Motor state detection method, device, computer equipment and readable storage medium

By collecting and analyzing the robot start and stop information, status information and historical fault information, and using machine learning methods to identify abnormal motors, the problem of low fault detection accuracy in the existing technology is solved, and the timely maintenance and fault prediction of the motor are achieved, and the Scrap happens.

CN118425774BActive Publication Date: 2024-11-01KAIWEI IND SERVICES (SHANGHAI) CO LTD
View PDF 4 Cites 0 Cited by

Patent Information

Application Number
CN202410896503.5
Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
Filing Date
2024-07-05
Publication Date
2024-11-01
Estimated Expiration
2044-07-05

AI Technical Summary

Technical Problem

The existing motor fault detection methods have low accuracy, which leads to the inability to maintain and repair the motor in time, and then scrapped.

Method used

By collecting the robot's start and stop information, status information and historical fault information, using machine learning methods such as neural networks, clustering models and support vector machines, we can identify abnormal motors, avoid large-scale boot detection, and improve detection accuracy.

Benefits of technology

It improves the accuracy and timeliness of motor fault detection, avoids scrapping due to inability to maintain in time, and ensures the normal operation of the robot system.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN118425774B_ABST
    Figure CN118425774B_ABST
Patent Text Reader

Abstract

The present application discloses a motor state detection method, device, computer equipment and readable storage medium, including: collecting the start and stop information of the robot; if it is determined that the first robot is in the running state, collecting the state information of the motor of the first robot; if it is determined that the state information of the first motor and the similarity between the historical fault information of the first motor exceed the state similarity threshold, then determining that the first motor is an abnormal motor; if it is determined that the second robot is in the shutdown state, collecting the production information of the second robot; if it is determined that the production information of the second robot is similar to the historical fault information of the second robot, then determining that the second robot is a robot to be identified; starting the robot to be identified and collecting the test information of the motor; if it is determined that the test information of the second motor is similar to the historical fault information of the second motor, then determining that the second motor is an abnormal motor. The present application not only improves the detection accuracy, but also avoids the situation where the motor is scrapped due to serious faults.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present application relates to the technical field of motor testing, and in particular to a motor state detection method, device, computer equipment and readable storage medium. Background Art

[0002] A robot is an automated, programmable mechanical device that can achieve various industrial processing and manufacturing functions by relying on its own power and control capabilities. The motor in the robot is the core of the robot's operation. Therefore, the reliability of the motor's fault detection method plays a key role in whether the robot can operate normally.

[0003] Traditional motor fault detection methods usually use methods such as observing whether the motor is damaged, whether the wiring is disconnected or loose, whether the insulation is burned, etc.; or judging whether the temperature or voltage of the robot's motor exceeds a threshold to achieve motor fault detection.

[0004] However, this method not only leads to low fault detection accuracy of the robot, but also when a parameter of the motor exceeds the threshold, it indicates that a serious fault has occurred in the motor, causing the motor to be easily scrapped due to lack of timely maintenance and repair. Summary of the invention

[0005] The purpose of the present invention is to provide a motor state detection method, device, computer equipment and readable storage medium to solve the problems of low fault detection accuracy in the prior art and the problem that the motor is easily scrapped due to failure to obtain timely maintenance and repair.

[0006] To achieve the above object, the present invention provides a motor state detection method for detecting at least one motor in at least one robot, the method comprising:

[0007] Collecting start and stop information of at least one of the robots; wherein the start and stop information includes a running state and a shutdown state;

[0008] If it is determined that the first robot is in the running state, then collecting state information of at least one motor of the first robot; wherein the first robot is one of at least one of the robots; the state information includes at least one voltage signal representing the running parameter of the robot, and at least one voltage signal representing the running condition of the robot;

[0009] If it is determined that the similarity between the state information of the first motor and the historical fault information of the first motor exceeds a preset state similarity threshold, the first motor is determined to be an abnormal motor; wherein the first motor is one of at least one of the motors; the historical fault information of the first motor is state information of a motor of the same motor type as the first motor when a fault occurred in the past;

[0010] If it is determined that the second robot is in a shutdown state, then collecting production information of the second robot; wherein the second robot is one of at least one of the robots; and the production information includes at least one piece of data information representing the operation status of the robot within a preset production time period;

[0011] If it is determined that the production information of the second robot is similar to the historical fault information of the second robot, the second robot is determined to be the robot to be identified; wherein the historical fault information of the second robot is production information of a robot of the same device type as the second robot when a fault occurred in the past;

[0012] Starting the robot to be identified and collecting test information of at least one motor in the robot to be identified; wherein the test information includes at least one voltage signal representing a test parameter of the robot within a test time period, and at least one voltage signal representing a test condition of the robot within the test time period;

[0013] If it is determined that the test information of the second motor is similar to the historical fault information of the second motor, the second motor is determined to be an abnormal motor; wherein, the second motor is a motor in the robot to be identified; the historical fault information of the second motor is status information of a motor of the same motor type as the second motor when a fault occurred in history.

[0014] In the above solution, collecting the start and stop information of at least one of the robots includes:

[0015] Accessing a cloud database, wherein the cloud database has at least one list, wherein the list represents at least one robot in a factory area, and wherein the list records a unique identifier of at least one robot;

[0016] Obtaining a unique identifier of each robot in a first list from the database, and sending an inquiry message to each of the device addresses; wherein the first list is one of at least one of the lists;

[0017] Receive the start and stop information sent by each of the robots according to the inquiry message.

[0018] In the above solution, collecting the status information of at least one motor of the first robot includes:

[0019] Acquire the operation collection rules and working condition collection rules of the operation collection motor; wherein the operation collection motor is one of at least one motor of the first robot; the operation collection rules record the instrument used to collect the operation parameters of the motor and the calling rules of the instrument; the working condition collection rules record the instrument used to collect the operation condition of the motor and the calling rules of the instrument;

[0020] Collecting the operating parameters of the operating acquisition motor according to the operating acquisition rule; wherein the operating parameters include: temperature signal, torque signal, voltage value signal, current value signal, resistance value signal and electromagnetic signal;

[0021] The operating condition of the operating motor is collected according to the operating condition collection rule; wherein the operating condition includes: a sound amplitude signal and a mechanical vibration signal.

[0022] In the above solution, if it is determined that the similarity between the state information of the first motor and the historical fault information of the first motor exceeds a preset state similarity threshold, then determining that the first motor is an abnormal motor includes:

[0023] Calling a preset mature neural network model to calculate the state information of the first motor to obtain a network similarity result; if it is determined that the network similarity result exceeds the network threshold of the state similarity threshold, determining that the first motor is an abnormal motor; wherein the mature neural network model is obtained by training an initial neural network model through at least one historical fault information of the first motor; or

[0024] Call a preset mature clustering model to perform clustering operation on the state information of the first motor to obtain the similarity distance between the state information of the first motor and the cluster center of each cluster in the mature clustering model; if it is determined that at least one of the similarity distances exceeds the clustering threshold in the state similarity threshold, the first motor is determined to be an abnormal motor; wherein the mature clustering model is a model with at least one cluster cluster formed by entering at least one historical fault information of the first motor into an initial clustering model, and calling the initial clustering model to cluster the entered historical fault information; the cluster center is the center point of the cluster cluster.

[0025] In the above solution, collecting the production information of the second robot includes:

[0026] accessing a log database of the second robot;

[0027] Acquire production information of the second robot within a preset production time period; wherein at least one piece of data information in the production information includes: operating time, operating multiple, operating rate, and equipment utilization rate.

