Test system and method for simulating different working conditions of relay

Abstract

The invention relates to a test system and method for simulating different working conditions of a relay, and the system is characterized in that the system comprises an environment simulation region and a dynamic load simulation region, and the dynamic load simulation region comprises an electrical simulation region, a data collection region, and a data processing and control module; the environment simulation area is used for simulating working environment parameters of the relay, including temperature, humidity and mechanical stress; the environment simulation area is provided with a shell, and heating devices located on the top, the bottom and the side face of the shell are arranged in the shell and used for accurately adjusting the internal temperature. A ventilation opening is formed for adjusting air circulation; a temperature sensor and a humidity sensor are installed and used for monitoring environment temperature and humidity parameters in real time. According to the test system and the test method for simulating different working conditions of the relay, various environmental stresses can be synchronously applied, load changes can be dynamically simulated, data fusion analysis can be realized, and the accuracy of reliability verification of the relay can be improved.

Description

Technical Field

[0001] The present invention belongs to the technical field of railway relay testing, and in particular relates to a multi-environment simulation testing system and method for railway relays. Background Art

[0002] Traditional testing equipment can only simulate a single environmental factor (such as high temperature or vibration) and cannot replicate the multi-factor coupling effects found on railway sites (such as low temperature, humidity, and electromagnetic interference). However, in actual operating conditions, relays are often exposed to a multi-factor environment. A single simulation results in significant deviations from the actual operating conditions, making it impossible to fully assess the adaptability and reliability of relays in complex environments. Existing equipment load simulations are mostly fixed parameters, unable to dynamically simulate transient conditions such as grid voltage fluctuations and harmonic distortion. The separation of environmental parameters from electrical performance data makes it difficult to quantify the impact of the comprehensive environment on relay life. Due to the limitations of environmental and electrical simulations, as well as the lack of data processing and feedback control, existing relay testing technologies result in test results that deviate from actual application requirements, hindering the development and quality improvement of relay products. Summary of the Invention

[0003] The purpose of the present invention is to provide a multi-environment simulation test system and method for railway relays, so as to simultaneously apply multiple environmental stresses, dynamically simulate load changes, and realize data fusion analysis of the test device and method, thereby improving the accuracy of relay reliability verification.

[0004] The technical solution of the present invention is: to relate to a test system for simulating different working conditions of a relay, characterized by comprising an environmental simulation area, an electrical simulation area, a data acquisition area and a data processing and control module; The environmental simulation area is used to simulate the environmental parameters of the relay operation, including temperature, humidity, and mechanical stress. The environmental simulation area is provided with a housing, and heating devices are arranged on the top, bottom, and sides of the housing to accurately adjust the internal temperature. Ventilation holes are provided to adjust air circulation. Temperature sensors and humidity sensors are installed to monitor the environmental temperature and humidity parameters in real time. The area also includes rotating parts, one of which is used to connect to the mechanical structure to simulate mechanical stress. The electrical simulation area is used to simulate the electrical parameters of the relay when it is working, including load current and voltage; the electrical simulation area includes a load simulation circuit connected to the relay under test, and a basic power supply that provides stable power support for the electrical simulation area; The data acquisition area is used to collect environmental parameters and electrical parameters; the data acquisition area includes a data collector that communicates with the data processing and control module via an RS485 bus and realizes data transmission with the help of TCP / IP protocol, as well as auxiliary acquisition equipment; The data processing and control module includes a data synchronization unit, a Kalman filter unit, a feature extraction unit, a neural network life prediction unit and a feedback control unit; The data synchronization unit is used to receive data from the data collection area and synchronize the collected environmental parameters and electrical parameters to ensure the consistency of each parameter data in the time dimension, providing a basis for subsequent accurate analysis; The Kalman filter unit is connected to the data synchronization unit, receives the synchronized data, and uses the Kalman filter algorithm to filter the data to remove noise caused by external interference, sensor errors and other factors, thereby improving the quality and reliability of the data; The feature extraction unit is connected to the Kalman filter unit to extract effective feature parameters closely related to relay performance and life from the filtered data, such as but not limited to temperature change rate, current peak value, voltage fluctuation amplitude, and humidity change trend; The neural network life prediction unit is connected to the feature extraction unit, and uses the extracted effective feature parameters in combination with a neural network model pre-trained with a large amount of experimental data to predict the remaining service life of the relay and generate a detailed life prediction report, which includes the predicted life span and performance change trends at different stages; The feedback control unit is connected to the neural network life prediction unit, the environmental simulation area and the electrical simulation area respectively, and analyzes the working status and performance changes of the relay according to the life prediction report. If the prediction result shows that the life of the relay may be adversely affected, the feedback control unit will automatically adjust the temperature, humidity, and mechanical stress parameters of the environmental simulation area, as well as the load current and voltage parameters of the electrical simulation area, thereby forming a closed-loop control to put the relay in a condition that conforms to actual working conditions and is conducive to accurately testing its performance.

