Robot fault detection method, device, storage medium and robot

Abstract

The present invention provides a robot fault detection method, device, storage medium and robot, and relates to the field of robot technology. The robot fault detection method includes: establishing a joint simulation model of the robot; obtaining fault data of the robot joint in a fault state; determining an actual noise spectrum according to the fault data; entering the fault data into the joint simulation model, and obtaining a simulated noise spectrum through the joint simulation model; and determining target fault information according to the actual noise spectrum and the simulated noise spectrum.

Description

Technical Field

[0001] The present invention relates to the field of robot technology, and in particular to a robot fault detection method, a device, a storage medium and a robot. Background Art

[0002] In the related art, when a robot joint fails in operation, it will make an abnormal sound. In theory, the fault problem of the joint can be judged based on the abnormal sound. At present, it is difficult to accurately locate the fault with the general noise analysis method and software, especially for the movement of complex structures such as robot joints, there is no good diagnosis method.

[0003] Therefore, how to overcome the above-mentioned technical defects has become a technical problem that needs to be solved urgently. Summary of the invention

[0004] The present invention aims to solve at least one of the technical problems existing in the prior art.

[0005] To this end, a first aspect of the present invention proposes a robot fault detection method.

[0006] A second aspect of the present invention provides a fault detection device for a robot.

[0007] A third aspect of the present invention provides a fault detection device for a robot.

[0008] A fourth aspect of the present invention provides a readable storage medium.

[0009] A fifth aspect of the present invention provides a robot.

[0010] In view of this, the first aspect of the present invention provides a robot fault detection method, which includes: establishing a joint simulation model of the robot; obtaining fault data of the robot's joints in a fault state; determining an actual noise spectrum based on the fault data; entering the fault data into the joint simulation model, and obtaining a simulated noise spectrum through the joint simulation model; and determining target fault information based on the actual noise spectrum and the simulated noise spectrum.

[0011] The present application proposes a fault detection method, which is used in conjunction with a robot. The robot includes multiple joints, and the fault detection method is used to detect fault information occurring in the joints.

[0012] Specifically, the fault detection method is as follows: first, a joint simulation model of the robot is established according to the structural information of the robot. The joint simulation model can simulate the working state of the joint through a mathematical formula, and specifically, the working noise of the joint can be simulated by entering the working data. When a joint of the robot fails, the fault data of the robot running in the fault state is collected, and the corresponding actual noise spectrum is extracted from the fault data. Thereafter, the fault data is entered into the joint simulation model to obtain a simulated noise spectrum through the joint simulation model, and the simulated noise spectrum corresponds to the noise generated when common possible faults occur in the joint. Finally, the actual noise spectrum and the simulated noise spectrum are compared to analyze the target fault information matched by the actual noise spectrum through the comparison relationship and the fault simulation model, thereby completing the fault detection of the robot joint.

[0013] In the related art, there is a technical solution for deriving joint failure problems by directly analyzing noise. However, the robot's operating conditions are relatively complex, and the structure is complex and has many parts, so that the causes of abnormal noises are complex. Direct analysis of noise cannot effectively associate the noise analysis results with the working state of the robot hardware, so that the final analysis results are biased. In this regard, the present application introduces a joint simulation model so that the robot's joint failure analysis process can be effectively associated with the working state of the robot hardware through the joint simulation model. The link of importing fault data into the joint simulation model to simulate the simulated noise spectrum is to screen the cause of the fault in accordance with the robot's hardware working state. Finally, by analyzing the matching degree of the actual noise spectrum and the simulated noise spectrum, the actual fault type can be determined from the possible fault types under the working condition, and the corresponding fault information can be obtained.

[0014] For example, when the robot's joints may have faults A and B, and the actual working conditions of joints one and two are different and the fault noises are similar, it is difficult to accurately distinguish faults A and B by directly analyzing the noise, and the probability of misjudgment is high. In the technical solution defined in the present application, the joint simulation model can effectively associate the actual working conditions of the joints with the simulated noise spectrum through the collected joint data, thereby pre-excluding faults A or B that do not match the fault data, and then accurately analyzing the fault information that matches the actual working conditions by comparing the actual noise spectrum with the simulated noise spectrum.

[0015] It can be seen that this application solves the technical problems existing in related technologies, such as the failure detection results are out of touch with the actual working conditions and the failure location accuracy is poor, by introducing the joint simulation model. This has achieved the technical effect of optimizing the fault detection method, improving the fault detection accuracy and reliability, and reducing the difficulty of robot maintenance.

[0016] In addition, the robot fault detection method provided by the present invention may also have the following additional technical features:

[0017] In the above technical scheme, the joint simulation model includes multiple different fault modes and fault information corresponding to each fault mode. The step of obtaining a simulated noise spectrum through the joint simulation model includes: controlling the joint simulation model to operate in a first fault mode to obtain a simulated noise spectrum corresponding to the first fault mode.

[0018] In this technical solution, the joint simulation model includes a variety of different common fault modes, each of which is preset with its corresponding fault information. After a joint operation failure occurs and the fault data is entered into the joint simulation model, the joint simulation model selects a fault mode from multiple fault modes as the first fault mode, and controls the joint simulation model to operate with the fault data in the first fault mode, thereby obtaining a simulated noise spectrum corresponding to the fault data and the first fault mode. Thereafter, if the simulated noise spectrum and the actual noise spectrum have a high degree of match, it means that the current fault of the joint corresponds to the first fault mode, otherwise it means that the current fault corresponds to other fault modes other than the first fault mode.

[0019] By presetting the fault mode, on the one hand, the fault detection range of the joint simulation model can cover the types of scenario faults; on the other hand, it is beneficial to improve the precision and accuracy of joint fault analysis and reduce the possibility of misjudgment or missed judgment of joint faults.

[0020] In any of the above technical solutions, the step of determining fault information based on the actual noise spectrum and the simulated noise spectrum includes: determining the similarity between the actual noise spectrum and the simulated noise spectrum; based on the similarity being greater than or equal to a threshold, taking the fault information corresponding to the first fault mode as the target fault information; based on the similarity being less than a threshold, controlling the joint simulation model to switch to operate in the second fault mode.