[0028] In the above solution, if it is determined that the production information of the second robot is similar to the historical fault information of the second robot, determining that the second robot is a robot to be identified includes:

[0029] Calling a preset mature support vector machine to operate on the production information of the second robot to obtain the production category of the second robot; wherein the mature support vector machine is formed by training an initial support vector machine pair through historical normal information of at least one second robot and historical fault information of at least one second robot; the historical normal information of the second robot is production information of a robot of the same equipment type as the second robot when operating normally in history;

[0030] If it is determined that the content of the production category is a risk category, the second robot is determined to be a robot to be identified; wherein the risk category is a category in which the historical fault information of the second robot in the mature support vector machine belongs.

[0031] In the above solution, if it is determined that the test information of the second motor is similar to the historical fault information of the second motor, then determining that the second motor is an abnormal motor includes:

[0032] Calling a preset mature random forest model to operate on the test information of the second motor to obtain at least one classification category of the test information of the second motor; wherein the mature random forest model is composed of at least one mature decision tree; the mature decision tree is a binary tree model with at least one decision condition generated by dividing the historical normal information of at least one second motor and the historical abnormal information of at least one second motor through an initial decision tree; the mature decision tree is used to divide the voltage signal corresponding to the decision condition in the test information according to at least one decision condition, and obtain a classification category of the voltage signal corresponding to the decision condition;

[0033] If it is determined that the content of at least one of the classification categories is a fault category, the second motor is determined to be an abnormal motor; the fault category is a category to which the historical fault information of the second motor in the mature random forest belongs.

[0034] To achieve the above object, the present invention further provides a motor state detection device for detecting at least one motor in at least one robot, the device comprising:

[0035] A start-stop acquisition module, used to collect start-stop information of at least one of the robots; wherein the start-stop information includes a running state and a shutdown state;

[0036] An operation acquisition module, configured to acquire state information of at least one motor of the first robot if it is determined that the first robot is in an operation state; wherein the first robot is one of at least one of the robots; and the state information includes at least one voltage signal representing an operation parameter of the robot, and at least one voltage signal representing an operation condition of the robot;

[0037] An operation determination module is used to determine that the first motor is an abnormal motor if it is determined that the similarity between the state information of the first motor and the historical fault information of the first motor exceeds a preset state similarity threshold; wherein the first motor is one of the at least one motor; and the historical fault information of the first motor is state information of a motor of the same motor type as the first motor when a fault occurs in history;

[0038] A shutdown collection module, configured to collect production information of the second robot if it is determined that the second robot is in a shutdown state; wherein the second robot is one of the at least one robot; and the production information includes at least one piece of data information representing the operation status of the robot within a preset production time period;

[0039] a robot determination module, configured to determine that the second robot is a to-be-identified robot if it is determined that the production information of the second robot is similar to the historical fault information of the second robot; wherein the historical fault information of the second robot is production information of a robot of the same device type as the second robot when a fault occurred in the past;

[0040] A test module, used to start the robot to be identified and collect test information of at least one motor in the robot to be identified; wherein the test information includes at least one voltage signal representing a test parameter of the robot within a test time period, and at least one voltage signal representing a test condition of the robot within the test time period;

[0041] A shutdown determination module is used to determine that the second motor is an abnormal motor if it is determined that the test information of the second motor is similar to the historical fault information of the second motor; wherein the second motor is a motor in the robot to be identified; and the historical fault information of the second motor is status information of a motor of the same motor type as the second motor when a fault occurred in history.

[0042] To achieve the above objectives, the present invention also provides a computer device, which includes a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor of the computer device implements the steps of the above motor state detection method when executing the computer program.

[0043] To achieve the above object, the present invention further provides a computer-readable storage medium, on which a computer program is stored. When the computer program stored in the computer-readable storage medium is executed by a processor, the steps of the above motor state detection method are implemented.

[0044] The present invention provides a motor state detection method, device, computer equipment and readable storage medium, which distinguish between robots in running state and shutdown state by collecting start and stop information of at least one robot, including: directly testing robots in running state to achieve the technical effect of real-time detection of running robots; and achieving targeted detection of robots in shutdown state, so as to ensure that robots with abnormal motors are detected while avoiding large-scale startup of robots due to testing, which will affect the normal production of various factory areas.

[0045] If it is determined that the first robot is in an operating state, state information of at least one motor of the first robot is collected, and if it is determined that the state information of the first motor has a similarity with historical fault information of the first motor that exceeds a preset state similarity threshold, the first motor is determined to be an abnormal motor, and the similarity between each motor in an operating state and other motors of the same motor type that have failed is compared, and then the probability of failure of the first motor in the operating state is predicted, which not only improves the detection accuracy, but also sets the motor that may fail as an abnormal motor, and repairs and maintains it in a timely manner, thereby avoiding serious failure of the motor and causing it to be scrapped due to failure to obtain timely maintenance and repair.

[0046] If it is determined that the second robot is in a shutdown state, the production information of the second robot is collected. If it is determined that the production information of the second robot and the historical fault information of the second robot belong to the same risk category, the second robot is determined to be a robot to be identified. This allows robots in a shutdown state that may have faulty motors to be screened based on the production information, and the screened-out motors are used as motors to be identified, thereby avoiding subsequent large-scale startup inspections of robots in a shutdown state, which would affect the production rhythm and order of the factory.

[0047] By starting the robot to be identified and collecting test information of at least one motor in the robot to be identified, if it is determined that the number of voltage signals belonging to the same category between the test information of the second motor and the historical fault information of the second motor reaches a preset test similarity threshold, the second motor is determined to be an abnormal motor. The motor in the robot to be identified is tested to identify the degree of similarity between the motor and its corresponding historical fault information, thereby ensuring the accuracy of identifying abnormal motors in the robot in a stopped state, and then predicting the probability of failure of the second motor in the stopped state. This not only improves the detection accuracy, but also sets the motor that may fail as an abnormal motor, and repairs and maintains it in a timely manner, thereby avoiding serious failure of the motor and causing it to be scrapped due to failure to be maintained and repaired in a timely manner. BRIEF DESCRIPTION OF THE DRAWINGS

[0048] Figure 1 This is a schematic block diagram of the environmental application of the motor state detection method of this application;

[0049] Figure 2 is a flow chart of the motor state detection method of the present application;

[0050] Figure 3 A schematic diagram of a program module of a motor state detection device of the present application;

[0051] Figure 4 This is a schematic diagram of the hardware structure of the computer device in the fourth embodiment of the computer device of the present application. DETAILED DESCRIPTION

[0052] In order to make the purpose, technical scheme and advantages of the present invention more clearly understood, the present invention is further described in detail below in conjunction with the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not intended to limit the present invention. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative work are within the scope of protection of the present invention.

[0053] Embodiment 1: Figure 1 The environmental application diagram according to the present application is schematically shown.

[0054] In an exemplary embodiment, the server 11 where the motor state detection method is located is connected to the robot 12 and the motor 13 installed in the robot 12 respectively.

[0055] Figure 2 It is a flow chart of a motor state detection method provided in the present application, which specifically includes steps S201 to S207.

[0056] See also Figure 2The present application provides a motor state detection method for detecting at least one motor in at least one robot, the method comprising:

[0057] S201: Collecting start and stop information of at least one robot; wherein the start and stop information includes a running state and a shutdown state.

[0058] In this example, by collecting the start and stop information of at least one robot, robots in the running state and the shutdown state are treated differently, including: directly testing the robot in the running state to achieve the technical effect of real-time detection of the running robot; and achieving targeted detection of the robot in the shutdown state, so as to ensure the detection of robots with abnormal motors while avoiding large-scale startup of robots due to testing, which will affect the normal production of various factory areas.

