Electric cylinder system fault warning system based on artificial intelligence

Abstract

The present invention discloses an electric cylinder system fault warning system based on artificial intelligence, which relates to the technical field of intelligent warning, and solves the technical problems that it is difficult to identify and analyze the operation trajectory of the electric cylinder system according to the collected and processed image data, and it is difficult to perform initial fault warning processing according to the recognition result; secondly, it is difficult to measure and analyze the operation variables of the electric cylinder system; then, it is difficult to comprehensively analyze the fault warning value of the electric cylinder according to various data; finally, it is difficult to dynamically set the threshold for real-time warning according to the fault warning value of the electric cylinder. The present invention includes the following modules: a data acquisition module, a data processing module, an initial fault warning module, a data analysis module and a fault warning module. The present invention analyzes the operation trajectory of the electric cylinder system, measures and analyzes the electric cylinder operation variables according to the analysis results; obtains the fault warning value of the electric cylinder through the comprehensive data, and performs real-time warning through the dynamic threshold.

Description

Technical Field

[0001] The present invention belongs to the technical field of intelligent early warning, and in particular is an electric cylinder system fault early warning system based on artificial intelligence. Background Art

[0002] With the rapid development of industrial automation, electric cylinder systems have been widely used in various mechanical equipment. However, the maintenance of traditional electric cylinder systems mainly relies on manual inspections and regular repairs, which is not only inefficient, but also difficult to detect potential faults in time, which may cause sudden shutdown or damage of equipment, causing unnecessary losses to production.

[0003] The existing electric cylinder system fault warning system has the following problems: first, it is difficult to identify and analyze the operation trajectory of the electric cylinder system based on the collected and processed image data, and it is difficult to perform initial fault warning processing based on the identification results; second, it is difficult to measure and analyze the operation variables of the electric cylinder system; then, it is difficult to comprehensively analyze the fault warning value of the electric cylinder based on various data; finally, it is difficult to dynamically set the threshold for real-time warning based on the fault warning value of the electric cylinder. Summary of the invention

[0004] The present invention aims to solve at least one of the technical problems existing in the prior art; to this end, the present invention proposes an electric cylinder system fault warning system based on artificial intelligence, which is used to solve the following technical problems:

[0005] First, it is difficult to identify and analyze the operation trajectory of the electric cylinder system based on the collected and processed image data, and it is difficult to perform initial fault warning processing based on the identification results; secondly, it is difficult to measure and analyze the operation variables of the electric cylinder system; then, it is difficult to comprehensively analyze the fault warning value of the electric cylinder based on various data; finally, it is difficult to dynamically set thresholds for real-time warning based on the fault warning value of the electric cylinder.

[0006] To solve the above problems, the first aspect of the present invention provides an electric cylinder system fault warning system based on artificial intelligence, comprising the following modules:

[0007] Data acquisition module: collects the operating data of the electric cylinder system, including: operating parameters and operating image data of the electric cylinder system;

[0008] Data processing module: cleaning and normalizing the collected operating parameters; preprocessing the video images of the electric cylinder system operation, including: denoising and image enhancement;

[0009] Initial fault warning module: Analyze the running trajectory of the electric cylinder according to the motion image data of the electric cylinder, identify and judge the normal trajectory and abnormal trajectory according to the generated running trajectory of the electric cylinder, and make initial fault warning;

[0010] Data analysis module: when the running track of the electric cylinder is in the normal running track, measure and analyze the running variables of the electric cylinder system; calculate and analyze the repeated displacement difference, electric cylinder stability and temperature rise influence coefficient of the electric cylinder based on the analysis results; calculate the contamination degree value based on the collected and processed running parameters; analyze the wear area of ​​the bearing in the electric cylinder system based on the collected and processed image data; calculate the wear area based on the analysis results;

[0011] Fault warning module: The fault warning value of the electric cylinder is calculated based on the repeated displacement difference of the electric cylinder, the stability of the electric cylinder, the temperature rise influence coefficient of the electric cylinder, the pollution degree value and the wear area; the threshold is dynamically adjusted according to the calculation result, and the fault warning is performed based on the dynamically adjusted threshold.

[0012] As a further solution of the present invention: according to the motion image data of the electric cylinder, the running trajectory of the electric cylinder is analyzed and the normal trajectory and the abnormal trajectory are identified and judged according to the generated running trajectory of the electric cylinder, and an initial fault warning is performed, including the following steps:

[0013] S1: Install a camera on the motion path of the electric cylinder to collect and process the image sequence of the electric cylinder in real time;

[0014] S2: Mark the end of the electric cylinder as a feature point, and identify the feature point of the electric cylinder position in the image by using an image processing algorithm; obtain the specific position of the electric cylinder according to the position of the feature point of the electric cylinder position in the image;

[0015] S3: The electric cylinder system fault warning system based on artificial intelligence monitors and collects the image sequence of the electric cylinder movement in real time to automatically detect the characteristic points of the electric cylinder position, and tracks the position changes of the characteristic points of the electric cylinder position in the image sequence; according to the position changes of the characteristic points of the electric cylinder position in the continuous images, the operation trajectory of the electric cylinder is generated;

[0016] S4: According to the generated running trajectory of the electric cylinder, the normal trajectory is compared with the abnormal trajectory, the running state of the electric cylinder is analyzed and judged, and an initial fault warning is performed.

