Chain transmission system fault detection and diagnosis test device, method and system

Abstract

The invention discloses a chain transmission system fault detection and diagnosis test device, method and system, and relates to the technical field of chain transmission system fault detection.The chain transmission system fault detection and diagnosis test device comprises a driving system, a transmission system, a load system, a signal detection system and an auxiliary system; the real-time monitoring and data acquisition of each key component of the chain transmission system are realized; more comprehensive fault information can be provided, the accuracy and reliability of fault detection can be improved, timely discovery and early warning can be realized, and the fault occurrence probability and loss can be reduced; through the adjustable load system, the running states of the chain transmission system under different working conditions can be simulated, the performance and reliability of the system under different loads can be verified, efficient and accurate fault detection and diagnosis of the chain transmission system are achieved, and the safety of equipment running is improved.

Description

Technical Field

[0001] The present invention relates to the technical field of chain transmission system fault detection, and in particular to a chain transmission system fault detection and diagnosis test device, method and system. Background Art

[0002] As one of the key equipment in the comprehensive mining face, the stable mechanical properties of the scraper conveyor are crucial to the safety of underground production. The performance of its chain drive system affects the operating efficiency and reliability of the entire system. The operating conditions of the scraper conveyor are very harsh, and the chain drive system has a long transmission distance. In addition to being affected by factors such as the bending and tilting of the working face, the meshing polygon effect, and the material, wear, and corrosion of each component, it is also affected by variable load conditions. This directly leads to a high failure rate of key components of the chain drive system, affecting the safe operation of the equipment. At present, the methods for fault detection and diagnosis of chain drive systems have limited detection accuracy, and it is difficult to achieve real-time detection of chain drive systems. A single detection method often cannot fully reflect the operating status of the chain drive system, and lacks comprehensiveness and comprehensiveness. Therefore, it is necessary to combine multiple detection methods to achieve efficient, accurate, and real-time monitoring and diagnosis of chain drive systems. Summary of the invention

[0003] The purpose of the present invention is to solve the above technical problems and provide a chain drive system fault detection and diagnosis test device, method and system to achieve efficient and accurate fault detection and diagnosis of the chain drive system.

[0004] In order to achieve the above technical objectives and meet the above technical requirements, the technical solution adopted by the present invention is: a chain transmission system fault detection and diagnosis test device, including a drive system, a transmission system, a load system, a signal detection system, and an auxiliary system, wherein the drive system includes a motor, the motor output shaft is connected to the reducer through an active end coupling, and the reducer output shaft is connected to the transmission system; The transmission system is a transmission system to be tested on a scraper conveyor, comprising a sprocket, wherein the sprocket is arranged on a sprocket vertical bearing seat through a sprocket shaft, the output shaft of the reducer drives the sprocket shaft to drive the sprocket to rotate, the chain ring is arranged on the sprocket, and a scraper is arranged between the chain rings on both sides; The load system includes a magnetic powder brake for braking the transmission system, one end of the magnetic powder brake is connected to the transmission system through a load coupling to transmit torque, and the other end is connected to a water cooling device; The signal detection system comprises: A current sensor, wherein the current sensor is used to detect the current value of the motor; A voltage sensor, wherein the voltage sensor is used to measure the voltage value of the motor; A strain sensor is arranged on the scraper and transmits data through a wireless gateway to a data acquisition and analysis system; A vibration sensor, which is arranged on the sprocket vertical bearing seat and is used to measure load end vibration data; A torque sensor, the torque sensor is used to measure the torque and rotation speed of the sprocket shaft; A temperature sensor is used to monitor the temperature of the magnetic powder brake in real time.

[0005] Preferably: the auxiliary system includes an L-shaped base, a type I base, a tensioning structure, a material drop bin and a support platform, the support platform is provided with a guide rail, the L-shaped base and the type I base are arranged on the guide rail through a slider, the tensioning structure is arranged on both sides of the L-shaped base and the type I base through a fixed bracket, the mounting holes on the support platform and the fixed bracket are connected by bolts, and nuts are tightened on the other side, and the tensioning structure is adjusted by an adjusting mechanism to achieve tensioning of the chain.

