Relay contact performance analysis and test method

Abstract

The invention provides a relay contact performance analysis and test method. The relay contact performance analysis and test method comprises the steps that S1, a test host loads a preset test parameter template; s2, preparing a test fixture I and a test fixture II; s3, setting the state of a control relay K1; s4, setting the state of a control relay K2; s5, setting the state of a control relay K3; s6, installing a tested relay; s7, the test signal source injects a test signal; s8, a high-speed analog-to-digital converter collects a contact signal; s9, the test host integrates three algorithms to process the digital signals, and calculates and optimizes contact performance parameters; and S10, executing an electrical life test of the tested relay. The core of the invention is to realize full-automatic and high-precision testing of multiple parameters of the relay contact through cooperation of hardware configuration and an algorithm.

Description

Technical Field

[0001] This invention relates to the field of relay technology, and in particular to a method for analyzing and testing the performance of relay contacts. Background Technology

[0002] Relays are core components in rail transit control systems. The reliability of their contacts directly affects the stable operation of the entire rail transit system, and contact life is a key indicator for evaluating relay performance. During the switching process, contacts generate phenomena such as arcing and bouncing. The parameters related to these processes (such as arc duration, bouncing period, and contact resistance stability) are the core factors that determine the contact wear rate, oxidation degree, and ultimate lifespan.

[0003] Current technologies for relay testing primarily focus on fundamental indicators such as switching life, static contact resistance, and insulation performance, lacking precise measurement of dynamic parameters directly related to lifespan, such as arc time and contact bounce time. Furthermore, existing testing methods have significant limitations: first, fixed test modes make it difficult to adapt to different types (AC / DC) and load requirements (inductive / capacitive) relays in rail transit scenarios; second, the testing process relies on manual intervention, from fixture replacement to data statistics, which is inefficient and prone to human error; third, there is a lack of dynamic calibration and consistency verification of test data, making test results susceptible to factors such as ambient temperature and power fluctuations, resulting in insufficient reliability; and fourth, comprehensive analysis of multiple parameters is impossible, making it difficult to establish the correlation between parameters and contact lifespan, and failing to provide comprehensive support for relay performance evaluation. Summary of the Invention

[0004] This invention provides a method for analyzing and testing the performance of relay contacts. The core of this method is to achieve fully automatic and high-precision testing of multiple parameters of relay contacts through hardware configuration and algorithm collaboration.

[0005] To achieve the above objectives, the present invention adopts the following technical solution: A method for analyzing and testing the performance of relay contacts, comprising: S1: After the test host initializes and performs a self-test, it loads a preset test parameter template, including test mode, load type, power supply type, cycle interval, and number of electrical life test cycles. S2: Configure test fixture one and test fixture two, so that test fixture one is adapted to the relay under test and test fixture two is equipped with a corresponding load unit; S3: Configure the state of control relay K1 according to the test mode, and confirm that the state matches; S4: Configure the state of control relay K2 according to the load type, connect it to the load unit of test fixture two and verify its effectiveness; S5: Configure the state of control relay K3 according to the power supply type, and confirm the state matching; S6: Install the relay under test through test fixture one; S7: Control the test signal source to generate a test signal, inject it into the target contact of the relay under test, and construct a test circuit containing two load units of the test fixture; S8: Collects the contact response signals in the test circuit, converts them into digital signals, and stores them in the test host; S9: The test host integrates three collaborative algorithms to perform closed-loop processing on the digital signal. The three collaborative algorithms include a touch feature point recognition and basic parameter extraction algorithm, a dynamic error calibration algorithm, and a parameter consistency verification and feedback optimization algorithm. First, the basic parameters are calculated by the touch feature point recognition and basic parameter extraction algorithm. Then, the basic parameters are calibrated by the dynamic error calibration algorithm to obtain calibrated parameters. Finally, the consistency verification and feedback optimization algorithm is used to verify the consistency of the calibrated parameters. Based on the verification results, the core parameters of the first two types of algorithms are optimized in reverse until the calibrated parameters that meet the accuracy requirements are output. S10: Keep the control relay state unchanged, control the relay under test to operate according to the cycle interval and the number of electrical life test cycles, obtain the calibrated parameters after each cycle, compare them with the failure characteristics and record the abnormality judgment result.

[0006] In this specification, the preset test parameter template also includes failure characteristics, predetermined performance values, and statistical standards for consistency among multiple samples; the analysis and testing methods for relay contact performance also include: S11: Summarize the calibration parameters and anomaly judgment results, and generate a single relay performance evaluation report by combining the performance preset values; S12: When multiple samples need to be tested, test multiple samples according to the procedures from S6 to S11, and generate a multi-sample consistency report based on the consistency statistics standard. S13: After disconnecting the test power, removing the relay under test, clearing temporary data, and storing the report, the test system enters standby mode.

[0007] In this specification, the test host, test fixture one, test fixture two, high-speed analog-to-digital converter, test signal source, selection switch two, selection switch one, control relay K1, control relay K2, control relay K3, AC power supply and DC power supply constitute the test system; the control relay K1 includes control relays K1-1, K1-2 and K1-3; Signal Injection and Synchronization Link: The signal output of the test signal source is divided into two paths. One path is connected to the signal input of selector switch one, and the other path is connected to the synchronization signal input of the high-speed analog-to-digital converter (to achieve timing synchronization between test signal output and signal acquisition). The signal output of selector switch one is connected to the test signal injection interface of test fixture one. Test fixture one has a built-in mounting position for the relay under test (K test). The contact circuit of the relay under test is connected in series with the contact terminals of control relays K1-1, K1-2, and K1-3 to realize the switching of the injection path of the test signal to the contacts of the relay under test. Signal acquisition link: The contact signal acquisition interface of test fixture one is connected to the signal input terminal of selector switch two, the signal output terminal of selector switch two is connected to the signal acquisition input terminal of high-speed analog-to-digital converter, the digital signal output terminal of high-speed analog-to-digital converter establishes a data transmission connection with the test host, and at the same time, the high-speed analog-to-digital converter and the test signal source maintain synchronous signal interaction, forming a synchronous acquisition and transmission path for the response signal of the relay under test contact. Power supply and load circuit: The output terminals of AC power supply and DC power supply are connected to the contact terminals of control relay K3. The output terminal of K3 is connected to the power input interface of test fixture one, providing working power for test fixture one, control relays K1-1 / K1-2 / K1-3 and test fixture two. The load interface of test fixture one is connected to the signal terminal of test fixture two. Test fixture two has built-in resistor-capacitor parallel load unit and resistor-inductor series load unit. The control interface of test fixture two is connected to the contact terminals of control relay K2. The connection control of the load circuit is realized by the on and off of K2. Control Link: The test host establishes bidirectional control and data transmission connections (including synchronous timing control and signal parameter configuration) with the high-speed analog-to-digital converter and the test signal source, respectively. At the same time, the test host establishes control connections with control relays K1-1, K1-2, K1-3, K2, and K3 to realize the switching of the on / off state of each relay, the adjustment of the output parameters of the test signal, and the reception and processing of the acquired data.

[0008] In this specification, the contact feature point identification and basic parameter extraction algorithm identifies key feature points of contact action by setting voltage change threshold and stable duration threshold. When the voltage change in the digital signal exceeds the voltage change threshold and the duration reaches the stable duration threshold, it is determined to be a valid feature point. The contact bounce time is calculated based on the timestamp difference of the valid feature points. The contact resistance is calculated by the ratio of the voltage amplitude of the test signal to the circuit current value. During load testing, feature points corresponding to the generation and extinction of the arc are additionally identified, and the arc duration is calculated.

[0009] In this specification, the calibration process of the dynamic error calibration algorithm includes zeroing calibration and drift compensation. Zeroing calibration is automatically performed during the contact disconnection phase of each test cycle, collecting the zero signal measurement value of the test circuit and comparing it with the standard zero impedance parameter to obtain the system fixed error. Drift compensation calculates the signal drift based on the test duration and the change in ambient temperature, and obtains the calibrated parameters through the calculation logic of "basic parameters - system fixed error - signal drift".