[0021] In this technical solution, the step of determining fault information based on the actual noise spectrum and the simulated noise spectrum is described in detail. Specifically, after the joint simulation model outputs the simulated noise spectrum, the similarity between the actual noise spectrum and the simulated noise spectrum is analyzed, wherein the actual noise spectrum includes a first time domain feature and a first frequency domain feature, and the simulated noise spectrum includes a second time domain feature and a second spectrum feature, and the time domain matching degree of the first time domain feature and the second time domain feature, as well as the frequency domain matching degree of the first frequency domain feature and the second frequency domain feature are determined respectively, and then the similarity is obtained by combining the time domain matching degree and the frequency domain matching degree.

[0022] After determining the similarity, compare the size relationship between the similarity and the preset threshold. If the similarity is greater than or equal to the threshold, it proves that the simulated noise spectrum generated in the first fault mode matches the actual noise spectrum, and the current fault of the joint corresponds to the first fault mode. Then the fault information corresponding to the first fault mode is used as the target joint information to push it to the client side, providing convenient conditions for the customer to troubleshoot the fault. On the contrary, if the similarity is less than the threshold, it proves that the simulated noise spectrum generated in the first fault mode does not match the actual noise spectrum, and then the joint simulation model is controlled to switch to the second fault mode, and the cycle is repeated to check the multiple fault modes preset in the joint simulation mode one by one.

[0023] This control process realizes the one-by-one troubleshooting of fault types. Compared with the technical solution of batch troubleshooting, one-by-one troubleshooting is not affected by other fault data, and its accuracy and reliability are relatively high. At the same time, the one-by-one troubleshooting process can complete the judgment task when the judgment result is a match, which is conducive to reducing the data processing burden of the system, thereby improving the practicality of the fault detection method.

[0024] Among them, the target fault information includes the location of the fault, the cause of the fault, the fault type, the fault level and other information. For this, the technical solution does not impose rigid restrictions on the specific components of the target fault information, and can provide convenient conditions for users to troubleshoot faults.

[0025] In any of the above technical solutions, the steps of controlling the joint simulation model to operate in a first fault mode and obtaining a simulated noise spectrum corresponding to the first fault mode include: determining the characteristic frequency of the component corresponding to the first fault mode; determining the fault frequency based on the fault data; and modulating the characteristic frequency by the fault frequency to obtain a simulated noise spectrum.

[0026] In this technical solution, the process of generating a simulated noise spectrum by a joint simulation model is described. Specifically, after controlling the joint simulation model to operate in a first fault mode, the characteristic frequency of the component corresponding to the first fault mode is determined, and the corresponding fault frequency is determined according to the fault data. Then, the fault frequency is used as a carrier signal and the fault frequency is used as a modulation signal to modulate and generate a simulated noise spectrum.

[0027] By introducing the fault frequency modulation process, the fault analysis process can be effectively linked to the actual working conditions of the joint, ensuring that the joint fault detection process does not deviate from the actual working conditions, thereby achieving the technical effect of improving the accuracy and reliability of the fault detection method.

[0028] Specifically, after obtaining the simulated noise spectrum, the simulated noise spectrum can be synthesized into a sound signal, and the sound signal can be pushed to the user side. The user can conduct a preliminary investigation of joint faults by comparing the sound signal with the actual noise of the joint, thereby improving the practicality of the fault detection method and reducing the difficulty of joint maintenance.

[0029] In any of the above technical solutions, the fault data includes: static data, dynamic data and load data of the joint.

[0030] In this technical solution, the fault data includes static data, which includes the posture data of each component, the position data of the joint, etc. The fault data also includes dynamic data, which includes the speed of the joint, the rotation speed of each component, etc. The fault data also includes the load data of the joint.

[0031] By introducing the above data types, the simulation accuracy of the joint simulation model can be improved, the error of the simulated noise spectrum output by the joint simulation mode can be reduced, and the fault detection process can be effectively linked to the actual working condition data, thereby achieving the technical effect of improving the fault detection accuracy and reliability.

[0032] In any of the above technical solutions, the steps of establishing a joint simulation model of the robot include: establishing a spectrum model based on the component type information and component connection information of the joint; obtaining the vibration data and noise spectrum data of the joint in a non-fault state; calibrating the spectrum model based on the vibration data and noise spectrum data to obtain the joint simulation model.

[0033] In this technical solution, the steps of establishing a joint simulation model are described. Specifically, first obtain the component type information and component connection information of each joint in the robot, different joints correspond to different component type information and different component connection information, then establish the speed order relationship between the components according to the component connection relationship, and determine the corresponding spectrum model according to the speed order relationship. Afterwards, obtain the vibration data and noise spectrum data of the joint when it runs at a constant speed in a non-fault state, and calibrate the spectrum order and amplitude in the spectrum model through the vibration data and noise spectrum data to obtain a joint simulation model that can accurately simulate the joint working condition.

[0034] By introducing component information, the fault detection method can be closely linked to the inherent structure of the joint to ensure that the analysis process matches the joint properties. By introducing the steady-state working data of the joint to calibrate the spectrum model, the errors caused by external factors can be eliminated. For example, the joint simulation model cannot take into account the environmental characteristics such as temperature and humidity in the scene where the joint is located. The introduction of steady-state data can reduce the accuracy error caused by such external factors, so that the simulated noise spectrum output by the joint simulation model can be close to the actual working conditions. This will achieve the technical effect of improving the accuracy of joint simulation and improving the practicality and reliability of the fault detection method.

[0035] Specifically, the robot has multiple different types of joints, so each robot contains multiple joint simulation models. When a robot fails, the extracted fault data includes the joint model data corresponding to this part of the data, so that the joint fault model to be used can be screened out from multiple joint simulation models through the joint model data.

[0036] In any of the above technical solutions, the fault detection method also includes: obtaining common fault data of the joint; importing the common fault data into the joint simulation model to generate multiple fault modes.

[0037] In this technical solution, after the joint simulation model is initially established based on the component type information, component connection information and steady-state data, the common fault data of the joint is obtained. The common fault data is the historical data of the joint of the same model, which corresponds to various types of faults commonly occurring in the joint of the model. The joint simulation model establishes a corresponding fault model based on the common fault data to generate a fault mode corresponding to the common fault type in the joint simulation model.

[0038] By introducing common fault data to generate fault modes, the fault detection range of the joint simulation model can effectively cover common fault problems of the joint, thereby reducing the possibility of misjudgment and missed judgment of the fault detection method and improving the practicality of the fault detection method.