[0059] In a preferred embodiment, collecting start and stop information of at least one robot includes:

[0060] Accessing a cloud database, wherein the cloud database has at least one list, the list represents at least one robot in a factory area, and the list records a unique identifier of at least one robot;

[0061] Obtaining a unique identifier of each robot in a first list from a database, and sending a query message to each device address; wherein the first list is one of at least one list;

[0062] Receive the start and stop information sent by each robot according to the inquiry message.

[0063] In this example, the cloud database is a database deployed in a cloud computing environment. It is deployed in a virtual computing environment. The cloud database stores lists of multiple factories. Therefore, you only need to access the cloud database to obtain lists of all factories.

[0064] A device address is a string of numbers or character combinations used to uniquely identify and locate a device in a network or communication system.

[0065] The purpose of the inquiry message is to query the current operating status of a device or system, that is, whether the device or system is in operation or shutdown. The main purpose of the inquiry message is to obtain real-time status information of the device or system for further operation, monitoring or management.

[0066] The inquiry message includes: the device address of the robot (such as IP address, MAC address, device ID, etc.).

[0067] Specific information about the start / stop status requested, such as "Request to obtain the current start / stop status of robot A".

[0068] Identification information for the sender (requester) so that the recipient can recognize and respond to the request.

[0069] The start and stop information generated by the robot according to the inquiry message is the response generated according to the inquiry message. When the robot receives the start and stop status inquiry message, it will generate a response message according to the current status. The response message contains the actual start and stop status information of the device or system. The response message contains the status code, status description or other relevant information so that the sender can accurately understand the current status of the device.

[0070] S202: If it is determined that the first robot is in an operating state, the state information of at least one motor of the first robot is collected; wherein the first robot is one of at least one robot; the state information includes at least one voltage signal representing an operating parameter of the robot, and at least one voltage signal representing an operating condition of the robot.

[0071] In a preferred embodiment, collecting status information of at least one motor of the first robot includes:

[0072] Acquire the operation collection rules and working condition collection rules of the operation collection motor; wherein the operation collection motor is one of at least one motor of the first robot; the operation collection rules record the instrument used to collect the operation parameters of the motor and the calling rules of the instrument; the working condition collection rules record the instrument used to collect the operation condition of the motor and the calling rules of the instrument;

[0073] Collecting the operating parameters of the motor according to the operating collection rules; wherein the operating parameters include: temperature signal, torque signal, voltage value signal, current value signal, resistance value signal and electromagnetic signal;

[0074] The operating conditions of the motor are collected according to the operating condition collection rules; wherein the operating conditions include: sound amplitude signals and mechanical vibration signals.

[0075] In this example, by obtaining the operation collection rules and operating condition collection rules of the operation collection motor as the basis for collecting the operating parameters and operating conditions of each motor, targeted data collection is achieved for each motor, ensuring the reliability and accuracy of data collection.

[0076] By setting the operating parameters to include: temperature signal, torque signal, voltage value signal, current value signal, resistance value signal and electromagnetic signal, comprehensive collection of the operating conditions of the motor is achieved.

[0077] Among them, for temperature abnormalities, excessively high motor temperature will cause accelerated aging of the internal insulation material of the motor and shorten the service life of the motor; high temperature may also affect the electromagnetic characteristics of the motor, causing the motor performance to deteriorate.

[0078] Abnormal torque may cause the robot to move unsteadily or even damage the transmission mechanism. If the robot runs under high torque for a long time, the motor may overheat easily and shorten its life.

[0079] For voltage abnormalities, voltage that is too high or too low may cause the motor to malfunction or even damage the motor driver; voltage fluctuations will affect the stability and life of the motor.

[0080] For abnormal current, excessive current may cause the motor to overheat or even burn out; unstable current will affect the performance and service life of the motor.

[0081] For abnormal resistance, too large resistance may cause the motor current to decrease, affecting the motor performance; too small resistance may cause the motor current to be too large, easily burning the motor.

[0082] Electromagnetic anomalies may cause the motor to generate an abnormal magnetic field, affecting the motor's performance and service life; electromagnetic interference may also affect other electronic equipment of the robot.

[0083] By setting the operating conditions to include: sound amplitude signal, mechanical vibration signal, comprehensive collection of the motor's operating conditions can be achieved.

[0084] Among them, for abnormal sound amplitude, high mechanical amplitude will accelerate the wear of the motor and shorten the service life of the motor; excessive mechanical amplitude may also cause unstable operation of the motor, affecting the efficiency and output power of the motor; excessive amplitude will cause excessive wear and deformation on the motor surface, which not only affects the appearance of the motor, but may also reduce its accuracy.

[0085] For abnormal mechanical vibration, the sound amplitude is usually directly related to the noise level of the motor; excessive noise not only affects the working environment, but may also indicate abnormalities or wear inside the motor; long-term exposure to high noise environment may cause damage to the operator's hearing.

[0086] Specifically, the operation collection rules and working condition collection rules of the operation collection motor are obtained, including:

[0087] The motor type of the running acquisition motor is obtained, and the running acquisition rule and the working condition acquisition rule corresponding to the motor type are obtained from a preset rule database.

[0088] Exemplarily, the instrument for collecting the operation collection rules of the temperature signal is a temperature sensor, which is used to detect the temperature of the motor; the calling method of the instrument is to send A temperature detection instructions to the temperature sensor within the temperature collection time period, and obtain A temperature feedback voltages, and use the average value of the A temperature feedback voltages as the temperature signal.

[0089] The instrument used to collect the running collection rules of the torque signal is a torque sensor, which is installed on the motor shaft of the motor and is used to measure the torque output by the motor. The calling method of the instrument is to send B torque detection instructions to the torque sensor within the torque collection time period, and obtain B torque feedback voltages, draw a torque characteristic curve with the B torque feedback voltages as characteristic points, and use the torque characteristic curve as the torque signal.

[0090] The instrument for collecting the operation collection rules of the torque signal is a power collection module, which is used to collect the current value and speed of the motor, and calculate the current value and speed to form a power value; the calling method of the instrument is to send B1 torque detection instructions to the torque sensor within the torque collection time period, and obtain B1 torque feedback voltages, draw a torque characteristic curve with B1 torque feedback voltages as feature points, and use the torque characteristic curve as the torque signal.

[0091] The instrument in the operation collection rule for collecting voltage value signals is a voltmeter, which is used to measure the voltage of the motor; the calling method of the instrument is to send C voltage value detection instructions to the voltage value sensor within the voltage value collection time period, and obtain C voltage value feedback voltages, draw a voltage characteristic curve with the C voltage value feedback voltages as characteristic points, and use the voltage characteristic curve as the voltage value signal.

[0092] The instrument in the operation collection rule for collecting current value signals is an ammeter, which is used to measure the current of the motor; the calling method of the instrument is to send D current value detection instructions to the current value sensor within the current value collection time period, and obtain D current value feedback voltages, draw a current characteristic curve with the D current value feedback voltages as characteristic points, and use the current characteristic curve as the current value signal.

[0093] The instrument in the operation collection rule for collecting resistance value signals is a multimeter, which is used to measure the resistance of the motor; the calling rule of the instrument is to send E resistance value detection instructions to the resistance value sensor within the resistance value collection time period, and obtain E resistance value feedback voltages, and use the average value of the E resistance value feedback voltages as the resistance value signal.