[0017] As a further solution of the present invention: the step S4 comprises the following steps:

[0018] When the electric cylinder is operating normally, an image sequence of the electric cylinder's running trajectory is collected as a data sample of the normal trajectory; an image sequence of the electric cylinder's running trajectory under different fault conditions is collected as a data sample of the abnormal trajectory;

[0019] By establishing a neural network deep learning model, normal trajectories and abnormal trajectories are identified and judged; data samples of normal trajectories and data samples of abnormal trajectories are input into the model for model training;

[0020] The electric cylinder system fault warning system based on artificial intelligence monitors and collects the image sequence of the electric cylinder movement in real time, and identifies and determines whether the running trajectory of the electric cylinder is normal or abnormal.

[0021] When the running track of the electric cylinder is identified as an abnormal track, an initial fault warning signal is issued;

[0022] When it is identified that the running trajectory of the electric cylinder is a normal trajectory, the running variables of the electric cylinder are analyzed to determine whether to issue a fault warning.

[0023] As a further solution of the present invention: when the running track of the electric cylinder is in the normal running track, measuring and analyzing the running variables of the electric cylinder system includes the following steps:

[0024] Deploy edge computing devices on the electric cylinder system, and use the GPU on the edge computing node to perform image recognition and processing on the real-time collected operating image data of the electric cylinder to obtain the actual displacement of the electric cylinder, the number of bolts, and the bolt diameter.

[0025] When the electric cylinder reaches the set position or the speed is zero, the electric cylinder system completes a motion cycle, and the counter is incremented by one; the number of times the electric cylinder has run is obtained by accumulating the value of the counter;

[0026] By installing an angle sensor, the angle value between the electric cylinder and the horizontal plane output by the sensor is collected;

[0027] A length sensor is installed on the electric cylinder. Before the electric cylinder starts to move, the distance between the two ends of the electric cylinder is measured by the sensor and the initial length of the electric cylinder is recorded as a reference value. During the movement of the electric cylinder, the length sensor is used to monitor the length change of the electric cylinder in real time. When the movement of the electric cylinder ends, the final length of the electric cylinder is recorded. By comparing the difference between the final length and the initial length, the length change of the electric cylinder during the movement is obtained.

[0028] As a further solution of the present invention: calculating and analyzing the repeated displacement difference of the electric cylinder, the stability of the electric cylinder and the influence coefficient of the temperature rise of the electric cylinder according to the analysis results, including the following steps:

[0029] By analyzing the repeated displacement difference formula:

[0030] Get the repeated displacement difference ;in, represents the actual displacement of the electric cylinder, S represents the target displacement of the electric cylinder, Indicates the number of times the electric cylinder has run. ;

[0031] By analyzing the electric cylinder stability formula:

[0032] Get the stability of the electric cylinder ; Among them, θ represents the angle between the electric cylinder and the horizontal plane, N represents the number of bolts fixing the electric cylinder, represents the stretch coefficient, Indicates the diameter of the bolt, Indicates the preload force of the bolt;

[0033] By analyzing the temperature rise influence coefficient formula of the electric cylinder:

[0034] The temperature rise influence coefficient T of the electric cylinder is obtained, where: Indicates the temperature value of the electric cylinder at time t, Indicates the temperature value of the electric cylinder at time t+1, Indicates the ambient temperature of the electric cylinder. represents the final length at the end of the work, Indicates the initial length of the electric cylinder. .

[0035] As a further solution of the present invention: analyzing and calculating the pollution degree value according to the collected and processed operating parameters, the following steps are included:

[0036] The air around the electric cylinder system is passed through the filter membrane through the air sampler, so that the dust particles in the air are trapped by the filter membrane; the mass of the filter membrane is collected before and after the operation of the electric cylinder system, and the flow rate passing through the filter membrane and the operation time of the electric cylinder are collected;

[0037] By analyzing the ambient dust concentration formula:

[0038] Get the ambient dust concentration F, where: Indicates the quality of the filter membrane before the electric cylinder system is running. Indicates the quality of the filter membrane after the electric cylinder system is completed. represents the flow rate through the filter membrane, Indicates the running time of the electric cylinder;

[0039] By analyzing the pollution degree value formula:

[0040] Get the pollution degree value R; where, Indicates the ambient temperature of the electric cylinder. Indicates the ambient humidity of the electric cylinder. Indicates the ambient dust concentration of the electric cylinder; , and They correspond to the weights of ambient temperature, ambient humidity and dust concentration respectively.

[0041] As a further solution of the present invention: the wear area of ​​the bearing in the electric cylinder system is analyzed based on the collected and processed video image recognition, including the following steps:

[0042] Based on the collected and processed video images of the electric cylinder system, a deep learning model is established to identify and analyze the wear areas of the bearings in the electric cylinder system.

[0043] Marking the wear areas of bearings in electric axis systems;

[0044] Input the annotated wear area into the deep learning model for model training;

[0045] The artificial intelligence-based electric cylinder system fault warning system collects processed video images in real time to identify and output the wear areas of the bearings in the electric cylinder system.

[0046] As a further solution of the present invention: calculating the wear area of ​​the bearing according to the analysis results comprises the following steps:

[0047] By analyzing the formula:

[0048] The area of ​​the worn area M is obtained; where q represents the total number of pixels contained in the worn area, Represents the actual area of ​​the xth pixel, x .