[0010] In this specification, the number of consecutive preset test cycles in the parameter consistency verification and feedback optimization algorithm is 5-20. The preset consistency thresholds include resistance parameter thresholds and time parameter thresholds. If the variance of the calibrated parameters is less than the corresponding threshold, the core parameters of the current algorithm are maintained. If the variance of the resistance parameter exceeds the threshold, the recognition weight of the contact feature point recognition and basic parameter extraction algorithm is reduced, while the calibration coefficient of the dynamic error calibration algorithm is increased. If the variance of the time parameter exceeds the threshold, the stable duration threshold is shortened, and the basic parameter extraction and calibration steps are re-executed.

[0011] In this specification, the verification results of the parameter consistency verification and feedback optimization algorithm are fed back to the first two types of algorithms in the form of quantified coefficients. During feedback optimization, the rule of "consistency qualified coefficient → maintain algorithm parameters, consistency unqualified coefficient → adjust algorithm parameters according to preset ratio" is followed. The adjustment range of the identification threshold is 5%-15% of the initial value, and the adjustment range of the calibration coefficient is 10%-25% of the initial value, to ensure the stability and effectiveness of the algorithm parameter adjustment.

[0012] In this specification, the contact action time in the basic parameters is the time difference between "the moment when the test host issues the pick-up / drop command" and "the moment when the effective closed / open feature point is first identified", the contact bounce time is the time difference between "the moment when the first effective closed feature point is identified" and "the moment when the last effective bounce feature point is identified", and the contact resistance is the ratio of the average voltage value to the average current value of the test signal during the stable contact closure phase, ensuring the accuracy of the basic parameter calculation.

[0013] In this specification, the verification results of the parameter consistency verification and feedback optimization algorithm are linked with the anomaly judgment results of S10. If the consistency verification result is unqualified and the anomaly judgment result is "abnormal" for three consecutive cycles, it is determined that the relay under test has a potential fault. The test host marks the sample and increases the adjustment range of the calibration coefficient in subsequent cycles to improve the pertinence of parameter calibration.

[0014] In this specification, the multi-sample consistency statistical standard is calculated based on the calibrated parameters of all qualified samples, including the parameter mean, standard deviation, and coefficient of variation. Before calculation, sample data with unqualified consistency verification results are removed. If the deviation of the calibrated parameter of a single sample from the mean exceeds 3 times the standard deviation, it is judged as an abnormal sample and is not included in the statistical scope, thus ensuring the accuracy of the multi-sample consistency report.

[0015] In summary, the present invention has at least the following beneficial effects: Comprehensive testing capabilities: Breaking through existing technological limitations, it enables simultaneous testing and comprehensive analysis of key parameters across multiple dimensions, such as contact arc time, bounce time, action time, and contact resistance, fully covering the core factors affecting contact lifespan and providing comprehensive data support for relay performance evaluation.

[0016] Enhanced adaptability: Through modularly designed test fixtures and switchable power and load control logic, it can flexibly adapt to various relays with different pin distributions, power types (AC / DC), and load characteristics (inductive / capacitive) in the rail transit field, meeting diverse testing needs without major modifications to the test system.

[0017] Test automation and intelligent upgrade: Achieve full-process automation from test parameter configuration, signal injection, data acquisition to parameter calculation, anomaly detection, and report generation, significantly reducing manual intervention and repetitive labor intensity; Through the closed-loop collaboration of three major algorithms, achieve dynamic calibration and consistency verification of data, effectively offsetting system errors and signal drift, and improving the stability and reliability of test results.

[0018] Supporting accurate evaluation: Through multi-parameter comprehensive analysis and multi-sample consistency testing, the distribution pattern of contact performance parameters can be established, and parameter anomalies and potential faults can be accurately identified. This provides a scientific basis for relay life prediction, quality screening and reliability optimization, and helps to improve the overall stability of rail transit control systems. Attached Figure Description

[0019] To more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings used in the following description of the embodiments will be briefly introduced. 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.

[0020] Figure 1 This is a schematic diagram illustrating the analysis and testing methods for the relay contact performance involved in this invention.

[0021] Figure 2 This is a schematic diagram of the testing system involved in the present invention.

[0022] Figure 3 This is a schematic diagram of the relay contact testing process involved in this invention.

[0023] Figure 4 This is a schematic diagram of the algorithm flow involved in this invention. Detailed Implementation

[0024] In the following description, only certain exemplary embodiments are briefly described. As those skilled in the art will recognize, the described embodiments can be modified in various ways without departing from the spirit or scope of the embodiments of the invention. Therefore, the drawings and description are considered to be exemplary in nature and not restrictive.

[0025] The following disclosure provides many different implementations or examples for carrying out different structures of the embodiments of the present invention. To simplify the disclosure of the embodiments of the present invention, specific examples of components and arrangements are described below. Of course, these are merely examples and are not intended to limit the embodiments of the present invention. Furthermore, reference numerals and / or reference letters may be repeated in different examples of the embodiments of the present invention; such repetition is for simplification and clarity and does not in itself indicate a relationship between the various implementations and / or arrangements discussed.

[0026] The embodiments of the present invention will now be described in detail with reference to the accompanying drawings.

[0027] like Figure 1 As shown, this embodiment provides a method for analyzing and testing the performance of relay contacts, including: S1: After the test host initializes and performs a self-test, it loads a preset test parameter template, including test mode, load type, power supply type, cycle interval, and number of electrical life test cycles. S2: Configure test fixture one and test fixture two, so that test fixture one is adapted to the relay under test and test fixture two is equipped with a corresponding load unit; S3: Configure the state of control relay K1 according to the test mode, and confirm that the state matches; S4: Configure the state of control relay K2 according to the load type, connect it to the load unit of test fixture two and verify its effectiveness; S5: Configure the state of control relay K3 according to the power supply type, and confirm the state matching; S6: Install the relay under test through test fixture one; S7: Control the test signal source to generate a test signal, inject it into the target contact of the relay under test, and construct a test circuit containing two load units of the test fixture; S8: Collects the contact response signals in the test circuit, converts them into digital signals, and stores them in the test host; S9: The test host integrates three collaborative algorithms to perform closed-loop processing on the digital signal. The three collaborative algorithms include a touch feature point recognition and basic parameter extraction algorithm, a dynamic error calibration algorithm, and a parameter consistency verification and feedback optimization algorithm. First, the basic parameters are calculated by the touch feature point recognition and basic parameter extraction algorithm. Then, the basic parameters are calibrated by the dynamic error calibration algorithm to obtain calibrated parameters. Finally, the consistency verification and feedback optimization algorithm is used to verify the consistency of the calibrated parameters. Based on the verification results, the core parameters of the first two types of algorithms are optimized in reverse until the calibrated parameters that meet the accuracy requirements are output. S10: Keep the control relay state unchanged, control the relay under test to operate according to the cycle interval and the number of electrical life test cycles, obtain the calibrated parameters after each cycle, compare them with the failure characteristics and record the abnormality judgment result.

[0028] In some embodiments, the preset test parameter template further includes failure characteristics, predetermined performance values, and multi-sample consistency statistical standards; the method for analyzing and testing the relay contact performance further includes: S11: Summarize the calibration parameters and anomaly judgment results, and generate a single relay performance evaluation report by combining the performance preset values; S12: When multiple samples need to be tested, test multiple samples according to the procedures from S6 to S11, and generate a multi-sample consistency report based on the consistency statistics standard. S13: After disconnecting the test power, removing the relay under test, clearing temporary data, and storing the report, the test system enters standby mode.