[0006] The chain drive system fault detection and diagnosis test method adopts the chain drive system fault detection and diagnosis test device, including the following steps: S100: Install and debug the chain drive system fault detection and diagnosis test device; S200: Start the motor, adjust the frequency, obtain the reading from the torque sensor, and determine the range; adjust the current of the magnetic powder brake until the test bench stops, and read the current reading corresponding to the torque; S300: Determine the test parameters, run stably for a period of time, the current sensor and the voltage sensor record the current and voltage of the motor; the strain sensor obtains the data of the chain link during the stable operation process; the vibration sensor tests the speed fluctuation during the operation process; the temperature sensor detects the temperature during the operation process and records the data; S400: cleaning and normalizing the data collected by the sensor; S500: Build an LSTM model, import data into the optimized LSTM prediction model, determine the warning threshold, and perform diagnosis and monitoring in the host computer.

[0007] Preferably, step S100 includes: S101: Determine the specific location where the chain drive system fault detection and diagnosis test device is installed; S102: Adjust the chain link tension, calculate the chain link tension through the hydraulic cylinder reading, compare it with the actual value, and determine the parameters.

[0008] Preferably, step S200 includes: S201: Start the motor and gradually increase the frequency to a preset value; S202: adjusting the operating frequency of the motor so that the torque sensor obtains readings and determines an initial range; S203: Adjust the current of the magnetic powder brake to 2000 N·m, monitor the stopping condition of the sprocket, and read the torque corresponding to the current indication.

[0009] Preferably, step S400 includes: S401: Collect current, voltage, strain, vibration, torque and temperature through a signal detection system; S402: Perform data cleaning, the cleaning process includes removing noise, processing missing data, and eliminating outliers; S403: performing min-max normalization processing on the cleaned data; S404: Divide the processed data into a training set, a validation set, and a test set according to proportion.

[0010] Preferably, step S500 includes: S501: Build LSTM model structure; S502: Set a training batch, use the training set data to train the LSTM prediction model, continuously iterate through the Adam algorithm to reach the maximum number of iterations, calculate the loss error to optimize the model parameters, obtain the optimal model, store the LSTM model parameters, and establish the LSTM prediction model; The data of the training set include: motor current value, motor voltage value, scraper strain value, load end sprocket vibration signal, sprocket shaft torque and speed, and magnetic powder brake temperature; S503: randomly divide the data set into a training set and a test set, fit and compare the prediction results of the LSTM prediction model with the actual operation data of the test bench, and select the mean square error to evaluate the LSTM prediction model; S504: inputting historical data into the trained LSTM prediction model, the LSTM prediction model outputs the prediction value, and the maximum residual is calculated by the sliding weighted average method as the warning threshold; S505: extracting normal operating data from the database and importing it into the LSTM model for prediction, and processing the residual; S506: The residual result and the warning threshold are displayed in the upper computer. When the model prediction residual exceeds this threshold, it is considered that the chain drive system may have a fault and a warning is triggered.

[0011] The chain transmission system fault detection and diagnosis test system includes a chain transmission system fault detection and diagnosis test device, a data acquisition and analysis system, an algorithm diagnosis and monitoring system, The data acquisition and analysis system includes a data acquisition module and a data analysis module. The data acquisition module is used to receive the current, voltage, strain, vibration, torque and temperature collected by the signal detection system. The data analysis module receives the data from the data acquisition module and cleans and normalizes the data. The algorithm diagnosis and monitoring system calculates and updates the network parameters of the LSTM model, constructs the LSTM-Adam prediction model, and uses the maximum residual of the same type of data under normal operating conditions as the warning threshold to perform fault diagnosis.

[0012] Compared with the traditional structure, the invention has the following beneficial effects: 1. The present invention realizes real-time monitoring and data collection of each key component of the chain transmission system by integrating multiple sensors; it can provide more comprehensive fault information, improve the accuracy and reliability of fault detection, discover and warn in time, and reduce the probability of fault occurrence and losses; through the adjustable load system, it can simulate the operating state of the chain transmission system under different working conditions, verify the performance and reliability of the system under different loads, realize efficient and accurate fault detection and diagnosis of the chain transmission system, and improve the safety of equipment operation.