[0039] In any of the above technical solutions, the fault data includes first noise data.

[0040] In this technical solution, when a joint fails, the abnormal sound emitted by the joint is collected to obtain first noise data, and then the corresponding actual noise spectrum is obtained through the first noise data to facilitate fault detection by comparison with the simulated noise spectrum.

[0041] In any of the above technical solutions, the step of determining the actual noise spectrum based on the fault data includes: filtering the first noise data to obtain the second noise data; extracting the time domain characteristics and frequency domain characteristics in the second noise data; and determining the actual noise spectrum based on the time domain characteristics and the frequency domain characteristics.

[0042] In this technical solution, in the process of determining the actual noise spectrum according to the fault data, the first noise data is first subjected to bandpass filtering to eliminate the noise in the first noise data to obtain the second noise data. Then, the time domain features and frequency domain features in the second noise data are extracted, and the actual noise spectrum is determined by the time domain features and frequency domain features, wherein the extracted time domain features correspond to the aforementioned first time domain features, and the extracted frequency domain features correspond to the aforementioned first frequency domain features, thereby completing the preprocessing of the first noise data to facilitate the detection of faults in conjunction with the simulated noise spectrum.

[0043] This preprocessing process can remove the interference components in the first noise data, reduce the data processing burden in the fault detection process by extracting the main features, and improve the fault detection efficiency.

[0044] In any of the above technical solutions, the fault data includes model data of the joint.

[0045] In this technical solution, the fault data also includes joint model data. Different models of joints have different types, quantities and connection relationships of components. The joint model data can be used to effectively distinguish different types of joints, reducing the possibility of misusing the joint simulation model.

[0046] In any of the above technical solutions, the robot includes multiple joint simulation models, and the step of entering fault data into the joint simulation model includes: determining a target joint simulation model from multiple joint simulation models according to the model data; and entering the fault data into the target joint simulation model.

[0047] In this technical solution, the robot has multiple different types of joints, so each robot includes multiple joint simulation models. When a robot fails, the extracted fault data includes the joint model data corresponding to this part of the data, so as to screen out the corresponding target joint simulation model from multiple joint simulation models through the joint model data, and enter this part of the fault data into the screened target joint simulation model. This avoids the problem of misuse of joint simulation models and improves the accuracy and reliability of the fault detection method.

[0048] The second aspect of the present invention provides a robot fault detection device, which includes: an establishment module for establishing a joint simulation model of the robot; an acquisition module for acquiring fault data of the robot's joints in a faulty state; a first determination module for determining an actual noise spectrum based on the fault data; a simulation module for entering the fault data into the joint simulation model and obtaining a simulated noise spectrum through the joint simulation model; and a second determination module for determining target fault information based on the actual noise spectrum and the simulated noise spectrum.

[0049] The present application proposes a fault detection device, which is used in conjunction with a robot. The robot includes multiple joints. The fault detection device is used to detect fault information occurring in the joints. The fault detection device includes an establishment module, an acquisition module, a first determination module, a simulation module, and a second determination module.

[0050] Specifically, the establishment module first establishes a joint simulation model of the robot according to the structural information of the robot. The joint simulation model can simulate the working state of the joint through a mathematical formula, and specifically simulate the working noise of the joint by entering the working data. When a fault occurs in the joint of the robot, the acquisition module collects the fault data of the robot running in the fault state, and the first determination module extracts the corresponding actual noise spectrum from the fault data. Thereafter, the simulation module enters the fault data into the joint simulation model to obtain a simulated noise spectrum through the joint simulation model, and the simulated noise spectrum corresponds to the noise generated when common possible faults occur in the joint. Finally, the second determination module compares the actual noise spectrum with the simulated noise spectrum, and analyzes the target fault information matched by the actual noise spectrum through the comparison relationship and the fault simulation model, thereby completing the fault detection of the robot joint.

[0051] In the related art, there is a technical solution for deriving joint failure problems by directly analyzing noise. However, the robot's operating conditions are relatively complex, and the structure is complex and has many parts, so that the causes of abnormal noises are complex. Direct analysis of noise cannot effectively associate the noise analysis results with the working state of the robot hardware, so that the final analysis results are biased. In this regard, the present application introduces a joint simulation model so that the robot's joint failure analysis process can be effectively associated with the working state of the robot hardware through the joint simulation model. The link of importing fault data into the joint simulation model to simulate the simulated noise spectrum is to screen the cause of the fault in accordance with the robot's hardware working state. Finally, by analyzing the matching degree of the actual noise spectrum and the simulated noise spectrum, the actual fault type can be determined from the possible fault types under the working condition, and the corresponding fault information can be obtained.

[0052] For example, when the robot's joints may have faults A and B, and the actual working conditions of joints one and two are different and the fault noises are similar, it is difficult to accurately distinguish faults A and B by directly analyzing the noise, and the probability of misjudgment is high. In the technical solution defined in the present application, the joint simulation model can effectively associate the actual working conditions of the joints with the simulated noise spectrum through the collected joint data, thereby pre-excluding faults A or B that do not match the fault data, and then accurately analyze the fault information that matches the actual working conditions by comparing the actual noise spectrum with the simulated noise spectrum.

[0053] It can be seen that this application solves the technical problems existing in related technologies, such as the failure detection results are out of touch with the actual working conditions and the failure location accuracy is poor, by introducing the joint simulation model. This has achieved the technical effect of optimizing the fault detection device, improving the fault detection accuracy and reliability, and reducing the difficulty of robot maintenance.

[0054] A third aspect of the present invention provides a robot fault detection device, which includes: a memory on which programs or instructions are stored; and a processor configured to implement the steps of the detection method in any of the above technical solutions when executing the programs or instructions.

[0055] In this technical solution, a fault detection device is proposed, which includes a memory and a processor. The processor executes the program or instruction stored in the memory to implement the detection method in any of the above technical solutions. Therefore, the fault detection device has the advantages of the fault detection method in any of the above technical solutions, and can achieve the technical effect that can be achieved by the fault detection method in any of the above technical solutions. To avoid repetition, it will not be repeated here.

[0056] A fourth aspect of the present invention provides a readable storage medium having a program or instruction stored thereon, which implements the steps of the detection method in any of the above technical solutions when the program or instruction is executed by a processor.