[0005] The environmental simulation area includes a temperature and humidity control cabin, a vibration simulation platform, and an electromagnetic interference generator. The temperature and humidity control cabin and the vibration simulation platform are fixed by mechanical brackets, and the electromagnetic interference generator is connected to the relay coil under test through a ring current probe. The electrical simulation area includes a programmable power supply module and a dynamic load box. The output end of the programmable power supply module is connected to the corresponding contacts of the relay coil under test. The dynamic load box is connected in parallel to the corresponding contacts of the relay coil under test. The data acquisition unit and data processing control unit include a multi-sensor array, an edge computing gateway, and a host computer; the multi-sensor array is connected to the edge computing gateway through an interface, and the edge computing gateway is electrically connected to the host computer through Ethernet; The data acquisition unit and the data processing control unit collect environmental parameters and electrical parameters such as temperature, vibration, electromagnetic interference in the environmental simulation area and voltage, current, and contact resistance in the electrical simulation area, and simultaneously issue relevant control instructions including temperature and humidity set value instructions and vibration frequency instructions; Temperature and humidity control chamber: Adopts a composite temperature control structure of semiconductor cooling chip and PTC heater to achieve rapid switching from -40℃ to 85℃ (heating rate ≥5℃ / min). It integrates ultrasonic humidifier and condensation dehumidification module, and the humidity control range is 5%~95% RH with an accuracy of ±2%; Vibration simulation platform: It uses a six-degree-of-freedom electromagnetic vibration platform with a frequency range of 0-200Hz and a maximum acceleration of 10g. It supports random vibration and sine sweep modes. It has a built-in relay mounting fixture and is suitable for relays of different sizes or models, such as JWXC-1700 and JYJXC-160 / 260. Electromagnetic interference generator: uses a programmable high-frequency pulse group generator (1MHz-100MHz, peak voltage ±5kV) to simulate signal cable coupling interference, equipped with a ring current probe to directly inject common mode / differential mode interference into the relay coil loop; The electrical simulation area, data acquisition area, and data processing and control module in the dynamic load simulation area complete the following work steps: a) Programmable power supply module: outputs in dual modes of 24V DC / 220V AC, supports voltage sag of 100% to 50% / 1ms, and THD of harmonic superposition is adjustable from 0-40%; b) Load box: multi-position resistance / inductance combination load 0.1Ω-10kΩ, 1mH-10H, simulates relay contact load capacity; The data acquisition and fusion unit includes: a) Multi-sensor array: high-precision temperature sensor: PT100, capacitive humidity sensor, vibration accelerometer, current probe with 1 GHz bandwidth; b) Edge computing gateway: Real-time synchronous collection of environmental parameters: temperature, vibration frequency and relay electrical parameters such as pull-in time and contact resistance, with timestamp alignment accuracy ≤1ms.

[0006] The vent is provided with a ventilation regulating device, which includes a fan or a damper for regulating air circulation.

[0007] The load simulation circuit can accurately simulate currents and voltages of different amplitudes, as well as electrical signals with complex waveforms according to actual needs.

[0008] The data collector area is connected to sensors in the environmental simulation area and the electrical simulation area, and the auxiliary collection equipment works in coordination with the data collector to ensure the comprehensiveness and accuracy of data collection.

[0009] A method for simulating a test system of different working environment conditions of a relay, characterized by comprising the following method steps: a multi-environment coupling control method, a dynamic load adaptive matching control method, and a data fusion and life prediction method; The steps of the multi-environment coupling control method are: a) Collaborative control algorithm: Based on fuzzy PID control, the algorithm dynamically adjusts the superposition timing of temperature, humidity, vibration frequency, and electromagnetic interference. For example, in a low-temperature environment, the algorithm superimposes 150Hz high-frequency vibration and the third harmonic 30% voltage harmonic to simulate the comprehensive operating conditions of a train passing through a switch in cold regions. b) Environmental-electrical parameter mapping model: Establish the degradation equation between relay contact time (T) and environmental stress (S): ; : Nominal pull-in time (reference value under stress-free conditions); : Temperature change (unit: °C), which represents the effect of thermal stress on material thermal expansion and coil resistance; RH: Relative humidity (unit: %), reflecting the effect of moisture on contact oxidation and accelerated aging of insulation performance; Avib: vibration acceleration (unit: g), quantifies the interference of mechanical vibration on the stability of armature motion; E: Electromagnetic interference intensity (unit: kV / m), which represents the disturbance of the alternating electric field or pulse magnetic field on the relay magnetic circuit and coil current; k1, k2, k3, k4: Accelerated aging coefficients, corresponding to the sensitivity to temperature, humidity, vibration, and electromagnetic interference acceleration factor, respectively, reflecting the sensitivity of changes in electromagnetic field strength to the pull-in time.