[0013] 2. The present invention optimizes the LSTM model by using the Adam algorithm. The Adam algorithm independently designs an adaptive learning rate for each model parameter, fully utilizing the influence of the learning rate on the model performance and improving the model training speed. An LSTM-Adam prediction model is proposed, and the mean square error is selected to evaluate the model. The sliding weighted average method uses a sliding time window for data calculation and considers the weighted average of the data, which can improve the reliability and accuracy of the data. The sliding weighted average method is used to perform residual analysis on the normal operating condition data to obtain its warning threshold. BRIEF DESCRIPTION OF THE DRAWINGS

[0014] Figure 1 Design a process flow chart for the test device and test method of the present invention; Figure 2 It is a schematic diagram of the structure of the test device of the present invention; Figure 3 A top view of the present invention; Figure 4 It is a schematic diagram of the structure of the drive system of the present invention; Figure 5 It is a schematic diagram of the transmission system structure of the present invention; Figure 6 It is a schematic diagram of the structure of the signal detection system of the present invention; Figure 7 It is a schematic diagram of the structure of the load system of the present invention; Figure 8 It is a schematic diagram of the auxiliary system structure of the present invention; Fig. 9 This is a framework diagram of the data acquisition and analysis system of the present invention; Fig.10 Diagnostic model diagram of the algorithm of the present invention; Fig.11 It is the framework diagram of the monitoring system of the present invention; Fig.12 Establish a process for the LSTM-Adam prediction model of the present invention; Fig.13 This is the early warning judgment process of the present invention; In the figure: 1-driving system, 1.1-motor, 1.2-motor pad, 1.3-active end coupling, 1.4-reducer, 1.5-reducer pad; 2-transmission system, 2.1-sprocket, 2.2-chain ring, 2.3-scraper, 2.4-sprocket shaft, 2.5-sprocket vertical bearing seat, 2.6-chain connecting ring; 3-signal detection system, 3.1-current sensor, 3.2-voltage sensor, 3.3-strain sensor, 3.4-vibration sensor, 3.5-torque speed sensor, 3.6-temperature sensor; 4-load system, 4.1-load coupling, 4.2-magnetic powder brake, 4.3-water cooling device, 4.4-vertical pad; 5-auxiliary system, 5.1-L-type base, 5.2-type I base, 5.3-tensioning structure, 5.4-dropping bin, 5.5-support table. DETAILED DESCRIPTION

[0015] The present invention will be further described below.

[0016] See attached Figure 2-3 , a chain transmission system fault detection and diagnosis test device, comprising a drive system 1, a transmission system 2, a load system 4, a signal detection system 3, and an auxiliary system 5, like Figure 4 As shown, the drive system 1 includes a motor 1.1, the motor output shaft is connected to the reducer 1.4 through the active end coupling 1.3, and is fixed on the motor pad 1.2; the reducer 1.4 is fixed on the reducer pad 1.5, and the power is transmitted to the transmission system 2 through the reducer output shaft 1.5 to realize the function of variable speed torque, and the motor pad 1.2 and the reducer pad 1.5 are fixed to the L-shaped base 5.1 by welding to provide power to the entire chain transmission system; like Figure 5As shown, the transmission system 2 includes a sprocket 2.1, which is arranged on a sprocket vertical bearing seat 2.5 through a sprocket shaft 2.4 and fixed by bolts. The sprocket 2.1 and the sprocket shaft 2.4 are connected by keys to prevent the axis from being misaligned. The output shaft of the reducer drives the sprocket shaft 2.4 to drive the sprocket 2.1 to rotate; the chain ring 2.2 is arranged on the sprocket 2.1, and the chain ring 2.2 is meshed with the sprocket 2.1 to form a complete double-row chain, and the flat ring and the vertical ring are alternately connected; a scraper 2.3 is arranged between the chain rings 2.2 on both sides, and the scraper 2.3 is connected to the vertical ring in the chain ring 2.2, and the upper and lower parts are detachable and fixed to the flat ring by bolts; the vertical connecting chain ring 2.6 is installed at the vertical ring, connecting the flat rings at both ends, and elastic pins are driven into the holes for fixing, which is convenient for installation and disassembly.