[0029] In some embodiments, such as Figure 2 As shown, the test host, test fixture one, test fixture two, high-speed analog-to-digital converter, test signal source, selection switch two, selection switch one, control relay K1, control relay K2, control relay K3, AC power supply and DC power supply constitute the test system; the control relay K1 includes control relays K1-1, K1-2 and K1-3; Signal Injection and Synchronization Link: The signal output of the test signal source is divided into two paths. One path is connected to the signal input of selector switch one, and the other path is connected to the synchronization signal input of the high-speed analog-to-digital converter (to achieve timing synchronization between test signal output and signal acquisition). The signal output of selector switch one is connected to the test signal injection interface of test fixture one. Test fixture one has a built-in mounting position for the relay under test (K test). The contact circuit of the relay under test is connected in series with the contact terminals of control relays K1-1, K1-2, and K1-3 to realize the switching of the injection path of the test signal to the contacts of the relay under test. Signal acquisition link: The contact signal acquisition interface of test fixture one is connected to the signal input terminal of selector switch two, the signal output terminal of selector switch two is connected to the signal acquisition input terminal of high-speed analog-to-digital converter, the digital signal output terminal of high-speed analog-to-digital converter establishes a data transmission connection with the test host, and at the same time, the high-speed analog-to-digital converter and the test signal source maintain synchronous signal interaction, forming a synchronous acquisition and transmission path for the response signal of the relay under test contact. Power supply and load circuit: The output terminals of AC power supply and DC power supply are connected to the contact terminals of control relay K3. The output terminal of K3 is connected to the power input interface of test fixture one, providing working power for test fixture one, control relays K1-1 / K1-2 / K1-3 and test fixture two. The load interface of test fixture one is connected to the signal terminal of test fixture two. Test fixture two has built-in resistor-capacitor parallel load unit and resistor-inductor series load unit. The control interface of test fixture two is connected to the contact terminals of control relay K2. The connection control of the load circuit is realized by the on and off of K2. Control Link: The test host establishes bidirectional control and data transmission connections (including synchronous timing control and signal parameter configuration) with the high-speed analog-to-digital converter and the test signal source, respectively. At the same time, the test host establishes control connections with control relays K1-1, K1-2, K1-3, K2, and K3 to realize the switching of the on / off state of each relay, the adjustment of the output parameters of the test signal, and the reception and processing of the acquired data.

[0030] In some embodiments, the contact feature point recognition and basic parameter extraction algorithm identifies key feature points of contact action by setting voltage change threshold and stable duration threshold. When the voltage change in the digital signal exceeds the voltage change threshold and the duration reaches the stable duration threshold, it is determined to be a valid feature point. The contact bounce time is calculated based on the timestamp difference of the valid feature points. The contact resistance is calculated by the ratio of the voltage amplitude of the test signal to the circuit current value. During load testing, feature points corresponding to the generation and extinction of the arc are additionally identified, and the arc duration is calculated.

[0031] In some embodiments, the calibration process of the dynamic error calibration algorithm includes zeroing calibration and drift compensation. Zeroing calibration is automatically performed during the contact disconnection phase of each test cycle, acquiring the zero signal measurement value of the test circuit and comparing it with the standard zero impedance parameter to obtain the system fixed error. Drift compensation calculates the signal drift based on the test duration and the change in ambient temperature, and obtains the calibrated parameters through the calculation logic of "basic parameters - system fixed error - signal drift".

[0032] In some embodiments, the number of consecutive preset test cycles in the parameter consistency verification and feedback optimization algorithm is 5-20, and the preset consistency threshold includes the resistance parameter threshold and the time parameter threshold. If the variance of the calibrated parameters is less than the corresponding threshold, the core parameters of the current algorithm are maintained. If the variance of the resistance parameter exceeds the threshold, the recognition weight of the contact feature point recognition and basic parameter extraction algorithm is reduced, while the calibration coefficient of the dynamic error calibration algorithm is increased. If the variance of the time parameter exceeds the threshold, the stable duration threshold is shortened, and the basic parameter extraction and calibration steps are re-executed.

[0033] In some embodiments, the verification results of the parameter consistency verification and feedback optimization algorithm are fed back to the first two types of algorithms in the form of quantified coefficients. During feedback optimization, the rule of "consistency qualified coefficient → maintain algorithm parameters, consistency unqualified coefficient → adjust algorithm parameters according to preset ratio" is followed. The adjustment range of the identification threshold is 5%-15% of the initial value, and the adjustment range of the calibration coefficient is 10%-25% of the initial value, to ensure the stability and effectiveness of the algorithm parameter adjustment.

[0034] In some embodiments, the contact action time in the basic parameters is the time difference between "the moment when the test host issues the pick-up / drop command" and "the moment when the effective closed / open feature point is first identified", the contact bounce time is the time difference between "the moment when the first effective closed feature point is identified" and "the moment when the last effective bounce feature point is identified", and the contact resistance is the ratio of the average voltage value to the average current value of the test signal during the stable contact closure phase, ensuring the accuracy of the basic parameter calculation.

[0035] In some embodiments, the verification result of the parameter consistency verification and feedback optimization algorithm is linked with the anomaly judgment result of S10. If the consistency verification result is unqualified and the anomaly judgment result is "abnormal" for three consecutive cycles, it is determined that the relay under test has a potential fault. The test host marks the sample and increases the adjustment range of the calibration coefficient in subsequent cycles to improve the pertinence of parameter calibration.

[0036] In some embodiments, the multi-sample consistency statistical standard is calculated based on the calibrated parameters of all qualified samples, including the parameter mean, standard deviation and coefficient of variation. Before calculation, sample data with unqualified consistency verification results are removed. If the deviation of the calibrated parameter of a single sample from the mean exceeds 3 times the standard deviation, it is determined to be an abnormal sample and is not included in the statistical scope, thus ensuring the accuracy of the multi-sample consistency report.

[0037] The technical concept of this invention is as follows: This invention discloses a method for analyzing and testing the performance of relay contacts. The core of this method is to achieve fully automatic and high-precision testing of multiple parameters of relay contacts through hardware configuration and algorithm collaboration.

[0038] The testing system is centered around a test host and includes hardware components such as test fixtures, control relays (K1, K2, K3), test signal sources, high-speed analog-to-digital converters, and selector switches. The test host loads preset parameter templates to automatically configure the test mode, power supply type, and load type. Modular test fixtures enable rapid adaptation of the relays under test and flexible replacement of the load.

[0039] During the test, the test signal source injects test signals into the contacts of the relay under test through a selection switch. The high-speed analog-to-digital converter synchronously acquires the contact status signals and converts them into digital signals. The test host calls three collaborative algorithms (contact feature point recognition and basic parameter extraction algorithm, dynamic error calibration algorithm, and parameter consistency verification and feedback optimization algorithm) to process the digital signals: first, it extracts the key feature points of the contact action and calculates the basic parameters; then, it eliminates system errors and drift through dynamic calibration; and finally, it judges the parameter stability through consistency verification and optimizes the parameters of the preceding algorithms in reverse, forming a closed-loop processing.

[0040] Subsequently, based on the calibrated and accurate parameters, electrical life cycle testing is performed, parameter anomalies are monitored in real time, performance judgment of individual samples and consistency analysis of multiple samples are completed, and finally, a pass report or a fault report is output, achieving comprehensive, accurate, and efficient testing of relay contact performance. The relay contact testing process is as follows: Figure 3 As shown.

[0041] S1: Load the preset test parameter template on the test host. After the test host starts, it automatically initializes the system hardware (including test signal source, high-speed analog-to-digital converter, selector switch, etc.). After the hardware self-test passes (no fault alarms, normal communication of all components), it loads the preset test parameter template. This template is based on the application scenarios and reliability requirements of relays in the rail transit field. It contains complete and comprehensive test configuration information, specifically covering test mode options (control relay K1 open, control relay K1 closed), load type options (inductive load, capacitive load), power supply type options (AC power supply, DC power supply), failure characteristic definitions (deviation of the test contact bounce time from the preset time, test contact resistance exceeding the preset value), preset contact performance values ​​(such as action time ≤ 5ms, contact resistance ≤ 100mΩ, etc.), statistical standards for multi-sample consistency analysis (such as parameter distribution interval coefficient of variation ≤ 5%), interval time of the tested relay's cyclic action (adjustable range 10ms-10s), and number of electrical life test cycles (adjustable range 100-1 million cycles). All parameters can be modified according to specific test requirements, but the default parameters of the template already cover the test requirements of mainstream rail transit relays.