[0010] The steps of the dynamic load adaptive matching control method are: a) Intelligent load switching logic: Automatically matches the standard load (4.8kΩ / 0.44H) according to the relay model (such as JZXC-480) and randomly inserts transient overload (2 times the rated current) during the test; b) Harmonic injection strategy: The actual harmonic spectrum of the power grid is analyzed through FFT to generate a synthetic waveform containing the 3rd, 5th, and 7th harmonics with a THD accuracy of ±0.5%.

[0011] The steps of the data fusion and life prediction method are: 1. Parameter settings a) Multi-source data correlation analysis: Using the Kalman filter algorithm to fuse vibration acceleration data with contact jitter signals, we can identify poor contact caused by mechanical wear. b) Remaining life prediction model: Based on an LSTM neural network, it inputs environmental stress sequences (temperature, vibration, humidity) and electrical parameters (coil resistance change rate, contact voltage drop) and outputs the remaining life of the relay (MTBF error ≤ 10%). Input environmental stress sequence and electrical parameters. Environmental parameter input includes temperature, humidity, vibration spectrum, and electromagnetic interference intensity. Electrical parameter input includes pull-in time, release time, contact voltage drop, and coil current; 2. Data processing layer Data alignment module, used for timestamp synchronization, accuracy ≤ 1ms; Kalman filtering, which fuses vibration acceleration and contact jitter signals to output an estimated contact resistance value; Feature extraction, extracting degradation features such as temperature coefficient k1 and vibration sensitivity k2; 3. Lifespan prediction layer LSTM neural network, with an input sequence length of 24 hours and 128 hidden layer nodes. The inputs include environmental stress sequence and electrical feature vectors, and the outputs include remaining life (hours) and confidence interval (±10%). Report generation in PDF / HTML format, including degradation curves and maintenance recommendations; 4. Feedback control layer It is used to dynamically adjust test conditions, such as "increasing the vibration frequency to the failure threshold". After the adjustment is completed, it points to the environmental simulation area and the electrical simulation area to form a closed-loop control.

[0012] The parameter setting includes setting the temperature, humidity, and mechanical stress parameters of the environmental simulation area, and the load current and voltage parameters of the electrical simulation area according to the actual working environment and electrical requirements of the relay; and starting the data acquisition system. Test operation: Place the relay under test on the test platform in the environmental simulation area and connect it to the load simulation circuit in the electrical simulation area. Start the equipment in the environmental simulation area and the electrical simulation area to simulate the working environment and electrical conditions of the relay. The data acquisition system collects environmental and electrical parameters in real time and transmits the data to the data processing and control module. Data processing and life prediction: The data synchronization unit synchronizes the collected data; the Kalman filter unit filters the synchronized data; the feature extraction unit extracts effective feature parameters from the filtered data; the neural network life prediction unit uses the extracted feature parameters in combination with a pre-trained neural network model to predict the life of the relay and generate a life prediction report; Feedback control: The feedback control unit analyzes the working status and performance changes of the relay based on the life prediction report, adjusts the parameters of the environmental simulation area and the electrical simulation area to put the relay under more appropriate test conditions; and repeats the test process until the predetermined test time is reached or the test requirements are met.

[0013] Compared with the prior art, the present invention has the following advantages: including but not limited to the construction of a reference model, including the physical construction between multiple test hardware and the test implementation process, etc.

[0014] 1. Improved testing efficiency: Composite environment testing can replace traditional single-factor sequential testing, shortening the testing cycle.

[0015] 2. Improved fault coverage: It can stimulate hidden faults that are difficult to expose in a single environment, such as contact sticking caused by low temperature and vibration.

[0016] 3. Reduced maintenance costs: The life prediction model supports preventive maintenance and reduces operational interruptions caused by sudden relay failures.