[0049] As a further solution of the present invention: comprehensively calculating the fault warning value of the electric cylinder according to the repeated displacement difference of the electric cylinder, the stability of the electric cylinder, the temperature rise influence coefficient of the electric cylinder, the pollution degree value and the wear area, including the following steps:

[0050] By analyzing the fault warning value formula of the electric cylinder:

[0051]

[0052] Get the fault warning value of the electric cylinder ;in, , , T, R and M represent the repeated displacement difference of the electric cylinder, the stability of the electric cylinder, the influence coefficient of the temperature rise of the electric cylinder, the pollution degree value and the wear area respectively; , , , and They respectively represent the weights corresponding to the repeated displacement difference, the electric cylinder stability, the electric cylinder temperature rise influence coefficient, the pollution degree value and the wear area.

[0053] As a further solution of the present invention: dynamically adjusting the threshold value according to the calculation result, and performing fault warning according to the dynamically adjusted threshold value, the following steps are included:

[0054] By analyzing the fault threshold formula:

[0055] Get the fault threshold ,in represents the initial failure threshold, represents the adjustment factor, Indicates the fault warning value of the electric cylinder;

[0056] Set the fault warning value of the electric cylinder and fault threshold Compare and check if the fault warning value of the electric cylinder is Greater than or equal to the fault threshold When the alarm is raised, the electric cylinder will continue to run.

[0057] Compared with the prior art, the present invention has the following beneficial effects:

[0058] The present invention uses the motion image data of the electric cylinder to analyze the running trajectory of the electric cylinder and identify and judge the normal trajectory and abnormal trajectory, and perform initial fault warning; this helps to timely discover potential fault risks, take preventive measures in advance, avoid the occurrence of faults, and improve the efficiency of fault warning of the running state of the electric cylinder;

[0059] The present invention improves the accuracy of fault detection by comprehensively analyzing the fault warning value of the electric cylinder operation through various data;

[0060] The present invention provides real-time warning by dynamically setting thresholds, which helps to improve the accuracy and timeliness of fault warnings, reduce equipment operation and maintenance costs, and increase economic and safety benefits. BRIEF DESCRIPTION OF THE DRAWINGS

[0061] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0062] Figure 1 It is a system module diagram of the present invention;

[0063] Figure 2 The present invention is a flow chart of the method. DETAILED DESCRIPTION

[0064] The technical solution of the present invention will be clearly and completely described below in conjunction with the embodiments. Obviously, the described embodiments are only part of the embodiments of the present invention, rather than all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0065] See also Figure 1-2 The first aspect of the present invention provides an electric cylinder system fault warning system based on artificial intelligence, comprising the following modules:

[0066] Data acquisition module: collects the operating data of the electric cylinder system, including: operating parameters and operating image data of the electric cylinder system;

[0067] Data processing module: cleaning and normalizing the collected operating parameters; preprocessing the video images of the electric cylinder system operation, including: denoising and image enhancement;

[0068] Initial fault warning module: Analyze the running trajectory of the electric cylinder according to the motion image data of the electric cylinder, identify and judge the normal trajectory and abnormal trajectory according to the generated running trajectory of the electric cylinder, and make initial fault warning;

[0069] Data analysis module: when the running track of the electric cylinder is in the normal running track, measure and analyze the running variables of the electric cylinder system; calculate and analyze the repeated displacement difference, electric cylinder stability and temperature rise influence coefficient of the electric cylinder based on the analysis results; calculate the contamination degree value based on the collected and processed running parameters; analyze the wear area of ​​the bearing in the electric cylinder system based on the collected and processed image data; calculate the wear area based on the analysis results;

[0070] Fault warning module: The fault warning value of the electric cylinder is calculated based on the repeated displacement difference of the electric cylinder, the stability of the electric cylinder, the temperature rise influence coefficient of the electric cylinder, the pollution degree value and the wear area; the threshold is dynamically adjusted according to the calculation result, and the fault warning is performed based on the dynamically adjusted threshold.

[0071] Specifically, the present invention includes five modules, namely, a data acquisition module, a data processing module, an initial fault warning module, a data analysis module and a fault warning module. The data acquisition module is used to collect the operating parameters and operating video images of the electric cylinder system; the data processing module is used to clean and normalize the collected operating parameters to eliminate the influence of noise and abnormal values; the collected video image data is de-noised, and the contrast, brightness, smoothness and edge information of the video image are adjusted to enhance the video image, so that the edge details of the video image are clearer. According to the motion image data of the electric cylinder, the operating trajectory of the electric cylinder is analyzed and an initial fault warning is performed. By measuring the operating variables of the electric cylinder, the repeated displacement difference, the stability of the electric cylinder and the temperature rise influence coefficient of the electric cylinder are analyzed and calculated; the contamination degree value is analyzed and calculated according to the operating parameters; the wear area of ​​the bearing in the electric cylinder system is analyzed according to the image data; the fault warning value of the electric cylinder is calculated by combining various data, and the threshold is dynamically adjusted according to the calculation result to perform a fault warning.