[0042] S2: Prepare test fixture one and test fixture two For the relay model under test, first obtain its pinout diagram (e.g., number of pins 2-16, pin spacing 2.54mm or 3.96mm, etc.). Adjust or replace the pin interface module of test fixture one according to the diagram. The interface module of test fixture one adopts a modular design. By changing the contact terminals with different pin spacings and adjusting the terminal arrangement order, precise adaptation with the relay under test can be achieved. After adaptation, test the relay under test into test fixture one and check the interface fit (no looseness, no misalignment), ensuring that the insertion and extraction force is within the range of 5-10N to meet the requirements of repeated testing. At the same time, according to the load type and load capacity determined in the preset test parameter template (inductive load range 1mH-10H, capacitive load range 1nF-100μF), select the corresponding load module and install it into test fixture two. Test fixture two has a built-in load detection circuit. After installation, it automatically detects the integrity and parameter accuracy of the load module. If the load capacity deviation exceeds ±2%, an alarm will be issued, and the load module needs to be replaced or calibrated to ensure that the test load meets the preset requirements.

[0043] S3: Set the state of control relay K1 The test host sends a status control command to the control relay K1 according to the test mode selected in the preset test parameter template. The control relay K1 is an electromagnetic relay with a response time ≤1ms. Its status has only two selectable states, which are fully enumerated: The first is to set the control relay K1 to the open state. In this state, the external test power supply is not connected to the contacts of the relay under test; only a low-power test voltage is provided through the test signal source to measure basic contact parameters without load interference. The second is to set the control relay K1 to the closed state. In this state, the external test power supply and the test signal source work together to apply a load-bearing test voltage to the contacts of the relay under test, simulating contact performance parameter measurement under actual working conditions. After the status is set, the test host confirms the actual status of K1 through a feedback signal to ensure consistency with the command. If the status is inconsistent, the command is resent and wiring faults are investigated.

[0044] S4: Set the state of control relay K2 The test host sends a load switching command to control relay K2 based on the load type option in the preset test parameter template. Control relay K2 is a double-throw relay, its core function being to achieve rapid switching of the test load type. Two switching states are fully enumerated and without omission: The first is switching the contacts of control relay K2 to the inductive load side, connecting the test system to the inductive load module installed in test fixture two. This load module simulates the electrical characteristics of actual inductive loads such as motors and transformers through an inductor coil, and its inductance value can be finely adjusted using the adjustment knob on the fixture. The second is switching the contacts of control relay K2 to the capacitive load side, connecting the test system to the capacitive load module installed in test fixture two. This load module simulates the electrical characteristics of actual capacitive loads such as capacitors and rectifier circuits through a capacitor array, and its capacitance value can be adjusted by replacing different capacitor units. After the load switching is completed, the test host verifies whether the load type is correctly connected by collecting the impedance value of the load circuit. The impedance value of the inductive load increases with increasing frequency, while the impedance value of the capacitive load decreases with increasing frequency, thus determining the effectiveness of the switching.

[0045] S5: Set the state of control relay K3 The test host sends a power switching command to the control relay K3 based on the power type option in the preset test parameter template. Control relay K3 is a high-power relay supporting switching between various test power supplies, including AC 220V and DC 24V. Two switching states are fully enumerated: the first is switching the contacts of control relay K3 to the AC power supply side, with the connected AC power voltage range set to 110V-220V and frequency 50Hz / 60Hz, meeting the testing requirements for AC relays in the rail transit field; the second is switching the contacts of control relay K3 to the DC power supply side, with the connected DC power voltage range set to 12V-110V, supporting constant voltage output mode and a ripple factor ≤0.5%, meeting the testing requirements for DC relays. After power switching, the test host collects the output parameters of the test power supply through a voltage sensor, confirming that the voltage value, frequency (AC), and other parameters are consistent with the preset values. If the deviation exceeds ±1%, a power calibration program is initiated to ensure the stability and accuracy of the test power supply.

[0046] S6: Install the relay under test The tester holds the relay under test and aligns it with the mating interface of Test Fixture 1, then inserts it smoothly in the vertical direction. During the insertion process, avoid tilting or excessive force that may cause pin bending. After the insertion is complete, the self-locking mechanism of Test Fixture 1 automatically locks the relay under test to prevent loosening due to vibration during the test. Subsequently, the test host starts the initial connection detection. By sending a weak detection current (1 mA) to the relay under test, the initial value of the contact resistance of the circuit is collected. If the initial value ≤ 1 Ω, it is determined that the installation is qualified; if the initial value > 1 Ω, an alarm prompt is issued. The tester needs to pull out the relay under test to check whether the pins are oxidized and whether there are foreign objects at the interface. After cleaning, reinstall it until the detection is qualified. After the installation is qualified, the test system records information such as the model and serial number of the relay under test, and stores it in association with the subsequent test data for easy traceability.

[0047] S7: Inject test signals from the test signal source The test host sends a signal generation instruction to the test signal source according to the signal configuration in the preset test parameter template (such as the test voltage amplitude is 5V - 24V, and the signal type is DC constant voltage or AC sine wave). After receiving the instruction, the test signal source generates a test voltage signal that meets the requirements through an internal oscillator. The amplitude error of the signal ≤ ±0.1% and the frequency error ≤ ±0.01% (for AC signals). Subsequently, the test host controls the operation of Selector Switch 1. Selector Switch 1 is a multi-channel analog switch that supports the switching of 4 - 16 contact signals. According to the number of contacts of the relay under test (such as 2 contacts, 4 contacts, etc.), the test voltage signal is automatically and accurately injected into the input end of each target contact. During the injection process, the test signal source monitors the output current in real time. If the current exceeds the preset range (such as ≤ 100 mA), it is determined that the contact is short-circuited, and the signal injection is immediately stopped and an alarm is issued to avoid damaging the equipment; if the current is normal, the test signal is continuously injected to provide a basis for subsequent contact signal acquisition.

[0048] S8: The high-speed analog-to-digital converter collects contact signals The test host synchronously controls the operation of selection switch two, which is also a multi-channel analog switch, corresponding one-to-one with the contact channels of selection switch one. It automatically switches to the output terminal of the contact currently receiving the test signal, acquiring the contact status signal. The high-speed analog-to-digital converter (ADC) has a sampling rate of 1MHz (i.e., one data point is acquired per microsecond) and a resolution of 16 bits, enabling precise capture of minute changes in the voltage signal during contact operation. The acquired signal is an analog voltage signal, ranging from 0V to 5V. The high-speed ADC converts it into a digital signal (16-bit binary data) and transmits it in real-time to the test host's buffer via a high-speed data bus (such as an SPI bus with a transmission rate of 10Mbps). CRC checks are used during transmission to ensure data integrity; if the check fails, the signal for the corresponding time period is reacquired to avoid data loss or errors. The test host's buffer capacity is ≥1GB, capable of storing 10 minutes of continuous acquired data, providing sufficient raw data support for subsequent algorithm processing.

[0049] S9: The test host integrates three major algorithms to process digital signals, calculate and optimize contact performance parameters. After receiving the digital signal transmitted by the high-speed analog-to-digital converter, the test host immediately activates Algorithm 1 (contact feature point identification and basic parameter extraction), Algorithm 2 (dynamic error calibration), and Algorithm 3 (parameter consistency verification and feedback optimization). These three algorithms work collaboratively in a closed loop. The basic parameters output by Algorithm 1 serve as the input to Algorithm 2, and the calibration results of Algorithm 2 serve as the verification object for Algorithm 3. The verification results of Algorithm 3 then optimize the identification threshold of Algorithm 1 and the calibration coefficients of Algorithm 2, thereby improving the accuracy and stability of parameter calculation. The algorithm flow is as follows: Figure 4 As shown.

[0050] (1) Algorithm 1: Touch feature point recognition and basic parameter extraction algorithm Model Construction Process: The core objective of this algorithm is to accurately identify key feature points of contact action from massive amounts of digital signals, and then calculate basic performance parameters. The model is built based on the electrical characteristics of contact action: when the contact closes, the voltage signal abruptly changes from a high level (close to the test voltage) to a low level (close to 0V); when the contact opens, the voltage signal abruptly changes from a low level to a high level; when the contact bounces, the voltage signal fluctuates multiple times between high and low levels; when an electric arc is generated, the voltage signal shows a gradual upward or downward trend. Based on these characteristics, the model sets a voltage abrupt change threshold. ( (For test voltage amplitude) and stability duration threshold When the voltage change exceeds And the duration reached When the signal changes, it is determined to be a valid feature point; for the gradual signal in the arc stage, a linear fitting method is used to capture the start and end points of the voltage change to ensure that no feature points are missed.