[0057] In this technical solution, a readable storage medium is proposed, which stores a program or instruction. The program or instruction is executed by a processor to implement the steps of the detection method in any of the above technical solutions. Therefore, the readable storage medium has the advantages of the fault detection method in any of the above technical solutions, and can achieve the technical effects that can be achieved by the fault detection method in any of the above technical solutions. To avoid repetition, it will not be repeated here.

[0058] A fifth aspect of the present invention provides a robot, comprising: a detection device as in any of the above technical solutions; or a readable storage medium as in any of the above technical solutions.

[0059] In this technical solution, a robot including the detection device in any of the above technical solutions or the terminal readable storage medium in any of the above technical solutions is proposed, so the robot has the advantages of the detection device in any of the above technical solutions and can achieve the technical effects that can be achieved by the detection device in any of the above technical solutions, or the robot has the advantages of the readable storage medium in any of the above technical solutions and can achieve the technical effects that can be achieved by the readable storage medium in any of the above technical solutions. To avoid repetition, it will not be repeated here.

[0060] Additional aspects and advantages of the present invention will become apparent from the following description or may be learned by practice of the present invention. BRIEF DESCRIPTION OF THE DRAWINGS

[0061] The above and / or additional aspects and advantages of the present invention will become apparent and easily understood from the description of the embodiments in conjunction with the following drawings, in which:

[0062] Figure 1One of the flowcharts of the robot fault detection method according to one embodiment of the present invention is shown;

[0063] Figure 2 A second flow chart showing a method for detecting a robot fault according to an embodiment of the present invention;

[0064] Figure 3 A third flowchart of a method for detecting a fault of a robot according to an embodiment of the present invention is shown;

[0065] Figure 4 A fourth flowchart of a method for detecting a fault of a robot according to an embodiment of the present invention is shown;

[0066] Figure 5 A fifth flowchart of a method for detecting a fault of a robot according to an embodiment of the present invention is shown;

[0067] Figure 6 A sixth flow chart showing a method for detecting a fault of a robot according to an embodiment of the present invention;

[0068] Figure 7 A seventh flowchart of a method for detecting a fault of a robot according to an embodiment of the present invention is shown;

[0069] Figure 8 An eighth flowchart of a method for detecting a fault of a robot according to an embodiment of the present invention is shown;

[0070] Fig. 9 One of the structural block diagrams of a robot fault detection device according to an embodiment of the present invention is shown;

[0071] Fig.10 A second structural block diagram of a robot fault detection device according to an embodiment of the present invention is shown;

[0072] Fig.11 A block diagram showing the structure of a robot according to an embodiment of the present invention. DETAILED DESCRIPTION

[0073] In order to more clearly understand the above-mentioned purpose, features and advantages of the present invention, the present invention is further described in detail below in conjunction with the accompanying drawings and specific embodiments. It should be noted that the embodiments of the present application and the features in the embodiments can be combined with each other without conflict.

[0074] In the following description, many specific details are set forth to facilitate a full understanding of the present invention. However, the present invention may also be implemented in other ways different from those described herein. Therefore, the protection scope of the present invention is not limited to the specific embodiments disclosed below.

[0075] Refer to the following Figures 1 to 11 A robot fault detection method, device, storage medium and robot according to some embodiments of the present invention are described.

[0076] like Figure 1 As shown, one embodiment of the present invention provides a robot fault detection method, the fault detection method comprising:

[0077] Step 102, establishing a joint simulation model of the robot;

[0078] Step 104, obtaining fault data of the joints of the robot in a fault state;

[0079] Step 106, determining an actual noise spectrum according to the fault data;

[0080] Step 108, inputting the fault data into the joint simulation model, and obtaining a simulated noise spectrum through the joint simulation model;

[0081] Step 110: determining target fault information according to the actual noise spectrum and the simulated noise spectrum.

[0082] The present application proposes a fault detection method, which is used in conjunction with a robot. The robot includes multiple joints, and the fault detection method is used to detect fault information occurring in the joints.

[0083] Specifically, the fault detection method is as follows: first, a joint simulation model of the robot is established according to the structural information of the robot. The joint simulation model can simulate the working state of the joint through a mathematical formula, and specifically, the working noise of the joint can be simulated by entering the working data. When a joint of the robot fails, the fault data of the robot running in the fault state is collected, and the corresponding actual noise spectrum is extracted from the fault data. Thereafter, the fault data is entered into the joint simulation model to obtain a simulated noise spectrum through the joint simulation model, and the simulated noise spectrum corresponds to the noise generated when common possible faults occur in the joint. Finally, the actual noise spectrum and the simulated noise spectrum are compared to analyze the target fault information matched by the actual noise spectrum through the comparison relationship and the fault simulation model, thereby completing the fault detection of the robot joint.

[0084] In the related art, there are embodiments of deriving joint failure problems by directly analyzing noise. However, the operating conditions of the robot are relatively complex, and the structure is complex and has many parts, so that the causes of abnormal noises are complex. Direct analysis of noise cannot effectively associate the noise analysis results with the working state of the robot hardware, so that the final analysis results are biased. In this regard, the present application introduces a joint simulation model so that the joint failure analysis process of the robot can be effectively associated with the working state of the robot hardware through the joint simulation model. The link of importing fault data into the joint simulation model to simulate the simulated noise spectrum is to screen the cause of the fault in accordance with the working state of the robot's hardware. Finally, by analyzing the matching degree of the actual noise spectrum and the simulated noise spectrum, the actual fault type can be determined from the possible fault types under the working condition, and the corresponding fault information can be obtained.

[0085] For example, when the robot's joints may have fault A and fault B, and the actual working conditions of joints one and two are different and the fault noises are similar, it is difficult to accurately distinguish fault A and fault B by directly analyzing the noise, and the probability of misjudgment is high. In the embodiments defined in the present application, the joint simulation model can effectively associate the actual working conditions of the joints with the simulated noise spectrum through the collected joint data, thereby pre-excluding fault A or fault B that does not match the fault data, and then accurately analyzing the fault information that matches the actual working conditions by comparing the actual noise spectrum with the simulated noise spectrum.

[0086] It can be seen that this application solves the technical problems existing in related technologies, such as the failure detection results are out of touch with the actual working conditions and the failure location accuracy is poor, by introducing the joint simulation model. This has achieved the technical effect of optimizing the fault detection method, improving the fault detection accuracy and reliability, and reducing the difficulty of robot maintenance.