[0094] The instrument in the operation collection rule for collecting electromagnetic signals is a magnetic field measuring instrument, which is used to measure the magnetic field strength generated by the motor; the instrument calling rule is to send F electromagnetic detection instructions to the electromagnetic sensor within the electromagnetic collection time period, and obtain F electromagnetic feedback voltages, draw an electromagnetic characteristic curve with the F electromagnetic feedback voltages as feature points, and set the electromagnetic characteristic curve as the electromagnetic signal.

[0095] The instrument in the working condition collection rule for collecting sound amplitude signals is a vibration measuring instrument, such as an accelerometer or a vibration sensor, which is directly fixed at an appropriate position of the motor for measurement; the instrument calling rule is to send G sound amplitude detection instructions to the sound amplitude sensor within the sound amplitude collection time period, and obtain G sound amplitude feedback voltages, and use the average value of the G sound amplitude feedback voltages as the sound amplitude signal.

[0096] The instrument in the working condition collection rule for collecting mechanical vibration signals is a sound level meter, which is used to measure the sound pressure level generated by the motor when it is running, which can reflect the size of the sound amplitude. The instrument calling rule is to send H mechanical vibration detection instructions to the mechanical vibration sensor within the mechanical vibration collection time period, and obtain H mechanical vibration feedback voltages, and use the average value of the H mechanical vibration feedback voltages as the mechanical vibration signal.

[0097] S203: If it is determined that the similarity between the state information of the first motor and the historical fault information of the first motor exceeds a preset state similarity threshold, the first motor is determined to be an abnormal motor; wherein the first motor is one of at least one motor; the historical fault information of the first motor is state information of a motor of the same motor type as the first motor when a fault occurred in history.

[0098] In this example, if it is determined that the first robot is in an operating state, state information of at least one motor of the first robot is collected, and if it is determined that the state information of the first motor has a similarity with the historical fault information of the first motor that exceeds a preset state similarity threshold, the first motor is determined to be an abnormal motor, and the degree of similarity between each motor in an operating state and other motors of the same motor type that have failed is compared, thereby predicting the probability of failure of the first motor in an operating state, which not only improves the detection accuracy, but also sets the motor that may fail as an abnormal motor, and repairs and maintains it in a timely manner, thereby avoiding serious failure of the motor, which is scrapped due to the inability to obtain timely maintenance and repair.

[0099] Exemplarily, the equipment types include: handling type, where handling robots are used to move autonomously along a preset path to complete operations such as handling, transportation, assembly, and stacking.

[0100] Disassembly type: Disassembly type robots are used for automated disassembly equipment.

[0101] Welding and welding machine robots: used for various types of welding, such as spot welding, gas shielded welding, brazing, etc.

[0102] SCARA (Selective Compliance Assembly Robot Arm), SCARA robots are four-axis robots with high-precision planar positioning capabilities. They are used for electronic component assembly and light load handling, and are suitable for semiconductor manufacturing, electronic assembly and other fields.

[0103] PUMA type, PUMA robots have 6 degrees of freedom and are suitable for high-speed precision machining operations.

[0104] Delta type, Delta type robot is a triangular structure with a higher picking speed. It is suitable for high-speed picking and packaging operations such as bottle and snack packaging.

[0105] Laser cutting category: laser cutting robots are used to accurately cut various shapes.

[0106] The categories of motors include: stepper motors, brushed DC motors, brushless DC motors, AC servo motors, frameless torque motors and coreless motors.

[0107] The locations of the motors include: the base of the robot (for the overall rotation of the robot), the robot arm joint (for the rotation of the robot arm), the robot wrist joint (for the rotation of the robot fingers), the robot arm (arm extension and retraction), and the robot fingers (for controlling the grasping of the robot fingers).

[0108] In a preferred embodiment, if it is determined that the similarity between the state information of the first motor and the historical fault information of the first motor exceeds a preset state similarity threshold, then determining that the first motor is an abnormal motor includes:

[0109] Calling a preset mature neural network model to calculate the state information of the first motor to obtain a network similarity result; if it is determined that the network similarity result exceeds a network threshold of a state similarity threshold, determining that the first motor is an abnormal motor; wherein the mature neural network model is obtained by training an initial neural network model through at least one historical fault information of the first motor; or

[0110] A preset mature clustering model is called to perform clustering operation on the state information of the first motor to obtain the similarity distance between the state information of the first motor and the cluster center of each clustering cluster in the mature clustering model; if it is determined that at least one similarity distance exceeds the clustering threshold in the state similarity threshold, the first motor is determined to be an abnormal motor; wherein the mature clustering model is a model having at least one clustering cluster formed by entering historical fault information of at least one first motor into an initial clustering model, and calling the initial clustering model to cluster the entered historical fault information; the cluster center is the center point of the clustering cluster.

[0111] In this example, the torque signal, voltage value signal, current value signal and electromagnetic signal are characteristic curves; the temperature signal, resistance value signal, sound amplitude signal and mechanical vibration signal are specific values.

[0112] In an exemplary embodiment, a preset mature neural network model is called to operate on the state information of the first motor to obtain a network similarity result.

[0113] Mature neural network models include: mature MLP model and mature CNN model; MLP model is a multi-layer perceptron, which is a classification model based on artificial neural network, which can classify according to the input numerical data. CNN (Convolutional Neural Network) model is a neural network specially designed to process data with grid-like structure (such as images). The core idea of ​​CNN is to use convolutional layer to extract local features of input data. Convolutional layer scans the input data by applying a series of learnable filters (also called convolution kernels) and generates feature maps, which describe the local characteristics of the input data.

[0114] Specifically, the initial MLP model is trained by using the historical temperature signal, the historical resistance value signal, the historical sound amplitude signal and the historical mechanical vibration signal in the historical fault information of at least one first motor to obtain a mature MLP model. The initial CNN model is trained by using the historical torque signal, the historical voltage value signal, the historical current value signal and the historical electromagnetic signal in the historical fault information of at least one first motor to obtain a mature CNN model.

[0115] The network similarity results include: MLP results and CNN results. The MLP results reflect the similarity between the temperature signal, resistance value signal, sound amplitude signal and mechanical vibration signal in the status information of the first motor and the historical temperature signal, historical resistance value signal, historical sound amplitude signal and historical mechanical vibration signal in the historical fault information of the first motor.

[0116] The CNN results reflect the similarity between the torque signal, voltage value signal, current value signal and electromagnetic signal of the first motor and the historical torque signal, historical voltage value signal, historical current value signal and historical electromagnetic signal in the historical fault information of the first motor.

[0117] The network threshold of the state similarity threshold includes: an MLP threshold and a CNN threshold; if it is determined that the MLP result exceeds the MLP threshold and / or the CNN result exceeds the CNN threshold, it is determined that the network similarity result exceeds the network threshold of the state similarity threshold.

[0118] In another exemplary embodiment, the initial clustering model includes: a clustering calculation module and a K-Means clustering model; the clustering calculation module sequentially calculates the mean of characteristic points in the historical torque signal, the historical voltage value signal, the historical current value signal, and the historical electromagnetic signal to obtain the historical torque mean signal, the historical voltage value mean signal, the historical current value mean signal, and the historical electromagnetic mean signal;

[0119] A historical temperature signal, a historical resistance value signal, a historical sound amplitude signal, a historical mechanical vibration signal, a historical torque mean signal, a historical voltage mean signal, a historical current mean signal and a historical electromagnetic mean signal of a historical fault information are entered into a K-Means clustering model as a data point to obtain a K-Means clustering model with at least one data point, the K-Means clustering model is called to perform clustering operation on at least one data point to obtain a mature clustering model with at least one cluster cluster, and the clustering calculation module is loaded into the mature clustering model.