[0017] like Figure 7 As shown, the load system 4 includes a magnetic powder brake 4.2 for braking the transmission system 2. The magnetic powder brake 4.2 is welded on a vertical pad 4.4 and fixed on a type-one base 5.2 by bolts. One end of the magnetic powder brake 4.2 is connected to the transmission system 2 through a load coupling 4.1 to transmit torque, and the other end is connected to a water cooling device 4.3 to prevent the magnetic powder brake 4.2 from overheating and ensure test safety.

[0018] like Figure 6 As shown, the signal detection system 3 includes: A current sensor 3.1, which is fixedly mounted on the L-shaped base 5.1 and is used to detect the current value of the motor 1.1; A voltage sensor 3.2, which is fixed to the other side of the L-shaped base 5.1 and is used to measure the voltage value of the motor 1.1; Strain sensor 3.3, which is arranged on the scraper 2.3 and transmits data through a wireless gateway to enter the data acquisition and analysis system; A vibration sensor 3.4, which is attached to the sprocket vertical bearing seat 2.5 and is used to measure load end vibration data; A torque sensor 3.5, which is located between the magnetic powder brake output shaft 4.1 and the sprocket shaft 2.4 and is used to measure the torque and rotation speed of the sprocket shaft 2.4; The temperature sensor 3.6 is attached to a base 5.2 of the magnetic powder brake 4.2 and is used for real-time monitoring of the temperature of the magnetic powder brake 4.2.

[0019] like Figure 8As shown, the auxiliary system 5 is used to install the chain transmission system fault detection and diagnosis test device, and the auxiliary system 5 includes an L-shaped base 5.1, a type-one base 5.2, a tensioning structure 5.3, a material drop bin 5.4, and a support platform 5.5. The support platform 5.5 is provided with a guide rail, and the L-shaped base 5.1 and the type-one base 5.2 are arranged on the guide rail through a slider to facilitate the adjustment and movement of the entire device. The tensioning structure 5.3 is arranged on both sides of the L-shaped base 5.1 and the type-one base 5.2 through a fixed bracket, and the mounting holes on the support platform 5.5 and the fixed bracket are connected by bolts, and nuts are tightened on the other side. The tensioning structure 5.3 is adjusted by the adjustment mechanism to achieve the tensioning of the chain and ensure the stable operation of the transmission system.

[0020] Chain drive system fault detection and diagnosis test system, including chain drive system fault detection and diagnosis test device, data acquisition and analysis system, algorithm diagnosis and monitoring system, The data acquisition and analysis system uses sensors, wireless gateways and other hardware devices to collect data, establish communication between the equipment and the host computer, ensure real-time data transmission, and display it in the host computer software through the data analysis module.

[0021] The data acquisition and analysis system includes: A data acquisition module, the data acquisition module is used to receive current, voltage, strain, vibration, torque and temperature collected by the signal detection system; A data analysis module receives data from the data acquisition module and performs cleaning and normalization processing on the data; The algorithm diagnosis and monitoring system saves the collected data to the SQL Server database in real time, uses the min-max normalization method to process the sample data, and divides the preprocessed data into training set and test set in a ratio of 8:2; the algorithm diagnosis and monitoring system calculates and updates the network parameters of the LSTM model, constructs the LSTM-Adam prediction model, and uses the maximum residual of the same type of data under normal operating conditions as the warning threshold for fault diagnosis.