[0087] In the above embodiment, the joint simulation model includes multiple different fault modes and fault information corresponding to each fault mode. The step of obtaining a simulated noise spectrum through the joint simulation model includes: controlling the joint simulation model to operate in a first fault mode to obtain a simulated noise spectrum corresponding to the first fault mode.

[0088] In this embodiment, the joint simulation model includes a variety of different common fault modes, each of which is preset with its corresponding fault information. After a joint operation failure occurs and the fault data is entered into the joint simulation model, the joint simulation model selects a fault mode from multiple fault modes as the first fault mode, and controls the joint simulation model to operate with the fault data in the first fault mode, thereby obtaining a simulated noise spectrum corresponding to the fault data and the first fault mode. Thereafter, if the simulated noise spectrum and the actual noise spectrum have a high degree of match, it means that the current fault of the joint corresponds to the first fault mode, otherwise it means that the current fault corresponds to other fault modes other than the first fault mode.

[0089] By presetting the fault mode, on the one hand, the fault detection range of the joint simulation model can cover the types of scenario faults; on the other hand, it is beneficial to improve the precision and accuracy of joint fault analysis and reduce the possibility of misjudgment or missed judgment of joint faults.

[0090] like Figure 2 As shown, in any of the above embodiments, the step of determining fault information according to the actual noise spectrum and the simulated noise spectrum includes:

[0091] Step 202, determining the similarity between the actual noise spectrum and the simulated noise spectrum;

[0092] Step 204: based on the similarity being greater than or equal to a threshold, taking the fault information corresponding to the first fault mode as target fault information;

[0093] Step 206 , based on the similarity being less than a threshold, controlling the joint simulation model to switch to operate in a second fault mode.

[0094] In this embodiment, the step of determining fault information based on the actual noise spectrum and the simulated noise spectrum is described in detail. Specifically, after the joint simulation model outputs the simulated noise spectrum, the similarity between the actual noise spectrum and the simulated noise spectrum is analyzed, wherein the actual noise spectrum includes a first time domain feature and a first frequency domain feature, and the simulated noise spectrum includes a second time domain feature and a second spectrum feature, and the time domain matching degree of the first time domain feature and the second time domain feature, as well as the frequency domain matching degree of the first frequency domain feature and the second frequency domain feature are determined respectively, and then the similarity is obtained by combining the time domain matching degree and the frequency domain matching degree.

[0095] After determining the similarity, compare the size relationship between the similarity and the preset threshold. If the similarity is greater than or equal to the threshold, it proves that the simulated noise spectrum generated in the first fault mode matches the actual noise spectrum, and the current fault of the joint corresponds to the first fault mode. Then the fault information corresponding to the first fault mode is used as the target joint information to push it to the client side, providing convenient conditions for the customer to troubleshoot the fault. On the contrary, if the similarity is less than the threshold, it proves that the simulated noise spectrum generated in the first fault mode does not match the actual noise spectrum, and then the joint simulation model is controlled to switch to the second fault mode, and the cycle is repeated to check the multiple fault modes preset in the joint simulation mode one by one.

[0096] This control process realizes the one-by-one troubleshooting of fault types. Compared with the batch troubleshooting implementation, the one-by-one troubleshooting is not affected by other fault data, and the accuracy and reliability are relatively high. At the same time, the one-by-one troubleshooting process can complete the judgment task when the judgment result is a match, which is conducive to reducing the data processing burden of the system, thereby improving the practicality of the fault detection method.

[0097] The target fault information includes the location of the fault, the cause of the fault, the type of fault, the level of the fault, and the like. This embodiment does not impose a rigid limit on the specific components of the target fault information, and only needs to provide convenient conditions for users to troubleshoot the fault.

[0098] like Figure 3 As shown, in any of the above embodiments, the step of controlling the joint simulation model to operate in the first fault mode and obtaining a simulated noise spectrum corresponding to the first fault mode includes:

[0099] Step 302, determining a component characteristic frequency corresponding to a first failure mode;

[0100] Step 304, determining the fault frequency according to the fault data;

[0101] Step 306: modulate the characteristic frequency by the fault frequency to obtain a simulated noise spectrum.

[0102] In this embodiment, the process of generating a simulated noise spectrum by a joint simulation model is described. Specifically, after controlling the joint simulation model to operate in a first fault mode, the characteristic frequency of the component corresponding to the first fault mode is determined, and the corresponding fault frequency is determined according to the fault data. Then, the fault frequency is used as a carrier signal and the fault frequency is used as a modulation signal to modulate and generate a simulated noise spectrum.

[0103] By introducing the fault frequency modulation process, the fault analysis process can be effectively linked to the actual working conditions of the joint, ensuring that the joint fault detection process does not deviate from the actual working conditions, thereby achieving the technical effect of improving the accuracy and reliability of the fault detection method.

[0104] Specifically, after obtaining the simulated noise spectrum, the simulated noise spectrum can be synthesized into a sound signal, and the sound signal can be pushed to the user side. The user can conduct a preliminary investigation of joint faults by comparing the sound signal with the actual noise of the joint, thereby improving the practicality of the fault detection method and reducing the difficulty of joint maintenance.

[0105] In any of the above embodiments, the fault data includes: static data, dynamic data and load data of the joint.

[0106] In this embodiment, the fault data includes static data, which includes posture data of each component, position data of joints, etc. The fault data also includes dynamic data, which includes the speed of the joint, the rotation speed of each component, etc. The fault data also includes load data of the joint.

[0107] By introducing the above data types, the simulation accuracy of the joint simulation model can be improved, the error of the simulated noise spectrum output by the joint simulation mode can be reduced, and the fault detection process can be effectively linked to the actual working condition data, thereby achieving the technical effect of improving the fault detection accuracy and reliability.

[0108] like Figure 4 As shown, in any of the above embodiments, the step of establishing a joint simulation model of the robot includes:

[0109] Step 402, establishing a spectrum model according to the component type information and component connection information of the joint;

[0110] Step 404, obtaining vibration data and noise spectrum data of the joint in a non-fault state;

[0111] Step 406, calibrating the spectrum model according to the vibration data and the noise spectrum data to obtain a joint simulation model.

