability result detection module, and a reliability result output module. The remote communication module uses standard communication protocols to establish a stable and high-speed data transmission channel between the experimental equipment and the remote control center, setting environmental parameters and threshold data, executing test programs, and receiving experimental results output by the main control module. The input terminals of the main control module are connected to the output terminals of the temperature control module, humidity control module, vibration module, and impact module, respectively; the output terminal of the remote communication module is connected to the input terminals of the temperature control module, humidity control module, vibration module, and impact module; and the main control module is connected to the remote control module.

2. The environmental reliability testing system for electronic components according to claim 1, characterized in that: The temperature control module and humidity control module are respectively composed of a temperature sensor and a humidity sensor, a control circuit and an actuator. The temperature sensor and humidity sensor detect the ambient temperature and humidity. The control circuit receives the temperature and humidity signals, processes the data, compares the set value with the actual value, and outputs a control signal based on the comparison result. The actuator adjusts the temperature and humidity according to the signal output by the control circuit.

3. The environmental reliability testing system for electronic components according to claim 1, characterized in that: The vibration module uses vibration sensors to monitor the vibration parameters of the components during vibration experiments, namely vibration frequency and amplitude, and uses displacement sensors to measure the vibration displacement data of the components during the vibration process.

4. The environmental reliability testing system for electronic components according to claim 1, characterized in that: The impact module uses an impact sensor to measure the magnitude and duration of the impact force on the components during the impact experiment, a strain sensor to measure the strain of the components when they are impacted, and an acceleration sensor to record the peak value and changes in acceleration of the components when they are impacted in real time.

5. The environmental reliability testing system for electronic components according to claim 1, characterized in that: The data receiving module includes a compatible data interface, an information encoding module, a receiving path planning module, and an information decoding module; the data processing module performs filtering, smoothing, and calibration processing on the data; the process monitoring module includes an information extraction module, a threshold comparison module, an alarm module, and a visualization display module.

6. The environmental reliability testing system for electronic components according to claim 5, characterized in that: The compatible data interface uses a USB 3.2 interface to enable adaptation between devices and instruments using different data interface standards and supports high-speed data transmission; the information encoding module is used to encode various information during the experiment, using the UTF-8 encoding standard to convert temperature, humidity, voltage, current, frequency, and acceleration data into a unified format and standard, and compressing the data to reduce storage space usage; the receiving path planning module plans the data receiving path according to the sensor's location, data type, and real-time requirements. The data decoding module decodes the data according to its encoding method and format, and then appends a checksum to the target data for verification.

7. The environmental reliability testing system for electronic components according to claim 5, characterized in that: The information extraction module is used to extract information from the data processed by the data processing module, including voltage, current, temperature, humidity, frequency, and acceleration data; the threshold comparison module is used to compare the extracted data with the normal operating range values ​​of the components to perform threshold comparison; the alarm module is used to issue an alarm when the received data exceeds the threshold; and the visualization display module is used to perform charting and curve processing on the data.

8. The environmental reliability testing system for electronic components according to claim 5, characterized in that: The parameter configuration module is used to improve the constant settings in the Weibull distribution model; the Weibull scaling module is used to adjust the scaling parameters of the improved Weibull distribution model; the Weibull regression model is used to analyze the failure modes of electronic components under specific environmental conditions; the Weibull adaptive adjustment module is used to adaptively adjust test conditions or test parameters based on the failure data and Weibull distribution characteristics of electronic components under specific environments; the reliability result detection module is used to detect the performance and reliability level of the improved Weibull distribution model under specific conditions; the reliability level result output section is used to output the reliability level results (output the reliability result detection section); the probability density of the improved Weibull stepwise method is: In formula (1), x is a random variable. 'k' is the scale parameter, and 'k' is the shape parameter. When the scale parameter is the same, the larger the shape parameter k, the more concentrated the failure data of the sample; conversely, the smaller the shape parameter k, the larger the area where the failure data falls, and the more dispersed the sample failure data. The steps of the Weibull analysis method are as follows: Determine sample size and time: Determine the sample size and testing time based on the actual usage environment of the product, potential problems, and expected lifespan; Collect fault data: Test the product according to the set sample size and test time, and record the fault data; Plotting the failure rate curve: Plot the product's failure rate curve based on the failure data; Fitting the failure rate curve: The failure rate curve is fitted using the Weibull distribution function to obtain the Weibull distribution parameters; Analyze the results and assess reliability: Evaluate the reliability of the product by analyzing the fitting results.

9. The environmental reliability testing system for electronic components according to claim 5, characterized in that: The reliability function of the electronic component is: ;In formula (2), This represents the number of failures by time t. Indicates the total number of test products. This represents the output characteristic value of the Weibull distribution function, where M represents the type of experiment. Indicates from unit time to Total number of products within the time range This represents the number of products at time i. The time interval is represented by i, which represents the unit of time from the start of the experiment. The failure rate is the probability that a product will fail within a unit of time after time t. The failure probability function is: ;In formula (3), This represents the number of products with normal performance at time t. The number of failed products within the monitoring time interval. The output characteristic value represents the failure probability function, and W represents a constant value under different environmental conditions. This represents the number of products that failed instantaneously at time t; The failure rate of electronic components follows an exponential distribution, and the relationship between reliability and failure rate is as follows: ;In formula (4), Indicates failure rate. b represents a constant value.

10. The environmental reliability testing system for electronic components according to claim 1, characterized in that: The parameter setting module is controlled by the ADE7758 chip to set the name, type, default value, allowed value range or control command of the Weibull distribution model parameters; The Weibull scaling module adjusts the proportions of the Weibull distribution model parameters through parameter estimation and optimization algorithms. The Weibull regression module fits the Weibull distribution parameters by maximizing the likelihood function; The Weibull adaptive adjustment module changes the model structure by adjusting model weights, changing parameter update frequency, or responding to specific events. The reliability result detection is achieved by filtering data parameter information through failure time, environmental conditions, life test, sequential test, and environmental stress. The reliability result output module outputs data information by removing outliers, filling in missing data, converting data formats, and calculating the mean, median, or variance.

Images