[0022] The chain drive system fault detection and diagnosis test method uses a chain drive system fault detection and diagnosis test device to conduct the test, including the following steps: S100: Install and debug the chain drive system fault detection and diagnosis test device; S200: Start the motor, adjust the frequency, obtain the reading from the torque sensor, and determine the range; adjust the magnetic powder brake current (2000N·m) until the test bench stops, and read the current reading corresponding to the torque; S300: Determine the test parameters, run stably for a period of time, the current sensor and the voltage sensor record the current and voltage of the motor; the strain sensor obtains the data of the chain link during the stable operation process; the vibration sensor tests the speed fluctuation during the operation process; the temperature sensor detects the temperature during the operation process and records the data; S400: importing the data collected by the sensor into the SQL Server database, extracting the data from the SQL Server database, and performing data cleaning and min-max normalization processing; S500: Build an LSTM model, import data into the optimized LSTM prediction model, determine the warning threshold, and perform diagnosis and monitoring in the host computer.

[0023] In this preferred embodiment, step S100 includes the following steps: S101: Determine the specific location for the chain drive system fault detection and diagnosis test device to be installed, install the auxiliary system, drive system, transmission system, load system, signal detection system, data acquisition and analysis system, and algorithm diagnosis and monitoring system, and ensure a stable foundation; S102: The chain tension is adjusted through the tensioning structure, and the chain link tension is calculated through the hydraulic cylinder indication and compared with the actual value to determine the parameters.

[0024] In this preferred embodiment, step S200 includes the following steps: S201: Start the motor using the control system and gradually increase the frequency to a preset value; S202: adjusting the operating frequency of the motor by means of a frequency converter so that the torque sensor obtains a reading and determines an initial range; S203: Adjust the current of the magnetic powder brake to 2000 N·m, monitor the stopping status of the test bench through the console, and read the torque corresponding to the current indication.

[0025] In this preferred embodiment, step S400 includes the following steps: S401: First, the signal detection system collects sensor data such as current, voltage, strain, vibration, torque and temperature. These data are transmitted to the data acquisition system through the wireless transmission module and then stored in the SQL Server database.

[0026] S402: After extracting the original data from the SQL Server database, data cleaning is performed. The cleaning process includes steps such as removing noise, processing missing data, and removing outliers to ensure data quality.

[0027] S403: Perform min-max normalization on the cleaned data. Normalization can scale the data to the same range, which helps to improve the training effect of the LSTM model.

[0028] S404: Divide the processed data into a training set, a validation set, and a test set according to a certain ratio. The training set accounts for 60%, the validation set accounts for 20%, and the test set accounts for 20%.

[0029] In this preferred embodiment, step S500 includes the following steps: S501: Build the LSTM model structure, including the input layer, multiple LSTM layers, a fully connected layer, and an output layer. The number of LSTM layers and the number of neurons in each layer can be determined through experiments to achieve the best model performance; S502: Set a training batch, use the training set data to train the LSTM prediction model, continuously iterate through the Adam algorithm to reach the maximum number of iterations, calculate the loss error to optimize the model parameters, obtain the optimal model, store the LSTM model parameters, and establish the LSTM prediction model; The data of the training set include: motor current value, motor voltage value, scraper strain value, load end sprocket vibration signal, sprocket shaft torque and speed, and magnetic powder brake temperature; S503: randomly divide the data set into a training set and a test set, fit and compare the prediction results of the LSTM prediction model with the actual operation data of the test bench, and select the mean square error to evaluate the LSTM prediction model; Fitting comparison: The prediction results of the LSTM-Adam prediction model are fitted and compared with the actual operation data of the test bench, and the monitoring point data at a certain interval are selected to compare the data differences at that location.

[0030] Evaluation process: The mean squared error (MSE) is selected as the evaluation indicator of the model. The mean squared error is the average of the squares of the differences between the predicted value and the actual value. The formula is as follows:

[0031] in, is the actual value, is the predicted value, n is the number of data in the test set. The smaller the MSE, the closer the model's prediction result is to the actual value, and the better the model performance is. It reflects the degree of difference between the actual value and the predicted value.

[0032] S504: inputting historical data into the trained LSTM prediction model, the LSTM prediction model outputs the prediction value, and the maximum residual is calculated by the sliding weighted average method as the warning threshold; The historical data includes fault data and normal working condition data. The fault data includes simulating the chain drive system fault on the scraper conveyor test bench by increasing the scraper running resistance, increasing the amount of coal gangue, and replacing the chain that is about to break due to fatigue, and the collected fault data is stored in the database. Collect multiple groups of data under working conditions as data sets, including current and voltage data, scraper strain data, vibration data, torque and speed, and temperature data.