[0017] The present invention will be further described below with reference to the accompanying drawings of the embodiments.

[0018] Figure 1 This is the overall structural diagram of the test bench; Figure 2 This is a schematic diagram of the internal structure of the temperature and humidity control cabin; Figure 3 It is a flow chart of data fusion and life prediction algorithm.

[0019] In the figure: 1-1, temperature and humidity control cabin, 1-2, vibration simulation table; 1-3, electromagnetic interference generator; 2-1, programmable power supply module; 2-2, dynamic load box; 3-1, multi-sensor array; 3-2, edge computing gateway; 3-3, host computer; 4-1, double-layer stainless steel insulation layer; 4-2, observation window; 5-1, semiconductor refrigeration chip array; 5-2, PTC heater; 5-3, temperature sensor; 6-1, ultrasonic humidifier; 6-2, condensation dehumidification module; 6-3, humidity sensor; 7-1, axial fan; 7-2 guide plate; 8-1, relay under test; 8-2, relay coil under test. DETAILED DESCRIPTION

[0020] In order to further illustrate the technical means and effects adopted by the present invention to achieve the predetermined purpose, the specific implementation methods, structural features and effects of the present invention are described in detail below with reference to the accompanying drawings and examples.

[0021] like Figure 1 and Figure 2As shown, the present invention relates to a test system and method for simulating different working conditions of a relay, which is characterized in that: it includes an environmental simulation area, an electrical simulation area, a data acquisition unit and a data processing control unit; the environmental simulation area includes a temperature and humidity control cabin 1-1, a vibration simulation platform 1-2, and an electromagnetic interference generator 1-3, the temperature and humidity control cabin 1-1 and the vibration simulation platform 1-2 are fixed by a mechanical bracket, and the electromagnetic interference generator 1-3 is connected to the relay coil 8-2 under test through a ring current probe; The electrical simulation area includes a programmable power supply module 2-1 and a dynamic load box 2-2. The output end of the programmable power supply module 2-1 is connected to the contact of the relay 8-1 under test, and the dynamic load box 2-2 is connected in parallel to the contact of the relay 8-1 under test. The data acquisition unit and the data processing control unit include a multi-sensor array 3-1, an edge computing gateway 3-2, and a host computer 3-3; the multi-sensor array 3-1 is connected to the edge computing gateway through an interface, and the edge computing gateway is electrically connected to the host computer through Ethernet.

[0022] The data acquisition unit and the data processing control unit collect the environmental parameters and electrical parameters of the temperature, vibration, electromagnetic interference of the environmental simulation area and the voltage, current, and contact resistance of the electrical simulation area, and at the same time issue relevant control instructions including temperature and humidity set value instructions and vibration frequency instructions.

[0023] Temperature and Humidity Control Chamber 1-1: Utilizes a composite temperature control structure featuring semiconductor cooling elements and PTC heaters, enabling rapid temperature switching from -40°C to 85°C (heating rate ≥ 5°C / min). Integrated with an ultrasonic humidifier and condensation dehumidification module, the humidity control range is 5% to 95% RH, with an accuracy of ±2%.

[0024] Vibration simulation table 1-2: uses a six-degree-of-freedom electromagnetic vibration table with a frequency range of 0-200Hz and a maximum acceleration of 10g. It supports random vibration and sine sweep modes, and has a built-in relay mounting fixture to adapt to electrical appliances of different sizes and models, such as JWXC-1700 and JYJXC-160 / 260.

[0025] Electromagnetic interference generator 1-3: uses a programmable high-frequency pulse group generator (1MHz-100MHz, peak voltage ±5kV) to simulate signal cable coupling interference, and is equipped with a circular current probe to directly inject common-mode / differential-mode interference into the relay coil loop.

[0026] The following is a dynamic load simulation system, including: programmable power supply module 2-1, load box, multi-sensor array 3-1 and edge computing gateway 3-2, where a) Programmable power supply module 2-1: Outputs DC 24V / AC 220V dual modes, supports voltage sag (100%→50% / 1ms), and harmonic superposition (THD 0-40% adjustable).

[0027] b) Load box: Multi-position resistance / inductance combination load (0.1Ω-10kΩ, 1mH-10H), simulating the load capacity of relay contacts.

[0028] 3. Data acquisition and fusion unit: a) Multi-sensor array 3-1: high-precision temperature sensor (PT100), humidity sensor (capacitive), vibration accelerometer (MEMS), current probe (bandwidth 1 GHz).