[0112] In this embodiment, the steps of establishing a joint simulation model are described. Specifically, first obtain the component type information and component connection information of each joint in the robot, different joints correspond to different component type information and different component connection information, then establish the speed order relationship between the components according to the component connection relationship, and determine the corresponding spectrum model according to the speed order relationship. Afterwards, obtain the vibration data and noise spectrum data of the joint when it runs at a constant speed in a non-fault state, and calibrate the spectrum order and amplitude in the spectrum model through the vibration data and noise spectrum data to obtain a joint simulation model that can accurately simulate the joint working condition.

[0113] By introducing component information, the fault detection method can be closely linked to the inherent structure of the joint to ensure that the analysis process matches the joint properties. By introducing the steady-state working data of the joint to calibrate the spectrum model, the errors caused by external factors can be eliminated. For example, the joint simulation model cannot take into account the environmental characteristics such as temperature and humidity in the scene where the joint is located. The introduction of steady-state data can reduce the accuracy error caused by such external factors, so that the simulated noise spectrum output by the joint simulation model can be close to the actual working conditions. This will achieve the technical effect of improving the accuracy of joint simulation and improving the practicality and reliability of the fault detection method.

[0114] Specifically, the robot has multiple different types of joints, so each robot contains multiple joint simulation models. When a robot fails, the extracted fault data includes the joint model data corresponding to this part of the data, so that the joint fault model to be used can be screened out from multiple joint simulation models through the joint model data.

[0115] like Figure 5 As shown, in any of the above embodiments, the fault detection method further includes:

[0116] Step 502, obtaining common fault data of the joint;

[0117] Step 504, importing common fault data into the joint simulation model to generate multiple fault modes.

[0118] In this embodiment, after the joint simulation model is initially established based on the component type information, component connection information and steady-state data, the common fault data of the joint is obtained. The common fault data is the historical data of the joint of the same model, which corresponds to various types of faults commonly occurring in the joint of the model. The joint simulation model establishes a corresponding fault model based on the common fault data to generate a fault mode corresponding to the common fault type in the joint simulation model.

[0119] By introducing common fault data to generate fault modes, the fault detection range of the joint simulation model can effectively cover common fault problems of the joint, thereby reducing the possibility of misjudgment and missed judgment of the fault detection method and improving the practicality of the fault detection method.

[0120] In any of the above embodiments, the fault data includes first noise data.

[0121] In this embodiment, when a joint fails, the abnormal sound emitted by the joint is collected to obtain first noise data, and then the corresponding actual noise spectrum is obtained through the first noise data to facilitate comparison with the simulated noise spectrum to detect the failure.

[0122] like Figure 6 As shown, in any of the above embodiments, the step of determining the actual noise spectrum according to the fault data includes:

[0123] Step 602, filtering the first noise data to obtain second noise data;

[0124] Step 604: time domain features and frequency domain features in the second noise data; determine the actual noise spectrum according to the time domain features and frequency domain features.

[0125] In this embodiment, in the process of determining the actual noise spectrum according to the fault data, the first noise data is first subjected to bandpass filtering to eliminate the noise in the first noise data to obtain the second noise data. Then, the time domain features and frequency domain features in the second noise data are extracted, and the actual noise spectrum is determined by the time domain features and frequency domain features, wherein the extracted time domain features correspond to the aforementioned first time domain features, and the extracted frequency domain features correspond to the aforementioned first frequency domain features, thereby completing the preprocessing of the first noise data to facilitate the detection of faults in conjunction with the simulated noise spectrum.

[0126] This preprocessing process can remove the interference components in the first noise data, reduce the data processing burden in the fault detection process by extracting the main features, and improve the fault detection efficiency.

[0127] In any of the above embodiments, the fault data includes model data of the joint.

[0128] In this embodiment, the fault data also includes joint model data. There are differences in the types, quantities and connection relationships of components in joints of different models. The joint model data can be used to effectively distinguish different types of joints, reducing the possibility of misusing the joint simulation model.

[0129] like Figure 7 As shown, in any of the above embodiments, the robot includes a plurality of joint simulation models, and the step of entering the fault data into the joint simulation models includes:

[0130] Step 702, determining a target joint simulation model from a plurality of joint simulation models according to the model data;

[0131] Step 704, entering the fault data into the target joint simulation model.

[0132] In this embodiment, the robot has multiple different types of joints, so each robot includes multiple joint simulation models. When a robot fails, the extracted fault data includes the joint model data corresponding to this part of the data, so that the corresponding target joint simulation model is screened out from multiple joint simulation models through the joint model data, and this part of the fault data is entered into the screened target joint simulation model. This avoids the problem of misuse of joint simulation models and improves the accuracy and reliability of the fault detection method.

[0133] like Figure 8 As shown, in a specific embodiment of the present invention, the fault detection process of the robot is as follows:

[0134] Step 802, establishing a spectrum model according to the component type information and component connection information of the joint;

[0135] Step 804, obtaining vibration data and noise spectrum data of the joint in a non-fault state;

[0136] Step 806, calibrating the spectrum model according to the vibration data and the noise spectrum data to obtain a joint simulation model;

[0137] Step 808, obtaining fault data of the joints of the robot in a fault state;

[0138] Step 810, determining a component characteristic frequency corresponding to a first failure mode;

[0139] Step 812, determining the fault frequency according to the fault data;

[0140] Step 814, modulating the characteristic frequency by the fault frequency to obtain a simulated noise spectrum;

[0141] Step 816, obtaining common fault data of the joint;

[0142] Step 818, importing common fault data into the joint simulation model to generate multiple fault modes;

[0143] Step 820, inputting the fault data into the joint simulation model;

[0144] Step 822, determining a target joint simulation model from a plurality of joint simulation models according to the model data;

[0145] Step 824, controlling the joint simulation model to operate in a first fault mode to obtain a simulated noise spectrum corresponding to the first fault mode;

[0146] Step 826, determining the similarity between the actual noise spectrum and the simulated noise spectrum;

[0147] Step 828, determine whether the similarity is greater than or equal to a threshold, if the determination result is yes, execute step 832, if the determination result is no, execute step 830;

[0148] Step 832, taking the fault information corresponding to the first fault mode as target fault information;

[0149] Step 830, controlling the joint simulation model to switch to operate in the second fault mode.