[0033] Calculate the maximum residual process:

[0034] Where: For calculation point The corresponding value; for The weight of .

[0035] The weight of each point also needs to be determined in the formula. Considering that the purpose of using the sliding weighted average method to process the residual is to use the residual mean obtained by sliding weighting to replace the residual of the point, it is chosen to give the current point a separate weight. , the remaining points are given the same weight = 1, then calculate the point The sliding weighted average of

[0036] The maximum residual process is to fit and compare the historical data obtained by the chain drive system sensor (motor current value, motor voltage value, scraper strain value, load-end sprocket vibration signal, sprocket shaft torque and speed, and magnetic powder brake temperature) with the predicted data of the established LSTM prediction model, calculate it through the sliding weighted average method, calculate the residual between the actual value and the predicted value, and select the maximum residual as the threshold.

[0037] After that, the scraper conveyor is operated normally, real-time data is collected, and the data is input into the established LSTM prediction model for prediction, so as to obtain new residual data values. As long as the residual data does not exceed the maximum threshold just obtained, it is considered to be normal operation.

[0038] S505: extracting normal operating data from the database and importing it into the LSTM model for prediction, and processing the residual; S506: The residual result and the warning threshold are displayed in the upper computer. When the model prediction residual exceeds this threshold, it is considered that the chain drive system may have a fault and a warning is triggered.

[0039] The above embodiments of the present invention are merely examples to clearly illustrate the present invention, but are not intended to limit the protection scope of the present invention. All equivalent technical solutions also belong to the scope of the present invention. The patent protection scope of the present invention should be defined by the claims.

Claims

1. Chain drive system fault detection and diagnosis test device, characterized by: The invention comprises a drive system (1), a transmission system (2), a load system (4), a signal detection system (3), and an auxiliary system (5); the drive system (1) comprises a motor (1.1); the motor output shaft is connected to a reducer (1.4) via an active end coupling (1.3); and the reducer output shaft is connected to the drive system (2); The transmission system (2) is a transmission system to be tested on a scraper conveyor, comprising a sprocket (2.1), wherein the sprocket (2.1) is arranged on a sprocket vertical bearing seat (2.5) via a sprocket shaft (2.4), the output shaft of the reducer drives the sprocket shaft (2.4) to drive the sprocket (2.1) to rotate, the chain ring (2.2) is arranged on the sprocket (2.1), and a scraper (2.3) is arranged between the chain rings (2.2) on both sides; The load system (4) comprises a magnetic powder brake (4.2) for braking the transmission system (2); one end of the magnetic powder brake (4.2) is connected to the transmission system (2) via a load coupling (4.1) to transmit torque, and the other end is connected to a water cooling device (4.3); The signal detection system (3) comprises: A current sensor (3.1), wherein the current sensor (3.1) is used to detect a current value of the motor (1.1); A voltage sensor (3.2), wherein the voltage sensor (3.2) is used to measure a voltage value of the motor (1.1); A strain sensor (3.3), wherein the strain sensor is arranged on the scraper (2.3) and transmits data through a wireless gateway to enter a data acquisition and analysis system; A vibration sensor (3.4), wherein the vibration sensor (3.4) is arranged on the sprocket vertical bearing seat (2.5) and is used to measure load end vibration data; A torque sensor (3.5), wherein the torque sensor (3.5) is used to measure the torque and rotation speed of the sprocket shaft (2.4); A temperature sensor (3.6) is used to monitor the temperature of the magnetic powder brake (4.2) in real time.

2. The chain drive system fault detection and diagnosis test system according to claim 1, characterized in that: The auxiliary system (5) comprises an L-shaped base (5.1), a type I base (5.2), a tensioning structure (5.3), a material drop bin (5.4) and a support platform (5.5); a guide rail is arranged on the support platform (5.5); the L-shaped base (5.1) and the type I base (5.2) are arranged on the guide rail via a slider; the tensioning structure (5.3) is arranged on both sides of the L-shaped base (5.1) and the type I base (5.2) via a fixed bracket; the tensioning structure (5.3) is adjusted via an adjustment mechanism to achieve tensioning of the chain.