[0051] Model training process: 1000 sets of relay test data of different types (AC / DC, inductive / capacitive loads) and different aging levels (new, used 10,000 times, used 100,000 times) were collected. Each set of data contained complete signals from 100 contact actions. Key feature points in the data (such as the control relay closing time, reliable contact closing time, etc.) were manually labeled. The labeled data was divided into a training set (800 sets) and a validation set (200 sets), and input into the model for training. During the training process, parameters such as voltage change threshold and stable duration threshold were continuously adjusted. The goal was to achieve a feature point recognition accuracy of ≥99.5% and a recognition delay of ≤1μs in the validation set. The final trained model should be able to adapt to the contact signal characteristics under different operating conditions.

[0052] Model application process: The test host calls the trained model to analyze the digital signal in the buffer point by point, identify key feature points and calculate basic parameters.

[0053] First, identify the closing time of the control relay. This moment marks the starting point at which the voltage at the contact input terminal begins to change after the test host sends a control relay closing command; subsequently, the moment of the first contact closure is identified. That is, the voltage at the contact output terminal suddenly drops to a low level for the first time and the duration reaches a certain threshold. The moment when the contacts finally stabilize and close. The moment when the voltage signal stabilizes at a low level and remains unchanged for 100 consecutive sampling points; the moment when the contact begins to loosen. The moment when the voltage at the contact output terminal begins to gradually rise from a low level; the moment when the voltage between the contacts equals the applied test voltage. For the voltage signal to rise to And a stable moment.

[0054] Simultaneously, the test host collects the loop current after the contacts are closed. The current is measured by a current sensor connected in series in the circuit, with an accuracy of ±0.1mA.

[0055] Based on the identified feature points and measured current, the basic parameters are calculated: When control relay K1 in S3 is in the off state: Contact action time (1) definition: The total time (in seconds) from the start of contact operation to the final stable closure after the control relay is closed. To control the relay closing time (unit: s); The time when the contact finally closes stably (unit: s).

[0056] Contact bounce time (2) definition: The time (in seconds) after the contact point closes for the first time and bounces multiple times until it finally closes stably. The moment when the contact first closes (unit: s).

[0057] Contact resistance (3) definition: Contact resistance after the contacts are stably closed (unit: Ω); Test voltage amplitude (unit: V); The current in the circuit after the contacts are closed (unit: A).

[0058] When control relay K1 in S3 is in the closed state: Contact arc time (4) definition: The time from when the contact begins to loosen until the voltage between the contacts reaches the test voltage (unit: s); The moment when the contact begins to loosen (unit: seconds); The moment when the voltage between the contacts equals the applied test voltage (unit: s).

[0059] Contact bounce time (5) Contact resistance under load current (6) definition: Contact resistance of the contact point under load current (unit: Ω); The current in the load circuit (unit: A) is collected in real time by a current sensor.

[0060] Core contribution: As the foundation of the entire parameter calculation process, this algorithm accurately extracts the key time nodes and initial resistance values ​​of the contact action from the original digital signal, solving the problems of complex signals and difficult identification of feature points during the contact action process. It provides a high-quality data source for subsequent error calibration and consistency verification, and its recognition accuracy directly determines the basic accuracy of the final test results.

[0061] (2) Algorithm 2: Dynamic error calibration algorithm Model Construction Process: The acquisition circuits of the test system (such as high-speed analog-to-digital converters and signal transmission lines) have inherent system errors. Simultaneously, environmental temperature changes (0℃-50℃) and power fluctuations during testing can cause signal drift, affecting the accuracy of parameter measurements. Therefore, the model constructs a two-stage calibration logic of "zeroing calibration + drift compensation": zeroing calibration uses the zero signal when the contacts are completely disconnected as a reference to eliminate inherent system errors; drift compensation establishes a dynamic compensation model based on the linear relationship between test duration and drift amount, correcting the errors caused by signal drift in real time. A calibration coefficient is introduced into the model. Used to adjust the intensity of drift compensation. The initial values ​​were obtained by fitting a large amount of experimental data. The algorithm can be dynamically adjusted based on the feedback results from Algorithm 3.

[0062] Model training process: Using a standard resistor with an accuracy of ±0.001Ω and an error-free signal generator, test signals were simulated under different test conditions (ambient temperatures of 0℃, 25℃, and 50℃; test durations of 1 min, 10 min, 60 min, and 24 h). Raw and true values ​​were collected for each condition, and the system error was calculated. and drift amount Using operating parameters (temperature, test duration) as input and error values ​​as output, the drift compensation function of the trained model is... ( (For the test duration), the training objective is to ensure that the calibrated measurement error is ≤ ±0.01Ω (resistance parameter) and ≤ ±0.1μs (time parameter), and the final trained model can adapt to the error compensation requirements under different working conditions.

[0063] Model application process: First, a zeroing calibration is performed: when the contacts of the relay under test are completely open, the test host controls the test signal source to output a zero voltage signal and continuously acquires the measured values ​​within 10ms. (Measurement results of Algorithm 1), with standard zero resistance ( Using as a benchmark, calculate the systematic error. This process is executed automatically during the contact disconnection phase of each test cycle, ensuring that system errors are updated in real time.

[0064] Then calculate the drift compensation amount: based on the current test duration. (Cumulative time from the start of the test to the current moment) and calibration coefficient Through formula Calculate the drift compensation amount, where The unit is s. It is a dimensionless coefficient.

[0065] Finally, the output parameters of Algorithm 1 are calibrated: Contact resistance after calibration (7) Contact action time after calibration (8) Contact bounce time after calibration (9) Contact arc time after calibration (10) definition: The calibrated contact resistance (unit: Ω); , , These are the calibrated motion time, bounce time, and arc time (unit: s); The systematic error of the time parameter is calibrated using a reference timer. (i.e., 100ns).

[0066] Interaction logic with Algorithm 1: Basic parameters output by Algorithm 1 , , , The data is directly used as input data for Algorithm 2, which calibrates it using formulas (7)-(10) to eliminate systematic errors and drift effects. The calibrated parameters... , , , On the one hand, it serves as verification data for Algorithm 3; on the other hand, it incorporates system errors. , The feedback is sent to Algorithm 1, which then adjusts the voltage change threshold for feature point recognition based on this feedback. (like When it is large, decrease To improve recognition sensitivity, and further enhance the accuracy of basic parameter extraction.

[0067] Core contribution: This algorithm specifically addresses the inherent errors and dynamic drift issues of the test system. Through real-time zeroing calibration and dynamic drift compensation, it improves the accuracy of parameter measurement by an order of magnitude, providing a reliable data foundation for subsequent consistency verification and ensuring that test results are not affected by changes in the environment and equipment status.

[0068] (3) Algorithm 3: Parameter Consistency Verification and Feedback Optimization Algorithm Model building process: Parameters in a single test cycle may be affected by random disturbances, resulting in accidental errors. Therefore, a variance analysis model based on a sliding window is constructed. The stability and consistency of the parameters are determined by analyzing calibrated parameters from multiple consecutive tests. The size of the sliding window... The number of test loops is such that it ensures real-time performance (verification every 10 loops) while effectively suppressing random interference (the statistical results from 10 loops are representative). A consistency threshold is set in the model. (Resistance parameter threshold) and (Time parameter threshold) , When the variance of the parameters within the window is less than or equal to the threshold, the model is considered to have passed the consistency test; otherwise, it is considered to have failed. Simultaneously, the model outputs the validation results. ( To be qualified, (For non-compliance), and based on Dynamically adjust the recognition weights of Algorithm 1 calibration coefficients of Algorithm 2 This forms a closed-loop optimization.

[0069] Model training process: Collect 100 consecutive test data sets of 50 sets of relays under the same test conditions. Each set of data includes calibrated resistance and time parameters. For each set of data, apply a sliding window... Calculate the variance, statistically analyze the variance distribution range of qualified relays, and determine the consistency threshold. and To ensure a pass rate of ≥99% for qualified relays, test data from 20 sets of unqualified relays (contact oxidation, loosening) were collected to train the mapping relationship between the verification results and algorithm parameter adjustments. At that time, by reducing and increase To enhance the algorithm's ability to suppress anomalous data, the training objective is to achieve an optimized verification accuracy of ≥98%.