[0072] In one embodiment of the present invention, according to the motion image data of the electric cylinder, the running trajectory of the electric cylinder is analyzed, and according to the generated running trajectory of the electric cylinder, the normal trajectory and the abnormal trajectory are identified and judged, and an initial fault warning is performed, including the following steps:

[0073] S1: Install a camera on the motion path of the electric cylinder to collect and process the image sequence of the electric cylinder in real time;

[0074] S2: Mark the end of the electric cylinder as a feature point, and identify the feature point of the electric cylinder position in the image by using an image processing algorithm; obtain the specific position of the electric cylinder according to the position of the feature point of the electric cylinder position in the image;

[0075] S3: The electric cylinder system fault warning system based on artificial intelligence monitors and collects the image sequence of the electric cylinder movement in real time to automatically detect the characteristic points of the electric cylinder position, and tracks the position changes of the characteristic points of the electric cylinder position in the image sequence; according to the position changes of the characteristic points of the electric cylinder position in the continuous images, the operation trajectory of the electric cylinder is generated;

[0076] S4: According to the generated running trajectory of the electric cylinder, the normal trajectory is compared with the abnormal trajectory, the running state of the electric cylinder is analyzed and judged, and an initial fault warning is performed.

[0077] Specifically, a camera is installed on the motion path of the electric cylinder, and the camera is used to capture continuous images of the electric cylinder during movement. An obvious mark is set at the end of the electric cylinder as a feature point. Then, the position information of this feature point is extracted from the image through the edge detection algorithm in image processing technology. The specific position of the electric cylinder can be calculated by analyzing the coordinates of the feature point in the image. Machine learning technology is used to automatically detect and track the position changes of the feature points of the electric cylinder, thereby generating the running trajectory of the electric cylinder. A model is established based on the normal motion trajectory of the electric cylinder. When the actual measured trajectory is significantly different from the normal trajectory, an initial fault warning is issued. This helps to detect potential problems in advance and avoid more serious failures.

[0078] In one embodiment of the present invention, the step S4 comprises the following steps:

[0079] When the electric cylinder is operating normally, an image sequence of the electric cylinder's running trajectory is collected as a data sample of the normal trajectory; an image sequence of the electric cylinder's running trajectory under different fault conditions is collected as a data sample of the abnormal trajectory;

[0080] By establishing a neural network deep learning model, normal trajectories and abnormal trajectories are identified and judged; data samples of normal trajectories and data samples of abnormal trajectories are input into the model for model training;

[0081] The electric cylinder system fault warning system based on artificial intelligence monitors and collects the image sequence of the electric cylinder movement in real time, and identifies and determines whether the running trajectory of the electric cylinder is normal or abnormal.

[0082] When the running track of the electric cylinder is identified as an abnormal track, an initial fault warning signal is issued;

[0083] When it is identified that the running trajectory of the electric cylinder is a normal trajectory, the running variables of the electric cylinder are analyzed to determine whether to issue a fault warning.

[0084] Specifically, when the electric cylinder is operating normally, image sequences of the running trajectory are collected, and these image sequences will be used as data samples of the normal trajectory. At the same time, the running trajectory image sequences of the electric cylinder under different fault states are simulated or collected as data samples of the abnormal trajectory. Ensure that the collected data samples are representative and can cover various normal operating states and fault states that may occur in the electric cylinder. Establish a neural network deep learning model to identify and judge whether the running trajectory of the electric cylinder is normal or abnormal. Input the data samples of the normal trajectory and the data samples of the abnormal trajectory into the model for training. By adjusting the model parameters, the model can accurately distinguish between the normal trajectory and the abnormal trajectory. The electric cylinder system fault warning system based on artificial intelligence monitors and collects the image sequence of the electric cylinder movement in real time. The trained neural network model is used to identify and judge the collected image sequence to determine whether the running trajectory of the electric cylinder is a normal trajectory or an abnormal trajectory. When the running trajectory of the electric cylinder is identified as an abnormal trajectory, an initial fault warning signal is issued to remind the operator to check and maintain. When the running trajectory of the electric cylinder is identified as a normal trajectory, the running variables of the electric cylinder will continue to be analyzed to determine whether further fault warning is needed.

[0085] In one embodiment of the present invention, when the running trajectory of the electric cylinder is in the normal running trajectory, measuring and analyzing the running variables of the electric cylinder system includes the following steps:

[0086] Deploy edge computing devices on the electric cylinder system, and use the GPU on the edge computing node to perform image recognition and processing on the real-time collected operating image data of the electric cylinder to obtain the actual displacement of the electric cylinder, the number of bolts, and the bolt diameter.

[0087] When the electric cylinder reaches the set position or the speed is zero, the electric cylinder system completes a motion cycle, and the counter is incremented by one; the number of times the electric cylinder has run is obtained by accumulating the value of the counter;

[0088] By installing an angle sensor, the angle value between the electric cylinder and the horizontal plane output by the sensor is collected;

[0089] A length sensor is installed on the electric cylinder. Before the electric cylinder starts to move, the distance between the two ends of the electric cylinder is measured by the sensor and the initial length of the electric cylinder is recorded as a reference value. During the movement of the electric cylinder, the length sensor is used to monitor the length change of the electric cylinder in real time. When the movement of the electric cylinder ends, the final length of the electric cylinder is recorded. By comparing the difference between the final length and the initial length, the length change of the electric cylinder during the movement is obtained.