[0150] like Fig. 9As shown, the second aspect of the present invention provides a robot fault detection device 900, and the robot fault detection device 900 includes: an establishment module 902, which is used to establish a joint simulation model of the robot; an acquisition module 904, which is used to obtain fault data of the robot's joints in a fault state; a first determination module 906, which determines the actual noise spectrum according to the fault data; a simulation module 908, which is used to enter the fault data into the joint simulation model, and obtain a simulated noise spectrum through the joint simulation model; a second determination module 910, which is used to determine the target fault information according to the actual noise spectrum and the simulated noise spectrum.

[0151] The present application proposes a robot fault detection device 900, which is used in conjunction with a robot. The robot includes multiple joints. The robot fault detection device 900 is used to detect fault information occurring in the joints. The robot fault detection device 900 includes an establishment module 902, an acquisition module 904, a first determination module 906, a simulation module 908 and a second determination module 910.

[0152] Specifically, the establishment module 902 first establishes a joint simulation model of the robot according to the structural information of the robot. The joint simulation model can simulate the working state of the joint through a mathematical formula, and specifically simulate the working noise of the joint by entering the working data. When a joint of the robot fails, the acquisition module collects the fault data of the robot running in the fault state, and the first determination module 906 extracts the corresponding actual noise spectrum from the fault data. Thereafter, the simulation module 908 enters the fault data into the joint simulation model to obtain a simulated noise spectrum through the joint simulation model, and the simulated noise spectrum corresponds to the noise generated when common possible faults occur in the joint. Finally, the second determination module 910 compares the actual noise spectrum with the simulated noise spectrum, so as to analyze the target fault information matched by the actual noise spectrum through the comparison relationship in conjunction with the fault simulation model, thereby completing the fault detection of the robot joint.

[0153] In the related art, there are embodiments of deriving joint failure problems by directly analyzing noise. However, the operating conditions of the robot are relatively complex, and the structure is complex and has many parts, so that the causes of abnormal noises are complex. Direct analysis of noise cannot effectively associate the noise analysis results with the working state of the robot hardware, so that the final analysis results are biased. In this regard, the present application introduces a joint simulation model so that the joint failure analysis process of the robot can be effectively associated with the working state of the robot hardware through the joint simulation model. The link of importing fault data into the joint simulation model to simulate the simulated noise spectrum is to screen the cause of the fault in accordance with the working state of the robot's hardware. Finally, by analyzing the matching degree of the actual noise spectrum and the simulated noise spectrum, the actual fault type can be determined from the possible fault types under the working condition, and the corresponding fault information can be obtained.

[0154] For example, when the robot's joints may have fault A and fault B, and the actual working conditions of joints one and two are different and the fault noises are similar, it is difficult to accurately distinguish fault A and fault B by directly analyzing the noise, and the probability of misjudgment is high. In the embodiments defined in the present application, the joint simulation model can effectively associate the actual working conditions of the joints with the simulated noise spectrum through the collected joint data, thereby pre-excluding fault A or fault B that does not match the fault data, and then accurately analyzing the fault information that matches the actual working conditions by comparing the actual noise spectrum with the simulated noise spectrum.

[0155] It can be seen that, by introducing the joint simulation model, this application solves the technical problems existing in the related art that the fault detection results are out of touch with the actual working conditions and the fault location accuracy is poor. This further achieves the technical effect of optimizing the robot fault detection device 900, improving the fault detection accuracy and reliability, and reducing the difficulty of robot maintenance.

[0156] In any of the above embodiments, the joint simulation model includes multiple different fault modes and fault information corresponding to each fault mode. The simulation module 908 also includes: controlling the joint simulation model to operate in a first fault mode to obtain a simulated noise spectrum corresponding to the first fault mode.

[0157] In any of the above embodiments, the second determination module 910 specifically includes: determining the similarity between the actual noise spectrum and the simulated noise spectrum; based on the similarity being greater than or equal to a threshold, taking the fault information corresponding to the first fault mode as the target fault information; based on the similarity being less than a threshold, controlling the joint simulation model to switch to operate in the second fault mode.

[0158] In any of the above embodiments, the second determination module 910 further includes: determining a component characteristic frequency corresponding to the first fault mode; determining the fault frequency according to the fault data; and obtaining a simulated noise spectrum by modulating the characteristic frequency by the fault frequency.

[0159] In any of the above embodiments, establishing module 902 specifically includes: establishing a spectrum model based on the component type information and component connection information of the joint; obtaining the vibration data and noise spectrum data of the joint in a non-fault state; calibrating the spectrum model based on the vibration data and noise spectrum data to obtain a joint simulation model.

[0160] In any of the above embodiments, establishing module 902 also includes: acquiring common fault data of the joint; importing the common fault data into the joint simulation model to generate multiple fault modes.

[0161] In any of the above embodiments, the fault data includes first noise data, and the first determination module 906 specifically includes: filtering the first noise data to obtain second noise data; time domain features and frequency domain features in the second noise data; and determining the actual noise spectrum based on the time domain features and frequency domain features.

[0162] In any of the above embodiments, the fault data includes model data of the joint, and the simulation module 908 further includes: determining a target joint simulation model from a plurality of joint simulation models according to the model data; and entering the fault data into the target joint simulation model.

[0163] like Fig.10 As shown, the third aspect of the present invention provides a robot fault detection device 1000, which includes: a memory 1002 on which programs or instructions are stored; a processor 1004, which is configured to implement the steps of the detection method in any of the above embodiments when executing the program or instructions.

[0164] In this embodiment, a fault detection device is proposed, which includes a memory 1002 and a processor 1004. The processor 1004 can implement the detection method in any of the above embodiments by executing the program or instruction stored in the memory 1002. Therefore, the fault detection device has the advantages of the fault detection method in any of the above embodiments, and can achieve the technical effects that can be achieved by the fault detection method in any of the above embodiments. To avoid repetition, it will not be described here.

[0165] A fourth aspect of the present invention provides a readable storage medium having a program or instruction stored thereon, which implements the steps of the detection method in any of the above embodiments when the program or instruction is executed by a processor.

[0166] In this embodiment, a readable storage medium is proposed, which stores a program or instruction. The program or instruction is executed by a processor to implement the steps of the detection method in any of the above embodiments. Therefore, the readable storage medium has the advantages of the fault detection method in any of the above embodiments, and can achieve the technical effects that can be achieved by the fault detection method in any of the above embodiments. To avoid repetition, it will not be described here.