[0070] Model application process: First, determine the sliding window data: select the most recently completed one. The calibrated parameters for each test cycle, including (No. Contact resistance after calibration in the next cycle. ), (No. (Calibration time after each cycle) (No. (bounce time after calibration in the next cycle) (No. (Calibrated arc time after each cycle).

[0071] Then calculate the average value of the parameters within the window: (11) (12) (13) (14) definition: This represents the average contact resistance within the window (unit: Ω). , , These are the average values ​​(in seconds) of the window action time, bounce time, and arc time.

[0072] Next, the variance of the parameters within the window is calculated. Variance reflects the degree of dispersion of the parameters; the smaller the variance, the better the consistency. (15) (16) (17) (18) definition: The variance of contact resistance (unit: ); , , The variances of action time, bounce time, and arc time, respectively (unit: ).

[0073] Finally, a consistency check is performed and feedback is provided for optimization. like and and and Then determine (Parameter consistency is satisfactory), now set the recognition weights for Algorithm 1. (Maintaining maximum recognition sensitivity), calibration coefficients for Algorithm 2 (Basic compensation strength); If any parameter variance exceeds the corresponding threshold, then it is determined that... (Parameter consistency fails), at this point, adjust the recognition weights of Algorithm 1. (Reducing the weight of outlier data), calibration coefficients of Algorithm 2 (Enhance drift compensation strength), and at the same time trigger Algorithm 1 to re-extract the basic parameters in the current window and Algorithm 2 to recalibrate until the parameters are consistent and qualified.

[0074] Interaction logic with Algorithm 1 and Algorithm 2: In the interaction with Algorithm 1, the results are verified. Through the weight adjustment formula Feedback is given to Algorithm 1 when hour, To reduce this, Algorithm 1 filters out signals with excessive fluctuations (such as voltage spikes exceeding a certain threshold) during feature point recognition. (Double the signal strength), reducing the impact of abnormal data on the basic parameters; the optimized basic parameters from Algorithm 1 are re-inputted into Algorithm 2, making the calibration data more accurate, thereby improving the accuracy of the next verification by Algorithm 3. During the interaction with Algorithm 2, Adjustment formula using calibration coefficient Feedback is given to Algorithm 2 when hour, Increase drift compensation amount The addition and enhancement of signal drift suppression improves the stability of the calibrated parameters, providing more reliable verification data for Algorithm 3.

[0075] Core contributions: This algorithm effectively distinguishes between random fluctuations and real anomalies in parameters through sliding window variance analysis, ensuring the stability and consistency of output parameters. At the same time, through closed-loop feedback optimization, it dynamically adjusts the core parameters of the first two algorithms, enabling the entire testing system to adapt to different test conditions and relay states, continuously improving test accuracy and reliability, and solving the problem that a single algorithm cannot cope with complex test environments.

[0076] Final output After the fusion processing and closed-loop optimization of the three major algorithms, the test host outputs accurate and stable contact performance parameters: when the control relay K1 is in the open state, the output calibrated and consistent contact action time is qualified. Contact bounce time Contact resistance When control relay K1 is in the closed state, the output is the calibrated and consistent contact arc time. Contact bounce time Contact resistance under load current These parameters will serve as the core basis for subsequent electrical life testing and data statistics.

[0077] S10: Perform electrical life test on the relay under test. The test host maintains the states of control relays K1, K2, and K3 set in S3, S4, and S5 unchanged. According to the interval time (e.g., 1 second) and electrical life test count (e.g., 100,000 cycles) defined in the preset test parameter template, it sends cyclic action commands to the relay under test, controlling it to repeatedly perform pick-up and drop-down actions. The process of each cycle is as follows: pick-up command → contact closure → stabilization for 50ms → drop-down command → contact opening → stabilization for 50ms, ensuring sufficient time for the contacts to complete the action and for the signal to stabilize. After each cycle is completed, the test host obtains the current contact performance parameters (final output in S9) in real time. , , , ), and compare it one by one with the failure feature definitions in the preset test parameter template to determine the contact status: the first case is the current contact bounce time If the time deviates from the preset time (e.g., 5ms) by more than ±20%, the contact state is considered abnormal, and the number of cycles and the abnormal parameter value are recorded. The second scenario involves the current contact resistance. If the value exceeds the predefined value (e.g., 100mΩ), the contact state is determined to be abnormal, and the relevant information is recorded. This test process continues until the preset number of electrical life tests is completed. If an abnormality occurs in 3 consecutive cycles during the test, the relay under test is determined to have failed prematurely, the test is stopped, and the number of failure cycles is recorded.

[0078] S11: Statistical Analysis and Anomaly Detection of Test Data for Individual Samples The test host retrieves all contact performance parameters output by S9, parameters for each cycle recorded by S10, and anomaly judgment results from the database, and starts the data statistical analysis program. The statistical analysis includes: the maximum, minimum, average, and standard deviation of each performance parameter; the number of occurrences, frequency, and first cycle number of abnormal parameters; and the trend of parameter changes with the number of cycles (e.g., whether the contact resistance increases with the number of cycles). Subsequently, the test host comprehensively compares the statistical analysis results with the predetermined values ​​of contact performance in the preset test parameter template: In the first case, if the statistical values ​​of all contact performance parameters do not deviate from the predetermined values ​​(e.g., average value ≤ predetermined value, maximum value ≤ predetermined value × 1.2), and the abnormality judgment result is no abnormality or the number of abnormalities ≤ 0.1% of the total number of cycles, then a contact performance qualification report for the relay under test is generated. The report includes the basic information of the relay under test (model, serial number), test configuration parameters (power supply type, load type, etc.), statistical results of each performance parameter, and electrical life test process record; In the second case, if the statistical value of any contact performance parameter deviates from the predetermined value, or the number of abnormalities > 0.1% of the total number of cycles, then a fault report is generated. In addition to the above basic information, the report also lists in detail the name of the abnormal parameter, statistical value, predetermined value, number of cycles in which the abnormality occurred, abnormality judgment basis (definition of failure characteristics), and possible fault cause analysis (e.g., contact oxidation, mechanical wear, etc.).

[0079] S12: Multi-sample consistency testing (if required by testing requirements) If the testing requirements involve consistency testing of multiple samples, the testers should follow the complete process from S6 to S11, sequentially testing multiple relay samples (recommended number ≥ 30, to meet statistical significance requirements) with the same parameters. The test configuration (power type, load type, number of cycles, etc.) for each sample should be completely identical to avoid differences in test conditions affecting the consistency analysis results. After testing, the test host collects the pass / fail reports for all samples and extracts the effective contact performance parameters for each sample (abnormal parameters from faulty samples are excluded and not included in the consistency analysis). Subsequently, the test host performs statistical analysis on the effective parameters and calculates the performance parameters of each contact (contact actuation time). Contact bounce time Contact arc time Contact resistance The distribution range (minimum - maximum), mean, standard deviation, and coefficient of variation (standard deviation / mean × 100%) of the samples are calculated. Finally, a multi-sample contact performance consistency report is generated, including: a list of basic information for the test samples, statistical results for each performance parameter (distribution range, mean, standard deviation, coefficient of variation), the percentage of qualified samples (number of qualified samples / total number of samples × 100%), a summary of anomalies in faulty samples, and a consistency evaluation (e.g., coefficient of variation ≤ 5% is considered good consistency, > 5% is considered poor consistency).

[0080] S13: Test End and Equipment Reset After testing a single or multiple samples, the test host stops sending action control commands to the relay under test and simultaneously disconnects control relay K3 to cut off the test power supply, preventing overheating caused by continuous power supply. The tester presses the unlock button on test fixture one, smoothly removes the relay under test, and places it in the designated sample box. Subsequently, depending on subsequent testing needs, if testing of the same type of relay continues, test fixture one and test fixture two are retained; if a different type of relay is replaced, test fixture one is replaced with one adapted to the new relay, or the load module of test fixture two is replaced according to the new load requirements. Finally, the test host clears the temporary data (cache data) of this test, stores the test reports (pass report, fault report, consistency report) in PDF format to the local hard drive, and backs them up to the cloud server, while keeping the content of the preset test parameter template unchanged. It then enters standby mode, awaiting the triggering of the next test command.