[0029] b) Edge Computing Gateway 3-2: Real-time synchronous collection of environmental parameters (temperature, vibration frequency) and relay electrical parameters (pull-in time, contact resistance), with timestamp alignment accuracy ≤1ms.

[0030] The following also includes the cabin structure, temperature control components, humidity control components, and air circulation design. Among them, 1. The cabin structure includes a double-layer stainless steel insulation layer 4-1, which is 50mm thick; the observation window 4-2, tempered glass + heated anti-fog coating.

[0031] 2. The temperature control component includes a semiconductor refrigeration chip array 5-1, model TEC1-12706; a PTC heater 5-2, power of 2kW; and a temperature sensor 5-3, PT100, installed near the relay installation position.

[0032] 3. The humidity control component includes an ultrasonic humidifier 6-1 with a frequency of 1.7 MHz and an atomization volume of 300 mL / h; a condensation dehumidification module 6-2 with semiconductor refrigeration sheets and aluminum fins; and a humidity sensor 6-3, which is capacitive and has an accuracy of ±2% RH.

[0033] 4. The air circulation design includes an axial flow fan 7-1, which is arranged on the side wall of the cabin and has an adjustable wind speed of 0.5-3m / s; the guide plate 7-2 ensures uniform distribution of temperature and humidity.

[0034] like Figure 2 and Figure 3 As shown, the present invention relates to a dynamic reliability evaluation method for a test system simulating different working conditions of a relay, comprising: a multi-environment coupling control method, a dynamic load adaptive matching control method, and a data fusion and life prediction method. The steps of the multi-environment coupling control method are: a) Collaborative control algorithm: Based on fuzzy PID control, it dynamically adjusts the superposition timing of temperature and humidity, vibration frequency, and electromagnetic interference.

[0035] Example: In a low-temperature environment (-20°C), high-frequency vibration (150 Hz) and voltage harmonics (3rd harmonic 30%) are superimposed to simulate the comprehensive operating conditions of a train passing through a switch in cold regions.

[0036] b) Environmental-electrical parameter mapping model: Establish the degradation equation between relay contact time (T) and environmental stress (S): ; : Nominal pull-in time (reference value under stress-free conditions); : Temperature change (unit: °C), which represents the effect of thermal stress on material thermal expansion and coil resistance; RH: Relative humidity (unit: %), reflecting the effect of moisture on contact oxidation and accelerated aging of insulation performance; Avib: vibration acceleration (unit: g), quantifies the interference of mechanical vibration on the stability of armature motion; E: Electromagnetic interference intensity (unit: kV / m), which represents the disturbance of the alternating electric field or pulse magnetic field on the relay magnetic circuit and coil current; k1, k2, k3, k4: Accelerated aging coefficients, corresponding to the sensitivity to temperature, humidity, vibration, and electromagnetic interference acceleration factor, respectively, reflecting the sensitivity of changes in electromagnetic field strength to the pull-in time.

[0037] The steps of the dynamic load adaptive matching control method are: a) Intelligent load switching logic: Automatically matches the standard load (4.8kΩ / 0.44H) according to the relay model (such as JZXC-480) and randomly inserts transient overload (2 times the rated current) during the test.

[0038] b) Harmonic injection strategy: The actual harmonic spectrum of the power grid is analyzed through FFT to generate a synthetic waveform containing the 3rd, 5th, and 7th harmonics with a THD accuracy of ±0.5%.

[0039] The steps of data fusion and life prediction method are: 1. a) Multi-source data correlation analysis: The Kalman filter algorithm is used to fuse vibration acceleration data with contact jitter signals to identify poor contact caused by mechanical wear.

[0040] b) Remaining life prediction model: Based on an LSTM neural network, it inputs environmental stress sequences (temperature, vibration, humidity) and electrical parameters (coil resistance change rate, contact voltage drop) and outputs the remaining life of the relay (MTBF error ≤ 10%). Input environmental stress sequence and electrical parameters. Environmental parameter input includes temperature, humidity, vibration spectrum, and electromagnetic interference intensity. Electrical parameter input includes pull-in time, release time, contact voltage drop, and coil current; 2. Data processing layer Data alignment module, used for timestamp synchronization, accuracy ≤ 1ms; Kalman filtering, which fuses vibration acceleration and contact jitter signals to output an estimated contact resistance value; Feature extraction, extracting degradation features such as temperature coefficient k1 and vibration sensitivity k2.