[0120] The clustering calculation module in the mature clustering model is called to calculate the mean of the characteristic points in the torque signal, voltage value signal, current value signal and electromagnetic signal in the state information in turn, and the torque mean signal, voltage mean signal, current mean signal and electromagnetic mean signal are obtained.

[0121] The temperature signal, resistance value signal, sound amplitude signal, mechanical vibration signal, torque mean signal, voltage mean signal, current mean signal and electromagnetic mean signal of a state information are entered into the mature clustering model as a target point, and the distance between the target point and the cluster center of each cluster in the mature clustering model is calculated to obtain the similarity distance between the target point and the cluster center.

[0122] If a similarity distance exceeds the clustering threshold in the state similarity threshold, it means that the target point belongs to the cluster where the center point corresponding to the similarity distance is located. Therefore, it means that the state information is similar or even identical to the historical fault information in the cluster, so the first motor is determined to be an abnormal motor.

[0123] S204: If it is determined that the second robot is in a shutdown state, the production information of the second robot is collected; wherein the second robot is one of at least one robot; and the production information includes at least one data information characterizing the operation status of the robot within a preset production time period.

[0124] In a preferred embodiment, collecting the production information of the second robot includes:

[0125] Accessing the log database of the second robot;

[0126] The production information of the second robot in a preset production time period is obtained; wherein at least one piece of data information in the production information includes: operating time, operating ratio, operating rate, and equipment utilization rate.

[0127] In this example, the log database is a database that records the operation status of the robot.

[0128] The running time of a robot refers to the length of time that the robot runs continuously in actual working conditions. When the robot is running idle, it is generally necessary to limit its running time to avoid excessive power consumption, which may cause the task to fail or the robot to malfunction. Generally speaking, the robot's idle running time is about 8-10 hours. The length of the robot's running time is affected by many factors, including the robot's power, the robot's movement mode, the robot's task type, the ambient temperature, the robot's load, etc.

[0129] The operation rate is the adjustment factor of the robot's movement speed. If we understand "operation rate" as the adjustment factor of the robot's movement speed, then it may be expressed as a percentage (%) to adjust the actual speed of the robot's movement. For example, when the speed rate is 100%, the robot moves at the fastest speed.

[0130] The operating rate is usually used to describe the ratio of the actual working time of the equipment to the possible working time. But in the field of robotics, the operating rate may refer to the ratio of the actual operating time of the robot to the total operating time in a period of time.

[0131] This indicator reflects the actual usage and production efficiency of the robot. A high operating rate usually means that the robot is fully utilized, while a low operating rate may indicate that there is a problem of idle equipment or low production efficiency.

[0132] Equipment utilization is the ratio of the time that a robot is actually used for production or task execution to the total time within a certain period of time. Equipment utilization is an important indicator to measure the efficiency of robot use. High equipment utilization means that the robot is fully utilized and can create more value for the enterprise. Low equipment utilization may indicate a waste of resources or low production efficiency.

[0133] S205: If it is determined that the production information of the second robot is similar to the historical fault information of the second robot, the second robot is determined to be the robot to be identified; wherein the historical fault information of the second robot is the production information of a robot of the same equipment type as the second robot when a fault occurred in the past.

[0134] In this example, if it is determined that the second robot is in a shutdown state, the production information of the second robot is collected; if it is determined that the production information of the second robot is similar to the historical fault information of the second robot, the second robot is determined to be the robot to be identified. Based on the production information, robots in a shutdown state that may have faulty motors are screened, and the screened-out motors are used as motors to be identified, thereby avoiding subsequent large-scale startup inspection of robots in a shutdown state, which affects the production rhythm and production order of the factory.

[0135] In a preferred embodiment, if it is determined that the production information of the second robot is similar to the historical fault information of the second robot, determining that the second robot is a robot to be identified includes:

[0136] Calling a preset mature support vector machine to operate on the production information of the second robot to obtain the production category of the second robot; wherein the mature support vector machine is formed by training an initial support vector machine pair through historical normal information of at least one second robot and historical fault information of at least one second robot; the historical normal information of the second robot is the production information of a robot of the same equipment type as the second robot when operating normally in history;

[0137] If the content of the production category is determined to be a risk category, the second robot is determined to be a robot to be identified; wherein the risk category is a category in which the historical fault information of the second robot in the mature support vector machine belongs

[0138] In this example, the Support Vector Machine (SVM) is a generalized linear classifier that performs binary classification of data in a supervised learning manner. It uses a decision boundary (i.e., a hyperplane) to separate data points into different categories. This decision boundary is determined by maximizing the margin between two categories.

[0139] The training method of support vector machine is to find a hyperplane given a set of training samples. The goal of SVM is to find a hyperplane so that the distance between the points closest to the hyperplane and the hyperplane is the largest. These points closest to the hyperplane are called support vectors. The decision-making process of SVM depends only on these support vectors, so it is more efficient in processing high-dimensional data.

[0140] Therefore, by using mature support vector machines to identify the production category of robots, robots that may have abnormal motors can be identified efficiently and quickly, thereby improving the efficiency of screening abnormal motors.

[0141] Specifically, historical normal information of at least one second robot and historical fault information of at least one second robot are classified by an initial support vector machine to obtain a mature support vector machine having a hyperplane capable of separating historical normal information and historical fault information, wherein the hyperplane separates the space of the mature support vector machine into an area where historical normal information is located and an area where historical fault information is located.

[0142] The production information of the second robot is entered into a mature support vector machine, and the mature support vector machine is called to determine whether the production information belongs to the area where historical normal information is located or the area where historical fault information is located; if it is the area where historical normal information is located, the content of the production category of the production information is determined to be a normal category; if it is the area where historical fault information is located, the content of the production category of the production information is determined to be a risk category.

[0143] The initial support vector machine pair is trained; the historical normal information of the second robot is the production information of a robot of the same equipment type as the second robot when it operates normally in history.

[0144] S206: Start the robot to be identified and collect test information of at least one motor in the robot to be identified; wherein the test information includes at least one voltage signal representing a test parameter of the robot within a test time period, and at least one voltage signal representing a test condition of the robot within the test time period.

[0145] Optionally, starting the robot to be identified and collecting test information of at least one motor in the robot to be identified includes:

[0146] Obtaining a startup strategy corresponding to the robot model of the robot to be identified, and starting the robot to be identified according to the startup strategy; wherein the startup strategy records the startup time, running direction, and running power of each motor in the robot;

[0147] Acquire the test operation rules and test condition rules corresponding to the motor type of the third motor in the robot to be identified; the test operation rules record the instrument used to collect the test parameters of the motor and the calling rules of the instrument; the test condition rules record the instrument used to collect the test condition of the motor and the calling rules of the instrument;

[0148] Collecting and running the test parameters of the motor according to the test running rules; wherein the test parameters include: one or more of a temperature signal, a torque signal, a voltage value signal, a current value signal, a resistance value signal and an electromagnetic signal;

[0149] The test working condition of the motor is collected according to the working condition collection rule; wherein the test working condition includes: sound amplitude signal and / or mechanical vibration signal.

[0150] In this example, the number of motors, motor types, and motor performances of robots of different device types are different. Therefore, a startup strategy needs to be compiled for each device type of robot to meet the testing requirements of the robots of each device type.

[0151] The startup strategy creates the startup time (for example, 1 minute, 3 minutes), startup direction (for example, 3000 revolutions forward, 5000 revolutions reverse), and operating power (for example, for a motor with a rated power of 5000W, the operating power is 4500W) for each motor in a robot of one device type.