[0081] To facilitate parameter adjustment during testing to achieve optimal results, please refer to the following debugging guidelines for the three core algorithm parameters: I. Algorithm 1: Touch Point Feature Point Recognition and Basic Parameter Extraction Algorithm (1) List of core parameters Voltage mutation threshold The initial value is ( (Measured voltage amplitude), dimensionless, used to determine the effective voltage change signal for contact operation.

[0082] Stable duration threshold The initial value is The unit is It is used to filter transient interference signals and identify valid feature points.

[0083] (2) Debugging steps 1. Initial calibration: Set as , Set as Ten sets of standard relay samples (qualified products with known contact performance parameters) were selected, and a single test was performed to collect the feature point identification results.

[0084] 2. Feature point recognition verification: Compare the identified control relay closing times. The moment when the contact first closes The identification deviation is calculated by comparing the characteristic points with the calibration values ​​of the standard sample.

[0085] 3. Parameter Iterative Adjustment: If there is a recognition deviation, adjust the parameters according to the following rules: If a feature point is missed (e.g., voltage fluctuations during the contact bounce phase are not detected), Downgraded to ,Keep If the test remains unchanged, retest and verify; if it still misses, then... shortened to .

[0086] If misidentification occurs (such as mistaking an interference signal for a valid feature point), Upgraded to ,Keep Unchanged; if misidentification still occurs, Extended to .

[0087] 4. Stability verification: After adjustment, select 20 sets of different types of relays (covering AC / DC, inductive / capacitive load combinations) and test them 5 times consecutively to ensure that the feature point recognition accuracy is ≥99.5% and the recognition delay is ≤1μs, then the debugging is complete.

[0088] (3) Optimization Objective Feature point recognition accuracy: ≥99.5% (no missed detections, no false recognitions).

[0089] Recognition delay: ≤1μs (the time difference between the recognition time and the actual feature point time).

[0090] Basic parameter calculation error: contact action time bounce time Error ≤ ±0.1μs, contact resistance Error ≤ ±0.01Ω.

[0091] (4) Troubleshooting If problem 1 occurs: the voltage change is not obvious when the contacts are closed, leading to... Recognition lag.

[0092] Solution: Lower to At the same time, check the stability of the output voltage of the test signal source to ensure that the signal amplitude error is ≤ ±0.1%.

[0093] Problem 2: Electromagnetic interference exists in the testing environment, leading to frequent misidentification of feature points.

[0094] Solution: Increase to ,extend to At the same time, check the grounding of the test system and enhance electromagnetic shielding.

[0095] II. Algorithm 2: Dynamic Error Calibration Algorithm (1) List of core parameters Initial value of calibration coefficient The initial value is Dimensionless, used to adjust the intensity of drift compensation.

[0096] Time parameter systematic error : The calibration value is (100ns), unit is It is pre-calibrated by a reference timer and used for fixed error calibration of time parameters.

[0097] (2) Debugging steps 1. Standard load calibration: Connect a standard resistor with an accuracy of ±0.001Ω as the load, set the test power supply to DC 24V and the load type to pure resistive load (simulating stable working conditions), and perform a continuous test for 1 hour.

[0098] 2. Error Calculation: Record the contact resistance output by Algorithm 1 every 10 minutes. After calibration with Algorithm 2 The calibration error is calculated by comparing it with the true value of the standard resistor. .

[0099] 3. Calibration coefficient adjustment: If calibration error (Resistance parameters), Upgraded to Retest for 1 hour and observe the change in error; if the error still exceeds the limit, continue to adjust upwards. Each increment is 0.00005, until... .

[0100] If the parameters fluctuate too much after calibration (e.g.) If the fluctuation is greater than 0.005Ω for five consecutive records, then... Downgraded to To balance the compensation strength and parameter stability.

[0101] 4. Multi-condition verification: Change the test power supply type (AC 220V) and load type (inductive load 1mH, capacitive load 10μF), repeat the above test, and ensure that the calibration error meets the requirements under different operating conditions to complete the debugging.

[0102] (3) Optimization Objective Resistance parameter calibration error: ≤ ±0.01Ω (compared with the true value of the standard resistor).

[0103] Time parameter calibration error: ≤ ±0.1μs (compared with the reference timer calibration value).

[0104] Drift suppression effect: After 24 hours of continuous testing, the parameter drift was ≤0.02Ω (resistance) and ≤0.2μs (time).

[0105] (4) Troubleshooting Problem 1: After the test duration exceeds 6 hours, the resistance calibration error increases significantly.

[0106] Solution: Upgraded to Strengthen drift compensation; at the same time, check the heat dissipation of the load module of test fixture two to avoid the load parameters drifting due to temperature rise.

[0107] If problem 2 occurs: a fixed deviation still exists after time parameter calibration (e.g.) It is always 0.2 μs larger than the true value.

[0108] Solution: Recalibrate 100 sets of time signals were acquired using a reference timer, the average error was calculated, and the results were updated. The value is then recalibrated.

[0109] III. Algorithm 3: Parameter Consistency Verification and Feedback Optimization Algorithm (1) List of core parameters Slide window size The initial value is 10 test cycles, dimensionless, used to determine the number of samples for consistency verification.

[0110] Resistance consistency threshold The initial value is The unit is It is used to determine the degree of dispersion of contact resistance.

[0111] Time consistency threshold The initial value is The unit is It is used to determine the degree of dispersion of time parameters.

[0112] Identify weights Adjust the formula to ( (1 indicates pass, 0 indicates fail) is a dimensionless value used to optimize the feature point recognition weights of Algorithm 1 (for consistency verification).

[0113] (2) Debugging steps 1. Verification of qualified samples: Select 10 sets of qualified relay samples and perform 20 cycles of testing on each set under the same test configuration (DC 24V, inductive load 1mH), and collect the calibrated parameter data.

[0114] 2. Consistency Verification: Click the sliding window. Calculate the parameter variance for each sample group. , Statistical consistency is qualified (etc.) ) proportion.

[0115] 3. Parameter adjustment: If the consistency pass rate of qualified samples is <99% (there are false negatives), Upgraded to , Upgraded to Recalculate the variance and verify it; if the proportion still does not meet the standard, continue to fine-tune the threshold, with each increase not exceeding 10% of the initial value.

[0116] If a non-conforming sample (contact oxidation) is judged as qualified ( ),Will Downgraded to , Downgraded to This enhances the sensitivity of anomaly detection.

[0117] 4. Feedback Optimization Verification: After adjusting the threshold, triggering 3 instances of consistency failure ( In scenarios such as manually connecting oxidized contacts to a relay, observe the performance of Algorithm 1. And Algorithm 2 Automatic adjustment and The parameters were adjusted and recalibrated to be consistent and qualified, thus completing the debugging process.

[0118] (3) Optimization Objective Consistency determination accuracy: ≥98% (the proportion of qualified samples being qualified is ≥99%, and the proportion of unqualified samples being unqualified is ≥97%).

[0119] Feedback to optimize response: Verification results After the changes, the parameter adjustment delay for Algorithms 1 and 2 is ≤ 1 test loop.

[0120] Reasonableness of sliding window: The variance of parameters within the window can effectively distinguish between random fluctuations and real anomalies, without oversensitivity or sluggishness.

[0121] (4) Troubleshooting If problem 1 occurs: the variance of parameters of qualified samples frequently exceeds the threshold, leading to... Misjudgment.

[0122] Solution: Adjust the sliding window size The number of iterations was increased to 15 to improve the stability of variance calculation; at the same time, the threshold was fine-tuned. , This reduces the probability of misjudgment.

[0123] If problem 2 occurs: the parameters are still unstable after feedback optimization ( (Repeatedly switching between 0 and 1).

[0124] Solution: The adjustment range has been narrowed to , The adjustment range has been narrowed to Reduce the adjustment range to avoid parameter oscillation.

[0125] The embodiments described above are for illustrative purposes only and are not intended to limit the invention. Therefore, any changes in numerical values ​​or substitutions of equivalent elements should still fall within the scope of this invention.