[0041] 3. Lifespan prediction layer LSTM neural network, the input sequence length is set to 24 hours, the hidden layer nodes are set to 128, the input includes: environmental stress sequence and electrical feature vector, and the output includes: remaining life (hours) and confidence interval (±10%).

[0042] Report generation in PDF / HTML format including degradation curves and maintenance recommendations.

[0043] 4. Feedback control layer It is used to dynamically adjust test conditions, such as "increasing the vibration frequency to the failure threshold". After the adjustment is completed, it points to the environmental simulation area and the electrical simulation area to form a closed-loop control.

[0044] Example 1: Using the present invention to test the cold wave resistance of high-speed rail relays 1. Test conditions: a) Temperature: -30°C → 25°C (cycle once every hour) b) Humidity: 85% RH (constant) c) Vibration: Random vibration (10-50Hz, RMS 0.5g) d) Electromagnetic interference: 1MHz pulse group (5 pulses per second) 2. Testing process: a) Install the JYJXC-160 relay on the test bench and start the multi-environment coupling mode.

[0045] b) Dynamic load module simulates overhead line voltage fluctuation (220V±20%).

[0046] c) The data fusion unit calculates the contact resistance change rate in real time. When it exceeds the threshold (ΔR > 15%), an alarm is triggered and the number of failures is recorded.

[0047] Example 2: Using the present invention to evaluate the vibration resistance life of subway relays: 1. Accelerated aging test: a) The vibration frequency is fixed at 100 Hz (simulating long-term micro-vibration of the track) and the temperature is maintained at 60°C.

[0048] b) The pull-in time and release time are collected every 24 hours and input into the LSTM model to predict the life decay curve.

[0049] 2. Output: Generates a PDF report containing the environmental stress spectrum, degradation trends of key parameters, and recommended replacement periods (e.g., "Remaining life: 8320 hours").

[0050] The above description is only a preferred embodiment of the present invention and is not intended to limit the scope of implementation of the present invention. Any changes, replacements or improvements made to the structures and features described in the scope of the present invention should be included in the scope of protection of the present invention.

Claims

1. A test system for simulating different working conditions of a relay, characterized in that: It includes an environmental simulation area and a dynamic load simulation area, wherein the dynamic load simulation area includes: an electrical simulation area, a data acquisition area, and a data processing and control module; The environmental simulation area is used to simulate the environmental parameters of the relay operation, including temperature, humidity, and mechanical stress. The environmental simulation area is provided with a housing, and heating devices are arranged on the top, bottom, and sides of the housing to accurately adjust the internal temperature. Ventilation holes are provided to adjust air circulation. Temperature sensors and humidity sensors are installed to monitor the environmental temperature and humidity parameters in real time. The area also includes rotating parts, one of which is used to connect to the mechanical structure to simulate mechanical stress. The electrical simulation area is used to simulate the electrical parameters of the relay when it is working, including load current and voltage; the electrical simulation area includes a load simulation circuit connected to the relay under test, and a basic power supply that provides stable power support for the electrical simulation area; The data acquisition area is used to collect environmental parameters and electrical parameters; the data acquisition area includes a data collector that communicates with the data processing and control module via an RS485 bus and realizes data transmission with the help of TCP / IP protocol, as well as auxiliary acquisition equipment; The data processing and control module includes a data synchronization unit, a Kalman filter unit, a feature extraction unit, a neural network life prediction unit and a feedback control unit; The data synchronization unit is used to receive data from the data collection area and synchronize the collected environmental parameters and electrical parameters to ensure the consistency of each parameter data in the time dimension, providing a basis for subsequent accurate analysis; The Kalman filter unit is connected to the data synchronization unit, receives the synchronized data, and uses the Kalman filter algorithm to filter the data to remove noise caused by external interference, sensor errors and other factors, thereby improving the quality and reliability of the data; The feature extraction unit is connected to the Kalman filter unit to extract effective feature parameters closely related to relay performance and life from the filtered data, such as but not limited to temperature change rate, current peak value, voltage fluctuation amplitude, and humidity change trend; The neural network life prediction unit is connected to the feature extraction unit, and uses the extracted effective feature parameters in combination with a neural network model pre-trained with a large amount of experimental data to predict the remaining service life of the relay and generate a detailed life prediction report, which includes the predicted life span and performance change trends at different stages; The feedback control unit is connected to the neural network life prediction unit, the environmental simulation area and the electrical simulation area respectively, and analyzes the working status and performance changes of the relay according to the life prediction report. If the prediction result shows that the life of the relay may be adversely affected, the feedback control unit will automatically adjust the temperature, humidity, and mechanical stress parameters of the environmental simulation area, as well as the load current and voltage parameters of the electrical simulation area, thereby forming a closed-loop control to put the relay in a condition that conforms to actual working conditions and is conducive to accurately testing its performance.