[0152] By setting the test parameters to one or more of: temperature signal, torque signal, voltage value signal, current value signal, resistance value signal and electromagnetic signal; and setting the test conditions to include: sound amplitude signal and / or mechanical vibration signal, the collection of voltage signals can be reduced, thereby improving the efficiency of collecting test information and reducing the test time.

[0153] Exemplarily, the instrument for collecting temperature signals for the test operation rules is a temperature sensor, which is used to detect the temperature of the motor; the calling method of the instrument is to send A1 temperature detection instructions to the temperature sensor within the temperature collection time period, and obtain A1 temperature feedback voltages, and use the average value of the A1 temperature feedback voltages as the temperature signal.

[0154] The instrument used to collect the test operation rules of the torque signal is a torque sensor, which is installed on the motor shaft of the motor to measure the torque output by the motor. The calling method of the instrument is to send B2 torque detection instructions to the torque sensor within the torque collection time period, and obtain B2 torque feedback voltages, draw a torque characteristic curve with the B2 torque feedback voltages as characteristic points, and use the torque characteristic curve as the torque signal.

[0155] The instrument for collecting the test operation rules of the torque signal is a power acquisition module, which is used to collect the current value and speed of the motor, and calculate the current value and speed to form the power value; the calling method of the instrument is to send B3 torque detection instructions to the torque sensor within the torque acquisition time period, and obtain B3 torque feedback voltages, draw a torque characteristic curve with B3 torque feedback voltages as characteristic points, and use the torque characteristic curve as the torque signal.

[0156] The instrument in the test operation rules for collecting voltage value signals is a voltmeter, which is used to measure the voltage of the motor; the calling method of the instrument is to send C1 voltage value detection instructions to the voltage value sensor within the voltage value collection time period, and obtain C1 voltage value feedback voltages, draw a voltage characteristic curve with C1 voltage value feedback voltages as characteristic points, and use the voltage characteristic curve as the voltage value signal.

[0157] The instrument in the test operation rules for collecting current value signals is an ammeter, which is used to measure the current of the motor; the calling method of the instrument is to send D1 current value detection instructions to the current value sensor within the current value collection time period, and obtain D1 current value feedback voltages, draw a current characteristic curve with D1 current value feedback voltages as characteristic points, and use the current characteristic curve as the current value signal.

[0158] The instrument in the test operation rule for collecting resistance value signals is a multimeter, which is used to measure the resistance of the motor; the instrument calling rule is to send E1 resistance value detection instructions to the resistance value sensor within the resistance value collection time period, and obtain E1 resistance value feedback voltages, and use the average value of the E1 resistance value feedback voltages as the resistance value signal.

[0159] The instrument in the test operation rules for collecting electromagnetic signals is a magnetic field measuring instrument, which is used to measure the magnetic field strength generated by the motor; the instrument calling rule is to send F1 electromagnetic detection instructions to the electromagnetic sensor within the electromagnetic collection time period, and obtain F1 electromagnetic feedback voltages, draw an electromagnetic characteristic curve with the F1 electromagnetic feedback voltages as feature points, and set the electromagnetic characteristic curve as the electromagnetic signal.

[0160] The instrument in the test condition rule for collecting sound amplitude signals is a vibration measuring instrument, such as an accelerometer or a vibration sensor, which is directly fixed at an appropriate position on the motor for measurement; the instrument calling rule is to send G1 sound amplitude detection instructions to the sound amplitude sensor within the sound amplitude collection time period, and obtain G1 sound amplitude feedback voltages, and use the average value of the G1 sound amplitude feedback voltages as the sound amplitude signal.

[0161] The instrument in the test condition rule for collecting mechanical vibration signals is a sound level meter, which is used to measure the sound pressure level generated by the motor when it is running, which can reflect the amplitude of the sound. The instrument calling rule is to send H1 mechanical vibration detection instructions to the mechanical vibration sensor within the mechanical vibration collection time period, and obtain H1 mechanical vibration feedback voltages, and use the average value of the H1 mechanical vibration feedback voltages as the mechanical vibration signal.

[0162] S207: If it is determined that the test information of the second motor is similar to the historical fault information of the second motor, the second motor is determined to be an abnormal motor; wherein the second motor is a motor in the robot to be identified; the historical fault information of the second motor is status information of a motor of the same motor type as the second motor when a fault occurred in history.

[0163] In this example, by starting the robot to be identified and collecting test information of at least one motor in the robot to be identified, if it is determined that the test information of the second motor is similar to the historical fault information of the second motor, the second motor is determined to be an abnormal motor, and the motor in the robot to be identified is tested to identify the degree of similarity between the motor and its corresponding historical fault information, thereby ensuring the accuracy of identifying abnormal motors in the robot in a stopped state, and then predicting the probability of failure of the second motor in the stopped state, which not only improves the detection accuracy, but also sets the motor that may fail as an abnormal motor, and repairs and maintains it in a timely manner, thereby avoiding serious failure of the motor and causing it to be scrapped due to failure to obtain timely maintenance and repair.

[0164] In a preferred embodiment, if it is determined that the test information of the second motor is similar to the historical fault information of the second motor, then determining that the second motor is an abnormal motor includes:

[0165] Calling a preset mature random forest model to operate on the test information of the second motor to obtain at least one classification category of the test information of the second motor; wherein the mature random forest model is composed of at least one mature decision tree; the mature decision tree is a binary tree model with at least one decision condition generated by dividing the historical normal information of at least one second motor and the historical abnormal information of at least one second motor through the initial decision tree; the mature decision tree is used to divide the voltage signal corresponding to the decision condition in the test information according to the at least one decision condition, and obtain the classification category of the voltage signal corresponding to the decision condition;

[0166] If it is determined that the content of at least one classification category is a fault category, the second motor is determined to be an abnormal motor; the fault category is a category to which historical fault information of the second motor in the mature random forest belongs.

[0167] In this example, Random Forest is an ensemble learning method based on decision trees. In machine learning and data mining, Random Forest is a classifier that contains multiple decision trees, and its output category is determined by the mode of the category output by the individual trees. In this embodiment, the random selection of the data set of Random Forest is to randomly extract k new self-help data sets from the original data set with replacement to construct k decision trees, and no pruning is performed during the construction process.

[0168] For the random selection of features in a random forest, it is assumed that there are m features in the original dataset. Then, each time n features (n << m) are randomly selected from the m features to construct a decision tree.

[0169] The initial random forest model includes: a forest calculation module and a random forest; the forest calculation module calculates the means of the feature points in the historical torque signal, historical voltage value signal, historical current value signal, and historical electromagnetic signal of the historical normal information and historical abnormal information, and obtains the historical torque mean signal, historical voltage value mean signal, historical current value mean signal, and historical electromagnetic mean signal.

[0170] The random forest extracts one of the initial trees, obtains one or more of them from the historical temperature signal, historical resistance value signal, historical sound amplitude signal, historical mechanical vibration signal, as well as the historical torque mean signal, historical voltage value mean signal, historical current value mean signal, and historical electromagnetic mean signal as decision elements, classifies the historical normal information and historical abnormal information to determine the decision conditions for classifying the historical normal information and historical abnormal information for each decision element, and obtains a decision tree with at least one decision element and its decision conditions. Among them, the decision element is the metadata used to classify the historical normal information and historical abnormal information, and the decision condition is the decision content based on the decision element for the historical normal information and historical abnormal information.