[0126] The above detailed description will enable those skilled in the art to understand that the present invention can indeed achieve the aforementioned objectives and has complied with the provisions of the Patent Law.

Claims

1. A method for analyzing and testing the performance of relay contacts, characterized in that, include: S1: After the test host initializes and performs a self-test, it loads a preset test parameter template, including test mode, load type, power supply type, cycle interval, and number of electrical life test cycles. S2: Configure test fixture one and test fixture two, so that test fixture one is adapted to the relay under test and test fixture two is equipped with a corresponding load unit; S3: Configure the state of control relay K1 according to the test mode, and confirm that the state matches; S4: Configure the state of control relay K2 according to the load type, connect it to the load unit of test fixture two and verify its effectiveness; S5: Configure the state of control relay K3 according to the power supply type, and confirm the state matching; S6: Install the relay under test through test fixture one; S7: Control the test signal source to generate a test signal, inject it into the target contact of the relay under test, and construct a test circuit containing two load units of the test fixture; S8: Collects the contact response signals in the test circuit, converts them into digital signals, and stores them in the test host; S9: The test host processes the digital signal: first, it calculates the basic parameters through the contact feature point recognition and basic parameter extraction algorithm, and then calibrates the basic parameters through the dynamic error calibration algorithm to obtain the calibrated parameters; S10: Keep the control relay state unchanged, control the relay under test to operate according to the cycle interval and the number of electrical life test cycles, obtain the calibrated parameters after each cycle, compare them with the failure characteristics and record the abnormality judgment result.

2. The method for analyzing and testing the performance of relay contacts according to claim 1, characterized in that, The preset test parameter template also includes failure characteristics, predetermined performance values, and statistical standards for consistency among multiple samples; The analysis and testing methods for the relay contact performance also include: S11: Summarize the calibration parameters and anomaly judgment results, and generate a single relay performance evaluation report by combining the performance preset values; S12: When multiple samples need to be tested, test multiple samples according to the procedures from S6 to S11, and generate a multi-sample consistency report based on the consistency statistics standard. S13: After disconnecting the test power, removing the relay under test, clearing temporary data, and storing the report, the test system enters standby mode.

3. The method for analyzing and testing the performance of relay contacts according to claim 1, characterized in that, The test host, test fixture one, test fixture two, high-speed analog-to-digital converter, test signal source, selection switch two, selection switch one, control relay K1, control relay K2, control relay K3, AC power supply and DC power supply constitute the test system; the control relay K1 includes control relays K1-1, K1-2 and K1-3; Signal Injection and Synchronization Link: The signal output of the test signal source is divided into two paths. One path is connected to the signal input of selector switch one, and the other path is connected to the synchronization signal input of the high-speed analog-to-digital converter. The signal output of selector switch one is connected to the test signal injection interface of test fixture one. Test fixture one has a built-in mounting position for the relay under test. The contact circuit of the relay under test is connected in series with the contact terminals of control relays K1-1, K1-2, and K1-3 to realize the switching of the injection path of the test signal to the contacts of the relay under test. Signal acquisition link: The contact signal acquisition interface of test fixture one is connected to the signal input terminal of selector switch two, the signal output terminal of selector switch two is connected to the signal acquisition input terminal of high-speed analog-to-digital converter, the digital signal output terminal of high-speed analog-to-digital converter establishes a data transmission connection with the test host, and at the same time, the high-speed analog-to-digital converter and the test signal source maintain synchronous signal interaction, forming a synchronous acquisition and transmission path for the response signal of the relay under test contact. Power supply and load circuit: The output terminals of AC power supply and DC power supply are connected to the contact terminals of control relay K3. The output terminal of K3 is connected to the power input interface of test fixture one, providing working power for test fixture one, control relays K1-1 / K1-2 / K1-3 and test fixture two. The load interface of test fixture one is connected to the signal terminal of test fixture two. Test fixture two has built-in resistor-capacitor parallel load unit and resistor-inductor series load unit. The control interface of test fixture two is connected to the contact terminals of control relay K2. The connection control of the load circuit is realized by the on and off of K2. Control Link: The test host establishes bidirectional control and data transmission connections with the high-speed analog-to-digital converter and the test signal source, respectively. At the same time, the test host establishes control connections with control relays K1-1, K1-2, K1-3, K2, and K3 to realize the switching of the on / off state of each relay, the adjustment of the output parameters of the test signal, and the reception and processing of the acquired data.

4. The method for analyzing and testing the performance of relay contacts according to claim 1, characterized in that, The contact feature point recognition and basic parameter extraction algorithm identifies key feature points of contact action by setting voltage change threshold and stable duration threshold. When the voltage change in the digital signal exceeds the voltage change threshold and the duration reaches the stable duration threshold, it is determined to be a valid feature point. The contact bounce time is calculated based on the timestamp difference of the valid feature points. The contact resistance is calculated by the ratio of the voltage amplitude of the test signal to the circuit current value. During load testing, feature points corresponding to the generation and extinction of the arc are additionally identified, and the arc duration is calculated.

5. The method for analyzing and testing the performance of relay contacts according to claim 1, characterized in that, The calibration process of the dynamic error calibration algorithm includes zeroing calibration and drift compensation. Zeroing calibration is automatically performed during the contact disconnection phase of each test cycle, collecting the zero signal measurement value of the test circuit and comparing it with the standard zero impedance parameter to obtain the system fixed error. Drift compensation calculates the signal drift based on the test duration and the change in ambient temperature, and obtains the calibrated parameters through the calculation logic of basic parameters - system fixed error - signal drift.

6. The method for analyzing and testing the performance of relay contacts according to claim 1, characterized in that, In S9, the consistency of the calibrated parameters is verified by the parameter consistency verification and feedback optimization algorithm, and the core parameters of the first two types of algorithms are optimized in reverse according to the verification results until the output calibrated parameters that meet the accuracy requirements are output; the number of consecutive preset test cycles in the parameter consistency verification and feedback optimization algorithm is 5-20, and the preset consistency threshold includes the resistance parameter threshold and the time parameter threshold; If the variance of all calibrated parameters is less than the corresponding threshold, maintain the core parameters of the current algorithm; if the variance of the resistance parameter exceeds the threshold, reduce the recognition weight of the contact feature point recognition and basic parameter extraction algorithms, and increase the calibration coefficient of the dynamic error calibration algorithm; if the variance of the time parameter exceeds the threshold, shorten the stable duration threshold, and re-execute the basic parameter extraction and calibration steps.

7. The method for analyzing and testing the performance of relay contacts according to claim 6, characterized in that, The verification results of the parameter consistency verification and feedback optimization algorithm are fed back to the first two types of algorithms in the form of quantified coefficients. During feedback optimization, the algorithm parameters are maintained according to the consistency pass coefficient, and the algorithm parameters are adjusted according to the preset ratio according to the consistency failure coefficient. The adjustment range of the identification threshold is 5%-15% of the initial value, and the adjustment range of the calibration coefficient is 10%-25% of the initial value.

8. The method for analyzing and testing the performance of relay contacts according to claim 1, characterized in that, The contact action time in the basic parameters is the time difference between the moment the test host issues the pick-up / drop command and the moment the effective closed / open feature point is first identified; the contact bounce time is the time difference between the moment the first effective closed feature point is identified and the moment the last effective bounce feature point is identified; and the contact resistance is the ratio of the average voltage value to the average current value of the test signal during the stable contact closure phase.

9. The method for analyzing and testing the performance of relay contacts according to claim 6, characterized in that, The verification results of the parameter consistency verification and feedback optimization algorithm are linked with the anomaly judgment results of S10. If the consistency verification result is unqualified and the anomaly judgment results of three consecutive cycles are abnormal, it is determined that the relay under test has a potential fault. The test host marks the sample and increases the adjustment range of the calibration coefficient in subsequent cycles.

10. The method for analyzing and testing the performance of relay contacts according to claim 2, characterized in that, The multi-sample consistency statistical standard is calculated based on the calibrated parameters of all qualified samples, including the parameter mean, standard deviation and coefficient of variation. Before calculation, sample data with unqualified consistency verification results are removed. If the deviation of the calibrated parameter of a single sample from the mean exceeds 3 times the standard deviation, it is judged as an abnormal sample.

Images