2. A test system for simulating different working conditions of a relay according to claim 1, characterized in that: The environmental simulation area includes a temperature and humidity control cabin (1-1), a vibration simulation table (1-2), and an electromagnetic interference generator (1-3); the temperature and humidity control cabin (1-1) and the vibration simulation table (1-2) are fixed by a mechanical bracket, and the electromagnetic interference generator (1-3) is connected to the relay coil (8-2) under test via a ring current probe; The electrical simulation area includes a programmable power supply module (2-1) and a dynamic load box (2-2), wherein the output end of the programmable power supply module (2-1) is connected to the corresponding contact of the relay coil (8-1) under test, and the dynamic load box (2-2) is connected in parallel to the corresponding contact of the relay coil (8-1) under test; The data acquisition unit and the data processing control unit include a multi-sensor array (3-1), an edge computing gateway (3-2), and a host computer (3-3); the multi-sensor array (3-1) is connected to the edge computing gateway (3-2) through an interface, and the edge computing gateway (3-2) is electrically connected to the host computer through Ethernet; The data acquisition unit and the data processing control unit collect environmental parameters and electrical parameters such as temperature, vibration, electromagnetic interference in the environmental simulation area and voltage, current, and contact resistance in the electrical simulation area, and simultaneously issue relevant control instructions including temperature and humidity set value instructions and vibration frequency instructions; Temperature and humidity control chamber (1-1): uses a composite temperature control structure of semiconductor refrigeration chip and PTC heater to achieve fast switching from -40℃ to 85℃ (heating rate ≥5℃ / min), integrates ultrasonic humidifier and condensation dehumidification module, and has a humidity control range of 5%~95% RH with an accuracy of ±2%; Vibration simulation table (1-2): uses a six-degree-of-freedom electromagnetic vibration table with a frequency range of 0-200Hz and a maximum acceleration of 10g. It supports random vibration and sine sweep modes. It has a built-in relay mounting fixture and is suitable for relays of different sizes, such as JWXC-1700 and JYJXC-160 / 260. Electromagnetic interference generator (1-3): uses a programmable high-frequency pulse group generator (1MHz-100MHz, peak voltage ±5kV) to simulate signal cable coupling interference, equipped with a ring current probe to directly inject common mode / differential mode interference into the relay coil loop; The electrical simulation area, data acquisition area, and data processing and control module in the dynamic load simulation area complete the following work steps: a) Programmable power supply module: outputs in dual modes of 24V DC / 220V AC, supports voltage sag of 100% to 50% / 1ms, and THD of harmonic superposition is adjustable from 0-40%; b) Load box: multi-position resistance / inductance combination load 0.1Ω-10kΩ, 1mH-10H, simulates relay contact load capacity; The data acquisition and fusion unit includes: a) Multi-sensor array: high-precision temperature sensor: PT100, capacitive humidity sensor, vibration accelerometer, current probe with 1 GHz bandwidth; b) Edge computing gateway: Real-time synchronous collection of environmental parameters: temperature, vibration frequency and relay electrical parameters such as pull-in time and contact resistance, with timestamp alignment accuracy ≤1ms.

3. A test system for simulating different working conditions of a relay according to claim 1, characterized in that: The vent is provided with a ventilation regulating device, which includes a fan or a damper for regulating air circulation.

4. A test system for simulating different working conditions of a relay according to claim 1, characterized in that: The load simulation circuit can accurately simulate currents and voltages of different amplitudes, as well as electrical signals with complex waveforms according to actual needs.

5. A test system for simulating different working conditions of a relay according to claim 1, characterized in that: The data collector area is connected to sensors in the environmental simulation area and the electrical simulation area, and the auxiliary collection equipment works in coordination with the data collector to ensure the comprehensiveness and accuracy of data collection.