[0090] Specifically, an edge computing device is installed on the electric cylinder system. The device has powerful data processing capabilities and can process a large amount of image data in real time. The edge computing device is connected to the control unit of the electric cylinder system to ensure that data can be transmitted and processed in real time. By collecting the running image data of the electric cylinder in real time, the GPU on the edge computing node is used for image recognition and processing to obtain the actual displacement, number of bolts and bolt diameter of the electric cylinder. When the electric cylinder reaches the set position or the speed is zero, it is considered that the electric cylinder system has completed a motion cycle. At this time, the counter is incremented by one to record the number of times the electric cylinder has run. By accumulating the value of the counter, the frequency of use and wear of the electric cylinder can be understood. An angle sensor is installed on the electric cylinder to collect the angle value between the electric cylinder and the horizontal plane output by the sensor. This angle value can help determine the movement direction and inclination of the electric cylinder, thereby evaluating its stability and safety. A length sensor is installed on the electric cylinder to measure the distance between the two ends of the electric cylinder. Before the electric cylinder starts to move, the sensor measures and records the initial length of the electric cylinder as a reference value; during the movement of the electric cylinder, the change in the length of the electric cylinder is monitored in real time; when the movement of the electric cylinder ends, the final length of the electric cylinder is recorded. By comparing the difference between the final length and the initial length, the length change of the electric cylinder during the movement can be obtained.

[0091] In one embodiment of the present invention, the repeated displacement difference of the electric cylinder, the stability of the electric cylinder and the influence coefficient of the temperature rise of the electric cylinder are calculated and analyzed according to the analysis results, including the following steps:

[0092] By analyzing the repeated displacement difference formula:

[0093] Get the repeated displacement difference ;in, represents the actual displacement of the electric cylinder, S represents the target displacement of the electric cylinder, Indicates the number of times the electric cylinder has run. ;

[0094] By analyzing the electric cylinder stability formula:

[0095] Get the stability of the electric cylinder ; Among them, θ represents the angle between the electric cylinder and the horizontal plane, N represents the number of bolts fixing the electric cylinder, represents the stretch coefficient, Indicates the diameter of the bolt, Indicates the preload force of the bolt;

[0096] By analyzing the temperature rise influence coefficient formula of the electric cylinder:

[0097] The temperature rise influence coefficient T of the electric cylinder is obtained, where: Indicates the temperature value of the electric cylinder at time t, Indicates the temperature value of the electric cylinder at time t+1, Indicates the ambient temperature of the electric cylinder. represents the final length at the end of the work, Indicates the initial length of the electric cylinder. .

[0098] Specifically, the difference in repeated displacement of the electric cylinder is calculated by the actual displacement, target displacement and number of operations of the electric cylinder during operation, and whether the mechanical structure of the electric cylinder is loose is determined. The angle between the electric cylinder and the horizontal plane will affect its force. Different angles will result in different gravity components and support reaction forces, thus affecting stability. The more bolts that fix the electric cylinder, the better stability it can usually provide. More bolts mean a larger contact area and more uniform stress distribution. The tensile coefficient reflects the performance of the bolts in a tensile state. When the stability of the electric cylinder is poor, it may cause the electric cylinder to move unsteadily or be damaged. The temperature value of the electric cylinder at time t and the temperature value of the electric cylinder at time t+1 are collected. These two parameters represent the temperature values ​​of the electric cylinder at two consecutive time points. By comparing the changes in these two temperature values, the temperature rise of the electric cylinder over a period of time can be understood. If the temperature value of the electric cylinder at time t+1 is higher than the temperature value of the electric cylinder at time t, it indicates that the electric cylinder has experienced a large temperature rise during this period. A higher ambient temperature may cause the temperature of the electric cylinder to rise, while a lower ambient temperature may help dissipate heat. The length of the cylinder at the end of the work reflects its elongation or contraction during the work. The change in the length of the cylinder is closely related to its internal temperature, because the thermal expansion and contraction effect will cause the material size to change. By measuring the final length, it is possible to infer the temperature change of the cylinder during the work. The initial length of the cylinder is its original length when it is not working. Comparing the final length with the initial length can determine the actual deformation of the cylinder during the work.

[0099] In one embodiment of the present invention, the pollution degree value is calculated based on the collected and processed operating parameters, including the following steps:

[0100] The air around the electric cylinder system is passed through the filter membrane through the air sampler, so that the dust particles in the air are trapped by the filter membrane; the mass of the filter membrane is collected before and after the operation of the electric cylinder system, and the flow rate passing through the filter membrane and the operation time of the electric cylinder are collected;

[0101] By analyzing the ambient dust concentration formula:

[0102] Get the ambient dust concentration F, where: Indicates the quality of the filter membrane before the electric cylinder system is running. Indicates the quality of the filter membrane after the electric cylinder system is completed. represents the flow rate through the filter membrane, Indicates the running time of the electric cylinder;

[0103] By analyzing the pollution degree value formula:

[0104] Get the pollution degree value R; where, Indicates the ambient temperature of the electric cylinder. Indicates the ambient humidity of the electric cylinder. Indicates the ambient dust concentration of the electric cylinder; , and They correspond to the weights of ambient temperature, ambient humidity and dust concentration respectively.