[0171] Furthermore, the test information of the second motor has one or more of the temperature signal, torque signal, voltage value signal, current value signal, resistance value signal, and electromagnetic signal, as well as the sound amplitude signal and / or mechanical vibration signal.

[0172] Therefore, at least one decision tree corresponding to the test information is obtained, and the test information is calculated through the obtained decision tree to obtain at least one classification category corresponding to at least one decision tree; among them, if the decision tree classifies the test information into the same category as the historical normal information, it is determined that the content of the classification category is the normal category; if the decision tree classifies the test information into the same category as the historical fault information, it is determined that the content of the classification category is the fault category; if the content of one of the obtained classification categories is the fault category, it is determined that the second motor is an abnormal motor.

[0173] Therefore, the mature random forest model can quickly and accurately determine whether the second motor is an abnormal motor based on limited test information, improving the identification efficiency of abnormal motors.

[0174] Exemplarily, assume that the decision elements of decision tree 1 are the historical temperature signal and the historical resistance value signal.

[0175] The decision elements of decision tree 2 are the historical sound amplitude signal and the historical torque mean signal;

[0176] The decision elements of decision tree 3 are the historical mechanical vibration signal and the historical voltage value mean signal;

[0177] The decision elements of decision tree 4 are the historical current value mean signal and the historical electromagnetic mean signal.

[0178] If the test information includes: temperature signal, resistance value signal, voltage value signal and mechanical vibration signal.

[0179] Then, the test information is classified using decision tree 1 and decision tree 3 to obtain the classification category output by decision tree 1 and the classification category of decision tree 3. If at least one of the two classification categories is a fault category, it is determined that the motor corresponding to the test information is an abnormal motor.

[0180] Example 2: Please refer to Figure 3 The present application also provides a motor state detection device 3, comprising: for detecting at least one motor in at least one robot, the motor state detection device 3 comprising:

[0181] A start-stop collection module 31 is used to collect start-stop information of at least one robot; wherein the start-stop information includes a running state and a shutdown state;

[0182] The operation acquisition module 32 is used to acquire state information of at least one motor of the first robot if it is determined that the first robot is in the operation state; wherein the first robot is one of the at least one robot; and the state information includes at least one voltage signal representing an operation parameter of the robot and at least one voltage signal representing an operation condition of the robot;

[0183] The operation determination module 33 is used to determine that the first motor is an abnormal motor if it is determined that the similarity between the state information of the first motor and the historical fault information of the first motor exceeds a preset state similarity threshold; wherein the first motor is one of the at least one motor; and the historical fault information of the first motor is state information of a motor of the same motor type as the first motor when a fault occurs in history;

[0184] A shutdown collection module 34 is used to collect production information of the second robot if it is determined that the second robot is in a shutdown state; wherein the second robot is one of the at least one robot; and the production information includes at least one data information representing the operation status of the robot within a preset production time period;

[0185] The robot determination module 35 is used to determine that the second robot is a robot to be identified if it is determined that the production information of the second robot is similar to the historical fault information of the second robot; wherein the historical fault information of the second robot is the production information of a robot of the same equipment type as the second robot when a fault occurred in the past;

[0186] A testing module 36, used to start the robot to be identified and collect test information of at least one motor in the robot to be identified; wherein the test information includes at least one voltage signal representing a test parameter of the robot within a test time period, and at least one voltage signal representing a test condition of the robot within the test time period;

[0187] The shutdown determination module 37 is used to determine that the second motor is an abnormal motor if the test information of the second motor is similar to the historical fault information of the second motor; wherein the second motor is a motor in the robot to be identified; and the historical fault information of the second motor is the status information of a motor of the same motor type as the second motor when a fault occurred in history.

[0188] Embodiment 3: To achieve the above-mentioned purpose, the present invention further provides a computer device 4. The components of the motor state detection device of Embodiment 3 can be dispersed in different computer devices. The computer device 4 can be a smart phone, tablet computer, laptop computer, desktop computer, rack server, blade server, tower server or cabinet server (including independent server, or server cluster composed of multiple application servers) that executes the program. The computer device of this embodiment includes at least but is not limited to: a memory 41 and a processor 42 that can be interconnected through a system bus, such as Figure 4 It should be pointed out that Figure 4 Only a computer device with components is shown, but it should be understood that implementing all of the components shown is not a requirement and more or fewer components may alternatively be implemented.

[0189] In this embodiment, the memory 41 (i.e., a readable storage medium) includes a flash memory, a hard disk, a multimedia card, a card-type memory (e.g., an SD or DX memory, etc.), a random access memory (RAM), a static random access memory (SRAM), a read-only memory (ROM), an electrically erasable programmable read-only memory (EEPROM), a programmable read-only memory (PROM), a magnetic memory, a disk, an optical disk, etc. In some embodiments, the memory 41 may be an internal storage unit of a computer device, such as a hard disk or a memory of the computer device. In other embodiments, the memory 41 may also be an external storage device of a computer device, such as a plug-in hard disk, a smart memory card (Smart Media Card, SMC), a secure digital (Secure Digital, SD) card, a flash card (Flash Card), etc. equipped on the computer device. Of course, the memory 41 may also include both an internal storage unit of the computer device and its external storage device. In this embodiment, the memory 41 is generally used to store an operating system and various application software installed on the computer device, such as the program code of the motor state detection device of the third embodiment, etc. In addition, the memory 41 may also be used to temporarily store various types of data that have been output or are to be output.

[0190] In some embodiments, the processor 42 may be a central processing unit (CPU), a controller, a microcontroller, a microprocessor, or other data processing chip. The processor 42 is generally used to control the overall operation of the computer device. In this embodiment, the processor 42 is used to run the program code stored in the memory 41 or process data, such as running the motor state detection device to implement the motor state detection method of the first embodiment.

[0191] Embodiment 4: To achieve the above purpose, the present invention also provides a computer-readable storage medium, such as flash memory, hard disk, multimedia card, card-type memory (for example, SD or DX memory, etc.), random access memory (RAM), static random access memory (SRAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), programmable read-only memory (PROM), magnetic memory, disk, CD, server, App application store, etc., on which a computer program is stored, and the program realizes the corresponding function when executed by the processor 42. The computer-readable storage medium of this embodiment is used to store a computer program for realizing the motor state detection method, and when executed by the processor 42, the motor state detection method of embodiment 1 is realized.

[0192] The serial numbers of the above embodiments of the present invention are only for description and do not represent the advantages or disadvantages of the embodiments.

[0193] Through the description of the above implementation methods, those skilled in the art can clearly understand that the above embodiment methods can be implemented by means of software plus a necessary general hardware platform, and of course by hardware, but in many cases the former is a better implementation method.

[0194] The above are only preferred embodiments of the present invention, and are not intended to limit the patent scope of the present invention. Any equivalent structure or equivalent process transformation made using the contents of the present invention specification and drawings, or directly or indirectly applied in other related technical fields, are also included in the patent protection scope of the present invention.