6. A method for testing a system for simulating different conditions of a relay working environment, characterized in that: The method comprises the following steps: a multi-environment coupling control method, a dynamic load adaptive matching control method, and a data fusion and life prediction method; The steps of the multi-environment coupling control method are: a) Collaborative control algorithm: Based on fuzzy PID control, the algorithm dynamically adjusts the superposition timing of temperature, humidity, vibration frequency, and electromagnetic interference. For example, in a low-temperature environment, the algorithm superimposes 150Hz high-frequency vibration and the third harmonic 30% voltage harmonic to simulate the comprehensive operating conditions of a train passing through a switch in cold regions. b) Environmental-electrical parameter mapping model: Establish the degradation equation between relay contact time (T) and environmental stress (S): ; Parameter Description: Nominal pull-in time (reference value under stress-free conditions); Temperature change (unit: °C), which represents the effect of thermal stress on material thermal expansion and coil resistance; RH: Relative humidity (unit: %), reflecting the effect of moisture on contact oxidation and accelerated aging of insulation performance; Avib: vibration acceleration (unit: g), quantifies the interference of mechanical vibration on the stability of armature motion; E: Electromagnetic interference intensity (unit: kV / m), which represents the disturbance of the alternating electric field or pulse magnetic field on the relay magnetic circuit and coil current; also called the pull-in time (base value under stress-free conditions); k1, k2, k3, k4: Accelerated aging coefficients, corresponding to the sensitivity to temperature, humidity, vibration, and electromagnetic interference acceleration factor, respectively, reflecting the sensitivity of changes in electromagnetic field strength to the pull-in time.

7. The method for testing a system for simulating different working conditions of a relay according to claim 6, characterized in that: The steps of the dynamic load adaptive matching control method are: a) Intelligent load switching logic: Automatically matches the standard load (4.8kΩ / 0.44H) according to the relay model (such as JZXC-480) and randomly inserts transient overload (2 times the rated current) during the test; b) Harmonic injection strategy: The actual harmonic spectrum of the power grid is analyzed through FFT to generate a synthetic waveform containing the 3rd, 5th, and 7th harmonics with a THD accuracy of ±0.5%.

8. The method for testing a system for simulating different working conditions of a relay according to claim 6, characterized in that: The steps of data fusion and life prediction method are:

1. Parameter settings a) Multi-source data correlation analysis: Using the Kalman filter algorithm to fuse vibration acceleration data with contact jitter signals, we can identify poor contact caused by mechanical wear. b) Remaining life prediction model: Based on an LSTM neural network, it inputs environmental stress sequences (temperature, vibration, humidity) and electrical parameters (coil resistance change rate, contact voltage drop) and outputs the remaining life of the relay (MTBF error ≤ 10%). Input environmental stress sequence and electrical parameters. Environmental parameter input includes temperature, humidity, vibration spectrum, and electromagnetic interference intensity. Electrical parameter input includes pull-in time, release time, contact voltage drop, and coil current; 2. Data processing layer Data alignment module, used for timestamp synchronization, accuracy ≤ 1ms; Kalman filtering, which fuses vibration acceleration and contact jitter signals to output an estimated contact resistance value; Feature extraction, extracting degradation features such as temperature coefficient k1 and vibration sensitivity k2; 3. Lifespan prediction layer LSTM neural network, with an input sequence length of 24 hours and 128 hidden layer nodes. The inputs include environmental stress sequence and electrical feature vectors, and the outputs include remaining life (hours) and confidence interval (±10%). Report generation in PDF / HTML format, including degradation curves and maintenance recommendations; 4. Feedback control layer It is used to dynamically adjust test conditions, such as "increasing the vibration frequency to the failure threshold." After the adjustment is completed, it points to the environmental simulation area and the electrical simulation area to form a closed-loop control.

9. The method for testing a system for simulating different working conditions of a relay according to claim 8, characterized in that: Parameter setting: According to the actual working environment and electrical requirements of the relay, set the temperature, humidity, and mechanical stress parameters of the environmental simulation area, as well as the load current and voltage parameters of the electrical simulation area; start the data acquisition system; Test operation: Place the relay under test on the test platform in the environmental simulation area and connect it to the load simulation circuit in the electrical simulation area. Start the equipment in the environmental simulation area and the electrical simulation area to simulate the working environment and electrical conditions of the relay. The data acquisition system collects environmental and electrical parameters in real time and transmits the data to the data processing and control module. Data processing and life prediction: The data synchronization unit synchronizes the collected data; the Kalman filter unit filters the synchronized data; the feature extraction unit extracts effective feature parameters from the filtered data; the neural network life prediction unit uses the extracted feature parameters in combination with a pre-trained neural network model to predict the life of the relay and generate a life prediction report; Feedback control: The feedback control unit analyzes the working status and performance changes of the relay based on the life prediction report, and adjusts the parameters of the environmental simulation area and the electrical simulation area to put the relay under more appropriate test conditions; Repeat the test process until the scheduled test time is reached or the test requirements are met.

Images