[0167] like Fig.11 As shown, the fifth aspect of the present invention provides a robot 1100, and the robot 1100 includes: a robot fault detection device 900 as in any of the above embodiments; or a readable storage medium 1120 as in any of the above embodiments.

[0168] In this embodiment, a robot 1100 including the robot fault detection device 900 in any of the above embodiments or the readable storage medium 1120 in any of the above embodiments is proposed, so the robot 1100 has the advantages of the robot fault detection device 900 in any of the above embodiments and can achieve the technical effects that can be achieved by the robot fault detection device 900 in any of the above embodiments, or the robot 1100 has the advantages of the readable storage medium 1120 in any of the above embodiments and can achieve the technical effects that can be achieved by the readable storage medium 1120 in any of the above embodiments. To avoid repetition, it will not be repeated here.

[0169] It should be clarified that in the claims, specification and drawings of the present invention, the term "multiple" refers to two or more than two. Unless otherwise clearly defined, the orientation or position relationship indicated by the terms "upper" and "lower" is based on the orientation or position relationship shown in the drawings, which is only for the purpose of more conveniently describing the present invention and making the description process simpler, rather than indicating or implying that the device or element referred to must have the specific orientation described, be constructed and operated in a specific orientation, so these descriptions cannot be understood as limiting the present invention; the terms "connect", "install", "fix" and the like should be understood in a broad sense. For example, "connection" can be a fixed connection between multiple objects, or a detachable connection between multiple objects, or an integral connection; it can be a direct connection between multiple objects, or an indirect connection between multiple objects through an intermediate medium. For ordinary technicians in this field, the specific meanings of the above terms in the present invention can be understood according to the specific circumstances of the above data.

[0170] In the claims, specification and drawings of the present invention, the description of the terms "one embodiment", "some embodiments", "specific embodiments" and the like means that the specific features, structures, materials or characteristics described in conjunction with the embodiment or example are included in at least one embodiment or example of the present invention. In the claims, specification and drawings of the present invention, the schematic representations of the above terms do not necessarily refer to the same embodiment or example. Moreover, the specific features, structures, materials or characteristics described may be combined in any one or more embodiments or examples in a suitable manner.

[0171] The above are only preferred embodiments of the present invention and are not intended to limit the present invention. For those skilled in the art, the present invention may have various modifications and variations. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present invention shall be included in the protection scope of the present invention.

Claims

1. A robot fault detection method, characterized in that: include: Establishing a joint simulation model of the robot; Acquiring fault data of the joints of the robot in a fault state; determining an actual noise spectrum according to the fault data; Entering the fault data into the joint simulation model, and obtaining a simulated noise spectrum through the joint simulation model; Determine target fault information according to the actual noise spectrum and the simulated noise spectrum; The step of establishing the joint simulation model of the robot comprises: Establishing a spectrum model according to the component type information and component connection information of the joint; Acquiring vibration data and noise spectrum data of the joint in a non-fault state; The spectrum model is calibrated according to the vibration data and the noise spectrum data to obtain the joint simulation model.

2. The fault detection method according to claim 1, characterized in that: The joint simulation model includes a plurality of different fault modes and fault information corresponding to each fault mode. The step of obtaining a simulated noise spectrum through the joint simulation model includes: The joint simulation model is controlled to operate in a first fault mode to obtain the simulated noise spectrum corresponding to the first fault mode.

3. The fault detection method according to claim 2, characterized in that: The step of determining target fault information according to the actual noise spectrum and the simulated noise spectrum comprises: Determining a similarity between the actual noise spectrum and the simulated noise spectrum; Based on the similarity being greater than or equal to a threshold, taking the fault information corresponding to the first fault mode as the target fault information; Based on the similarity being less than the threshold, the joint simulation model is controlled to switch to operate in a second fault mode.

4. The fault detection method according to claim 3, characterized in that: The step of controlling the joint simulation model to operate in a first fault mode to obtain the simulated noise spectrum corresponding to the first fault mode comprises: Determining a component characteristic frequency corresponding to the first failure mode; determining a fault frequency according to the fault data; The characteristic frequency is modulated by the fault frequency to obtain the simulated noise spectrum.

5. The fault detection method according to claim 4, characterized in that: The fault data includes: static data, dynamic data and load data of the joint.

6. The fault detection method according to claim 1, characterized in that: Also includes: Acquiring common fault data of the joint; The common fault data is imported into the joint simulation model to generate multiple fault modes.

7. The fault detection method according to any one of claims 1 to 6, characterized in that: The fault data includes first noise data.

8. The fault detection method according to claim 7, characterized in that: The step of determining the actual noise spectrum according to the fault data comprises: Performing filtering on the first noise data to obtain second noise data; Extracting time domain features and frequency domain features from the second noise data; The actual noise spectrum is determined according to the time domain characteristics and the frequency domain characteristics.

9. The fault detection method according to any one of claims 1 to 6, characterized in that: The fault data includes model data of the joint.

10. The fault detection method according to claim 9, characterized in that: The robot includes a plurality of joint simulation models, and the step of entering the fault data into the joint simulation models includes: Determining a target joint simulation model from among the plurality of joint simulation models according to the model data; The fault data is entered into the target joint simulation model.

11. A robot fault detection device, characterized in that: include: A building module, used to build a joint simulation model of the robot; An acquisition module, used for acquiring fault data of the joints of the robot in a fault state; A first determination module determines an actual noise spectrum according to the fault data; A simulation module, used for inputting the fault data into the joint simulation model, and obtaining a simulated noise spectrum through the joint simulation model; A second determination module, configured to determine target fault information according to the actual noise spectrum and the simulated noise spectrum; The establishment module specifically includes: establishing a spectrum model according to the component type information and component connection information of the joint; obtaining the vibration data and noise spectrum data of the joint in a non-fault state; The spectrum model is calibrated according to the vibration data and the noise spectrum data to obtain the joint simulation model.

12. A robot fault detection device, characterized in that: include: A memory on which programs or instructions are stored; A processor, configured to implement the steps of the detection method according to any one of claims 1 to 10 when executing the program or instruction.

13. A readable storage medium having a program or instruction stored thereon, characterized in that: When the program or instruction is executed by a processor, the steps of the detection method according to any one of claims 1 to 10 are implemented.

14. A robot, characterized in that: include: The detection device according to claim 11 or 12; or The readable storage medium as claimed in claim 13.

Images