Claims

1. A motor state detection method, characterized in that: For detecting at least one motor in at least one robot, the method comprises: Collecting start and stop information of at least one of the robots; wherein the start and stop information includes a running state and a shutdown state; If it is determined that the first robot is in the running state, then collecting state information of at least one motor of the first robot; wherein the first robot is one of at least one of the robots; the state information includes at least one voltage signal representing the running parameter of the robot, and at least one voltage signal representing the running condition of the robot; If it is determined that the similarity between the state information of the first motor and the historical fault information of the first motor exceeds a preset state similarity threshold, the first motor is determined to be an abnormal motor; wherein the first motor is one of at least one of the motors; the historical fault information of the first motor is state information of a motor of the same motor type as the first motor when a fault occurred in the past; If it is determined that the second robot is in a shutdown state, then collecting production information of the second robot; wherein the second robot is one of at least one of the robots; and the production information includes at least one piece of data information representing the operation status of the robot within a preset production time period; If it is determined that the production information of the second robot is similar to the historical fault information of the second robot, the second robot is determined to be the robot to be identified; wherein the historical fault information of the second robot is production information of a robot of the same device type as the second robot when a fault occurred in the past; Starting the robot to be identified and collecting test information of at least one motor in the robot to be identified; wherein the test information includes at least one voltage signal representing a test parameter of the robot within a test time period, and at least one voltage signal representing a test condition of the robot within the test time period; If it is determined that the test information of the second motor is similar to the historical fault information of the second motor, the second motor is determined to be an abnormal motor; wherein, the second motor is a motor in the robot to be identified; the historical fault information of the second motor is status information of a motor of the same motor type as the second motor when a fault occurred in history.

2. The motor state detection method according to claim 1, characterized in that: Collecting start and stop information of at least one of the robots, including: Accessing a cloud database, wherein the cloud database has at least one list, wherein the list represents at least one robot in a factory area, and wherein the list records a unique identifier of at least one robot; Obtaining a unique identifier of each robot in a first list from the database, and sending a query message to each device address; wherein the first list is one of at least one of the lists; Receive the start and stop information sent by each of the robots according to the inquiry message.

3. The motor state detection method according to claim 1, characterized in that: Collecting state information of at least one motor of the first robot includes: Acquire the operation collection rules and working condition collection rules of the operation collection motor; wherein the operation collection motor is one of at least one motor of the first robot; the operation collection rules record the instrument used to collect the operation parameters of the motor and the calling rules of the instrument; the working condition collection rules record the instrument used to collect the operation condition of the motor and the calling rules of the instrument; Collecting the operating parameters of the operating acquisition motor according to the operating acquisition rule; wherein the operating parameters include: temperature signal, torque signal, voltage value signal, current value signal, resistance value signal and electromagnetic signal; The operating condition of the operating motor is collected according to the operating condition collection rule; wherein the operating condition includes: a sound amplitude signal and a mechanical vibration signal.

4. The motor state detection method according to claim 1, characterized in that: If it is determined that the similarity between the state information of the first motor and the historical fault information of the first motor exceeds a preset state similarity threshold, determining that the first motor is an abnormal motor includes: Calling a preset mature neural network model to calculate the state information of the first motor to obtain a network similarity result; if it is determined that the network similarity result exceeds the network threshold of the state similarity threshold, determining that the first motor is an abnormal motor; wherein the mature neural network model is obtained by training an initial neural network model through at least one historical fault information of the first motor; or Call a preset mature clustering model to perform clustering operation on the state information of the first motor to obtain the similarity distance between the state information of the first motor and the cluster center of each cluster in the mature clustering model; if it is determined that at least one of the similarity distances exceeds the clustering threshold in the state similarity threshold, the first motor is determined to be an abnormal motor; wherein the mature clustering model is a model with at least one cluster cluster formed by entering at least one historical fault information of the first motor into an initial clustering model, and calling the initial clustering model to cluster the entered historical fault information; the cluster center is the center point of the cluster cluster.

5. The motor state detection method according to claim 1, characterized in that: Collecting production information of the second robot, including: accessing a log database of the second robot; Acquire production information of the second robot within a preset production time period; wherein at least one piece of data information in the production information includes: operating time, operating multiple, operating rate, and equipment utilization rate.

6. The motor state detection method according to claim 1, characterized in that: If it is determined that the production information of the second robot is similar to the historical fault information of the second robot, determining that the second robot is a robot to be identified includes: Calling a preset mature support vector machine to operate on the production information of the second robot to obtain the production category of the second robot; wherein the mature support vector machine is formed by training an initial support vector machine pair through historical normal information of at least one second robot and historical fault information of at least one second robot; the historical normal information of the second robot is production information of a robot of the same equipment type as the second robot when operating normally in history; If it is determined that the content of the production category is a risk category, the second robot is determined to be a robot to be identified; wherein the risk category is a category in which the historical fault information of the second robot in the mature support vector machine belongs.

7. The motor state detection method according to claim 1, characterized in that: If it is determined that the test information of the second motor is similar to the historical fault information of the second motor, determining that the second motor is an abnormal motor includes: Calling a preset mature random forest model to operate on the test information of the second motor to obtain at least one classification category of the test information of the second motor; wherein the mature random forest model is composed of at least one mature decision tree; the mature decision tree is a binary tree model with at least one decision condition generated by dividing the historical normal information of at least one second motor and the historical abnormal information of at least one second motor through an initial decision tree; the mature decision tree is used to divide the voltage signal corresponding to the decision condition in the test information according to at least one decision condition, and obtain a classification category of the voltage signal corresponding to the decision condition; If it is determined that the content of at least one of the classification categories is a fault category, the second motor is determined to be an abnormal motor; the fault category is a category to which the historical fault information of the second motor in the mature random forest belongs.

8. A motor state detection device, characterized in that: For detecting at least one motor in at least one robot, the device comprises: A start-stop acquisition module, used to collect start-stop information of at least one of the robots; wherein the start-stop information includes a running state and a shutdown state; An operation acquisition module, configured to acquire state information of at least one motor of the first robot if it is determined that the first robot is in an operation state; wherein the first robot is one of at least one of the robots; and the state information includes at least one voltage signal representing an operation parameter of the robot, and at least one voltage signal representing an operation condition of the robot; An operation determination module is used to determine that the first motor is an abnormal motor if it is determined that the similarity between the state information of the first motor and the historical fault information of the first motor exceeds a preset state similarity threshold; wherein the first motor is one of the at least one motor; and the historical fault information of the first motor is state information of a motor of the same motor type as the first motor when a fault occurs in history; A shutdown collection module, configured to collect production information of the second robot if it is determined that the second robot is in a shutdown state; wherein the second robot is one of the at least one robot; and the production information includes at least one piece of data information representing the operation status of the robot within a preset production time period; a robot determination module, configured to determine that the second robot is a to-be-identified robot if it is determined that the production information of the second robot is similar to the historical fault information of the second robot; wherein the historical fault information of the second robot is production information of a robot of the same device type as the second robot when a fault occurred in the past; A test module, used to start the robot to be identified and collect test information of at least one motor in the robot to be identified; wherein the test information includes at least one voltage signal representing a test parameter of the robot within a test time period, and at least one voltage signal representing a test condition of the robot within the test time period; A shutdown determination module is used to determine that the second motor is an abnormal motor if it is determined that the test information of the second motor is similar to the historical fault information of the second motor; wherein the second motor is a motor in the robot to be identified; and the historical fault information of the second motor is status information of a motor of the same motor type as the second motor when a fault occurred in history.

9. A computer device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that: When the processor of the computer device executes the computer program, the steps of the motor state detection method according to any one of claims 1 to 7 are implemented.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that: The computer program stored in the readable storage medium implements the steps of the motor state detection method according to any one of claims 1 to 7 when executed by a processor.

Citation Information

Patent Citations

  • Motor small sample fault diagnosis method based on multi-feature fusion under variable working conditions

    CN115859077A

  • Motor operation control method and system

    CN117458955A

  • Motor detection and repair method and device, electronic equipment and storage medium

    CN118091406A

  • Predicting failures in electrical submersible pumps using pattern recognition

    US20190317488A1