[0105] Specifically, the filter membrane in the air sampler is used to intercept dust particles in the air. Install the filter membrane on the sampling head of the air sampler to ensure that it is firmly installed and well sealed. Before the electric cylinder system is running, start the air sampler to allow the air around the electric cylinder system to pass through the filter membrane; after the electric cylinder system is finished running, start the air sampler again to collect air samples of the same length of time. The mass of the filter membrane before and after the operation of the electric cylinder system is measured to obtain the formula for environmental dust concentration. Dust particles may enter the interior of the electric cylinder, causing increased wear on key parts such as slide rails, transmission components, and electrical connection points, and may even clog pores or damage seals. Excessive humidity may cause metal parts inside the servo electric cylinder to rust, while too low humidity may make plastic parts brittle. Excessive temperature may cause the lubricating oil to deteriorate, increase friction and wear; while too low temperature may make the lubricating oil become viscous, affecting the normal operation of the electric cylinder. The pollution degree value is calculated comprehensively based on the ambient temperature, ambient humidity, and ambient dust concentration, where, , and They are 0.3, 0.25 and 0.45 respectively.

[0106] In one embodiment of the present invention, the wear area of ​​the bearing in the electric cylinder system is analyzed based on the collected and processed video image recognition, including the following steps:

[0107] Based on the collected and processed video images of the electric cylinder system, a deep learning model is established to identify and analyze the wear areas of the bearings in the electric cylinder system.

[0108] Marking the wear areas of bearings in electric axis systems;

[0109] Input the annotated wear area into the deep learning model for model training;

[0110] The artificial intelligence-based electric cylinder system fault warning system collects processed video images in real time to identify and output the wear areas of the bearings in the electric cylinder system.

[0111] Specifically, the wear area of ​​the bearing is manually marked in the video image, and the wear area in each image is accurately marked to ensure the accuracy and consistency of the marking, so as to facilitate subsequent model training and evaluation. A deep learning model is designed and constructed to learn the features in the image and accurately predict the location and shape of the wear area. The marked wear area is input into the deep learning model for training. Through multiple iterations of training, the model gradually learns the features in the image and the rules of the wear area. During the training process, the validation set is regularly used to evaluate the performance of the model, and the model parameters are adjusted according to the evaluation results. The artificial intelligence-based electric cylinder system fault warning system is integrated with the deep learning model. The running video images of the electric cylinder system are collected in real time through the fault warning system, and the processed images are input into the deep learning model for identification and analysis. According to the output results of the model, the wear area of ​​the bearing in the electric cylinder system is identified and output.

[0112] In one embodiment of the present invention, calculating the wear area of ​​the bearing according to the analysis results includes the following steps:

[0113] By analyzing the wear area formula:

[0114] The area of ​​the worn area M is obtained; where q represents the total number of pixels contained in the worn area, Represents the actual area of ​​the xth pixel, x .

[0115] Specifically, we first need to identify all pixels in the image that belong to the worn area and count the total number of these pixels. The actual area of ​​each pixel may vary depending on factors such as image resolution and unit conversion. The total area of ​​the worn area can be obtained by adding up the actual areas of all pixels in the worn area. By calculating the total area of ​​the worn area, we can quantify the degree of wear on the surface of the electric cylinder and evaluate its usage or remaining life.

[0116] In one embodiment of the present invention, the fault warning value of the electric cylinder is comprehensively calculated according to the repeated displacement difference of the electric cylinder, the stability of the electric cylinder, the temperature rise influence coefficient of the electric cylinder, the pollution degree value and the wear area, including the following steps:

[0117] By analyzing the fault warning value formula of the electric cylinder:

[0118]

[0119] Get the fault warning value of the electric cylinder ;in, , , T, R and M represent the repeated displacement difference of the electric cylinder, the stability of the electric cylinder, the influence coefficient of the temperature rise of the electric cylinder, the pollution degree value and the wear area respectively; , , , and They respectively represent the weights corresponding to the repeated displacement difference, the electric cylinder stability, the electric cylinder temperature rise influence coefficient, the pollution degree value and the wear area.

[0120] Specifically, the repeated displacement difference reflects the displacement change of the electric cylinder during multiple cycles. A large displacement difference may mean that the electric cylinder has mechanical wear or inaccurate control problems. The stability of the electric cylinder is one of the key factors to ensure its normal operation. An unstable electric cylinder may cause vibration, increased noise, or even complete failure. The temperature rise coefficient of the electric cylinder has a significant impact on the performance of the electric cylinder. Excessive temperature will lead to a decrease in material properties and poor lubrication effect. The degree of pollution is obtained based on the temperature, humidity and dust concentration around the electric cylinder. By analyzing the area of ​​the electric cylinder wear zone, the degree of wear of the electric cylinder equipment can be accurately evaluated. Among them, the weight of the repeated displacement difference is 0.15, the weight of the electric cylinder stability The weight of the influence coefficient of the temperature rise of the electric cylinder is 0.25. The weight of the pollution degree value is 0.25. The weight of the wear area is 0.25 and is 0.15; , , , and The sum of these weights is 1. These weights reflect the importance of each parameter in evaluating the electric cylinder failure. Different application scenarios may require different weight assignments to more accurately reflect the actual situation. The fault value of the electric cylinder operation is calculated by comprehensively considering multiple key performance indicators.

[0121] In one embodiment of the present invention, the threshold is dynamically adjusted according to the calculation result, and a fault warning is performed according to the dynamically adjusted threshold, including the following steps:

[0122] By analyzing the fault threshold formula:

[0123] Get the fault threshold ,in represents the initial failure threshold, represents the adjustment factor, Indicates the fault warning value of the electric cylinder;

[0124] Set the fault warning value of the electric cylinder and fault threshold Compare and check if the fault warning value of the electric cylinder is Greater than or equal to the fault threshold When the alarm is raised, the electric cylinder will continue to run.