3. Chain drive system fault detection and diagnosis test method, characterized in that: The chain drive system fault detection and diagnosis test device according to claim 1 comprises the following steps: S100: Install and debug the chain drive system fault detection and diagnosis test device; S200: Start the motor, adjust the frequency, obtain the reading from the torque sensor, and determine the range; adjust the current of the magnetic powder brake until the test bench stops, and read the current reading corresponding to the torque; S300: Determine the test parameters, run stably for a period of time, the current sensor and the voltage sensor record the current and voltage of the motor; the strain sensor obtains the data of the chain link during the stable operation process; the vibration sensor tests the speed fluctuation during the operation process; the temperature sensor detects the temperature during the operation process and records the data; S400: cleaning and normalizing the data collected by the sensor; S500: Build an LSTM model, import data into the optimized LSTM prediction model, determine the warning threshold, and perform diagnosis and monitoring in the host computer.

4. The chain drive system fault detection and diagnosis test method according to claim 3, characterized in that: Step S100 includes: S101: Determine the specific location where the chain drive system fault detection and diagnosis test device is installed; S102: Adjust the chain link tension, calculate the chain link tension through the hydraulic cylinder reading, compare it with the actual value, and determine the parameters.

5. The chain drive system fault detection and diagnosis test method according to claim 3, characterized in that: Step S200 includes: S201: Start the motor and gradually increase the frequency to a preset value; S202: adjusting the operating frequency of the motor so that the torque sensor obtains readings and determines an initial range; S203: Adjust the current of the magnetic powder brake to 2000 N·m, monitor the stopping condition of the sprocket, and read the torque corresponding to the current indication.

6. The chain drive system fault detection and diagnosis test method according to claim 3, characterized in that: Step S400 includes: S401: Collect current, voltage, strain, vibration, torque and temperature through a signal detection system; S402: Perform data cleaning, the cleaning process includes removing noise, processing missing data, and eliminating outliers; S403: performing min-max normalization processing on the cleaned data; S404: Divide the processed data into a training set, a validation set, and a test set according to proportion.

7. The chain drive system fault detection and diagnosis test method according to claim 3, characterized in that: Step S500 includes: S501: Build LSTM model structure; S502: Set a training batch, use the training set data to train the LSTM prediction model, iterate continuously through the Adam algorithm to reach the maximum number of iterations, calculate the loss error to optimize the model parameters, obtain the optimal model, store the LSTM model parameters, and establish the LSTM prediction model; The data of the training set include: motor current value, motor voltage value, scraper strain value, load end sprocket vibration signal, sprocket shaft torque and speed, and magnetic powder brake temperature; S503: randomly divide the data set into a training set and a test set, fit and compare the prediction results of the LSTM prediction model with the actual operation data of the test bench, and select the mean square error to evaluate the LSTM prediction model; S504: inputting historical data into the trained LSTM prediction model, the LSTM prediction model outputs the prediction value, and the maximum residual is calculated by the sliding weighted average method as the warning threshold; S505: extracting normal operating data from the database and importing it into the LSTM model for prediction, and processing the residual; S506: The residual result and the warning threshold are displayed in the upper computer. When the model prediction residual exceeds this threshold, it is considered that the chain drive system may have a fault and a warning is triggered.

8. A chain drive system fault detection and diagnosis test system, comprising the chain drive system fault detection and diagnosis test device according to claim 1, characterized in that: It also includes data acquisition and analysis systems, algorithm diagnosis and monitoring systems, The data acquisition and analysis system includes a data acquisition module and a data analysis module. The data acquisition module is used to receive the current, voltage, strain, vibration, torque and temperature collected by the signal detection system. The data analysis module receives the data from the data acquisition module and cleans and normalizes the data. The algorithm diagnosis and monitoring system calculates and updates the network parameters of the LSTM model, constructs the LSTM-Adam prediction model, and uses the maximum residual of the same type of data under normal operating conditions as the warning threshold to perform fault diagnosis.

Images