[0125] Specifically, the fault threshold is a dynamically adjusted value, which is determined based on the initial basic fault threshold and the adjustment coefficient and the current fault warning value of the electric cylinder. The initial fault threshold depends on the actual operation and working environment of the electric cylinder system and is determined based on historical data, industry standards or expert opinions to ensure its accuracy and reliability. The adjustment coefficient is calculated according to the operating status and environmental conditions of the electric cylinder system. The adjustment coefficient can reflect the impact of factors such as the health status, frequency of use, ambient temperature, humidity, etc. of the electric cylinder system on the fault threshold. By comprehensively considering these factors, the failure risk of the electric cylinder system can be more accurately assessed. The fault warning value of the electric cylinder is compared with the calculated fault threshold. If the fault warning value is greater than or equal to the fault threshold, it means that the electric cylinder system may have a risk of failure, and a warning signal needs to be issued and corresponding measures need to be taken; otherwise, the electric cylinder system will continue to operate normally.

[0126] The above embodiments are only used to illustrate the technical method of the present invention rather than to limit it. Although the present invention has been described in detail with reference to the preferred embodiments, those skilled in the art should understand that the technical method of the present invention may be modified or replaced by equivalents without departing from the spirit and scope of the technical method of the present invention.

Claims

1. An electric cylinder system fault warning system based on artificial intelligence, characterized in that: Includes the following modules: Data acquisition module: collects the operating data of the electric cylinder system, including: operating parameters and operating image data of the electric cylinder system; Data processing module: cleaning and normalizing the collected operating parameters; preprocessing the video images of the electric cylinder system operation, including: denoising and image enhancement; Initial fault warning module: Analyze the running trajectory of the electric cylinder according to the motion image data of the electric cylinder, identify and judge the normal trajectory and abnormal trajectory according to the generated running trajectory of the electric cylinder, and make initial fault warning; Data analysis module: when the running track of the electric cylinder is in the normal running track, measure and analyze the running variables of the electric cylinder system; calculate and analyze the repeated displacement difference, electric cylinder stability and temperature rise influence coefficient of the electric cylinder based on the analysis results; calculate the contamination degree value based on the collected and processed running parameters; analyze the wear area of ​​the bearing in the electric cylinder system based on the collected and processed image data; calculate the wear area based on the analysis results; Fault warning module: comprehensively calculate the fault warning value of the electric cylinder based on the repeated displacement difference of the electric cylinder, the stability of the electric cylinder, the temperature rise influence coefficient of the electric cylinder, the pollution degree value and the wear area; dynamically adjust the threshold value according to the calculation result, and perform fault warning based on the dynamically adjusted threshold value; The calculation and analysis of the repeated displacement difference of the electric cylinder, the stability of the electric cylinder and the temperature rise influence coefficient of the electric cylinder according to the analysis results includes the following steps: By analyzing the repeated displacement difference formula: Get the repeated displacement difference ;in, represents the actual displacement of the electric cylinder, S represents the target displacement of the electric cylinder, Indicates the number of times the electric cylinder has run. ; By analyzing the electric cylinder stability formula: Get the stability of the electric cylinder ; where θ represents the angle between the electric cylinder and the horizontal plane, and N represents the number of bolts fixing the electric cylinder. represents the stretch coefficient, Indicates the diameter of the bolt, Indicates the preload force of the bolt; By analyzing the temperature rise influence coefficient formula of the electric cylinder: The temperature rise influence coefficient T of the electric cylinder is obtained, where: Indicates the temperature value of the electric cylinder at time t, Indicates the temperature value of the electric cylinder at time t+1, Indicates the ambient temperature of the electric cylinder. represents the final length at the end of the work, Indicates the initial length of the electric cylinder. ; The calculation of the pollution degree value according to the collected and processed operating parameters includes the following steps: The air around the electric cylinder system is passed through the filter membrane through the air sampler, so that the dust particles in the air are trapped by the filter membrane; the mass of the filter membrane is collected before and after the operation of the electric cylinder system, and the flow rate passing through the filter membrane and the operation time of the electric cylinder are collected; By analyzing the ambient dust concentration formula: Get the ambient dust concentration F, where: Indicates the quality of the filter membrane before the electric cylinder system is running. Indicates the quality of the filter membrane after the electric cylinder system is completed. represents the flow rate through the filter membrane, Indicates the running time of the electric cylinder; By analyzing the pollution degree value formula: Get the pollution degree value R; where, Indicates the ambient temperature of the electric cylinder. Indicates the ambient humidity of the electric cylinder. Indicates the ambient dust concentration of the electric cylinder; , and They correspond to the weights of ambient temperature, ambient humidity and dust concentration respectively.

2. According to the artificial intelligence-based electric cylinder system fault warning system of claim 1, it is characterized in that: According to the motion image data of the electric cylinder, the running trajectory of the electric cylinder is analyzed, and according to the generated running trajectory of the electric cylinder, the normal trajectory and the abnormal trajectory are identified and judged, and the initial fault warning is performed, including the following steps: S1: Install a camera on the motion path of the electric cylinder to collect and process the image sequence of the electric cylinder in real time; S2: Mark the end of the electric cylinder as a feature point, and identify the feature point of the electric cylinder position in the image by using an image processing algorithm; obtain the specific position of the electric cylinder according to the position of the feature point of the electric cylinder position in the image; S3: The electric cylinder system fault warning system based on artificial intelligence monitors and collects the image sequence of the electric cylinder movement in real time to automatically detect the characteristic points of the electric cylinder position, and tracks the position changes of the characteristic points of the electric cylinder position in the image sequence; according to the position changes of the characteristic points of the electric cylinder position in the continuous images, the operation trajectory of the electric cylinder is generated; S4: According to the generated running trajectory of the electric cylinder, the normal trajectory is compared with the abnormal trajectory, the running state of the electric cylinder is analyzed and judged, and an initial fault warning is performed.

3. The electric cylinder system fault warning system based on artificial intelligence according to claim 2 is characterized in that: The step S4 comprises the following steps: When the electric cylinder is operating normally, an image sequence of the electric cylinder's running trajectory is collected as a data sample of the normal trajectory; an image sequence of the electric cylinder's running trajectory under different fault conditions is collected as a data sample of the abnormal trajectory; By establishing a neural network deep learning model, normal trajectories and abnormal trajectories are identified and judged; data samples of normal trajectories and data samples of abnormal trajectories are input into the model for model training; The electric cylinder system fault warning system based on artificial intelligence monitors and collects the image sequence of the electric cylinder movement in real time, and identifies and determines whether the running trajectory of the electric cylinder is normal or abnormal. When the running track of the electric cylinder is identified as an abnormal track, an initial fault warning signal is issued; When it is identified that the running trajectory of the electric cylinder is a normal trajectory, the running variables of the electric cylinder are analyzed to determine whether to issue a fault warning.

4. The electric cylinder system fault warning system based on artificial intelligence according to claim 1 is characterized in that: Measuring and analyzing the electric cylinder operation variables based on the collected and processed video images of the electric cylinder system operation includes the following steps: Deploy edge computing devices on the electric cylinder system, and use the GPU on the edge computing node to perform image recognition and processing on the real-time collected operating image data of the electric cylinder to obtain the actual displacement of the electric cylinder, the number of bolts, and the bolt diameter. When the electric cylinder reaches the set position or the speed is zero, the electric cylinder system completes a motion cycle, and the counter is incremented by one; the number of times the electric cylinder has run is obtained by accumulating the value of the counter; By installing an angle sensor, the angle value between the electric cylinder and the horizontal plane output by the sensor is collected; A length sensor is installed on the electric cylinder. Before the electric cylinder starts to move, the distance between the two ends of the electric cylinder is measured by the sensor and the initial length of the electric cylinder is recorded as a reference value. During the movement of the electric cylinder, the length sensor is used to monitor the length change of the electric cylinder in real time. When the movement of the electric cylinder ends, the final length of the electric cylinder is recorded. By comparing the difference between the final length and the initial length, the length change of the electric cylinder during the movement is obtained.

5. The electric cylinder system fault warning system based on artificial intelligence according to claim 1 is characterized in that: The wear area of ​​the bearing in the electric cylinder system is analyzed based on the collected and processed video images, including the following steps: Based on the collected and processed video images of the electric cylinder system, a deep learning model is established to identify and analyze the wear areas of the bearings in the electric cylinder system. Marking the wear areas of bearings in electric axis systems; Input the annotated wear area into the deep learning model for model training; The artificial intelligence-based electric cylinder system fault warning system collects processed video images in real time to identify and output the wear areas of the bearings in the electric cylinder system.

6. The electric cylinder system fault warning system based on artificial intelligence according to claim 5 is characterized in that: The wear area of ​​the bearing is calculated based on the analysis results, including the following steps: By analyzing the wear area formula: The area of ​​the worn area M is obtained; where q represents the total number of pixels contained in the worn area, Represents the actual area of ​​the xth pixel, x .

7. The electric cylinder system fault warning system based on artificial intelligence according to claim 1 is characterized in that: The fault warning value of the electric cylinder is calculated comprehensively according to the repeated displacement difference of the electric cylinder, the stability of the electric cylinder, the influence coefficient of the temperature rise of the electric cylinder, the pollution degree value and the wear area, including the following steps: By analyzing the fault warning value formula of the electric cylinder: Get the fault warning value of the electric cylinder ;in, , , T, R and M represent the repeated displacement difference of the electric cylinder, the stability of the electric cylinder, the influence coefficient of the temperature rise of the electric cylinder, the pollution degree value and the wear area respectively; , , , and They respectively represent the weights corresponding to the repeated displacement difference, the electric cylinder stability, the electric cylinder temperature rise influence coefficient, the pollution degree value and the wear area.

8. The electric cylinder system fault warning system based on artificial intelligence according to claim 7 is characterized in that: Dynamically adjusting the threshold value according to the calculation result, and performing fault warning according to the dynamically adjusted threshold value, including the following steps: By analyzing the fault threshold formula: Get the fault threshold ,in represents the initial failure threshold, represents the adjustment factor, Indicates the fault warning value of the electric cylinder; Set the fault warning value of the electric cylinder and fault threshold Compare and check if the fault warning value of the electric cylinder is Greater than or equal to the fault threshold When the alarm is raised, the electric cylinder will continue to run.

Images