An intelligent fault monitoring and diagnosis system and method for industrial robots

Abstract

The present invention provides an intelligent fault monitoring and diagnosis system and method for industrial robots, which belongs to the field of measurement data analysis technology. The system includes: a real-time data acquisition subsystem for acquiring real-time data of a target sensor of a target industrial robot; a fault monitoring subsystem for performing fault monitoring according to real-time data; a fault diagnosis subsystem for performing fault diagnosis according to real-time data based on data analysis and machine learning technology during fault monitoring; and a diagnostic report output subsystem for outputting a diagnostic report. The intelligent fault monitoring and diagnosis system and method for industrial robots of the present invention deploy target sensors at the monitoring part of the target industrial robot to acquire real-time data; based on machine learning technology, determine the diagnostic fault and analyze the cause according to the real-time data, output a diagnostic report, and specifically determine the monitoring part to collect fault detection data, thereby improving the efficiency of subsequent fault detection.

Description

Technical Field

[0001] The present invention relates to the technical field of measurement data analysis, and in particular to an intelligent fault monitoring and diagnosis system and method for industrial robots. Background Art

[0002] An industrial robot is a multifunctional, multi-degree-of-freedom manipulator that can be automatically controlled and reprogrammed. It can carry materials, workpieces, or hold tools to complete various tasks. Industrial robots are widely used in many fields, including but not limited to automobile manufacturing, electronic manufacturing, logistics, and petrochemicals. With the advancement of technology, industrial robots are becoming more intelligent and flexible, and can adapt to more complex application scenarios. The widespread use of industrial robots has improved production efficiency and accuracy. The intelligent fault monitoring and diagnosis system can monitor the working status of the robot in real time, detect and handle faults in a timely manner, reduce production interruption time, and ensure the continuity and stability of production.

[0003] The invention patent with application number: CN202111062865.7 discloses an early fault detection method for industrial robot harmonic reducer based on WLCTD and CNN-LSTM, wherein the method comprises: step 1: obtaining a data set of the vibration state of the harmonic reducer; step 2: using the WLCTD algorithm to remove the noise from the data set of step 1; step 3: using the CNN-LSTM network to process the noise-removed data set; step 4: performing constraint optimization on the data set after the CNN-LSTM network data processing in step 3; step 5: classifying and fault detection on the data set after the constraint optimization in step 4. The above invention is used to solve the downtime loss caused by the failure of the harmonic reducer.

[0004] However, the above-mentioned prior art does not allocate fault detection resources according to the characteristics of each part of the industrial robot in a targeted manner, and has the problem of low detection efficiency.

[0005] In view of this, there is an urgent need for an intelligent fault monitoring and diagnosis system and method for industrial robots to at least solve the above-mentioned deficiencies. Summary of the invention

[0006] One of the purposes of the present invention is to provide an intelligent fault monitoring and diagnosis system and method for industrial robots, deploy target sensors at the monitoring location of the target industrial robot, and obtain real-time data of the target sensors; based on data analysis and machine learning technology, perform fault diagnosis on the target industrial robot according to the real-time data; determine the diagnosed fault and perform cause analysis, output a diagnosis report, and specifically determine the monitoring location to collect fault detection data, thereby improving the efficiency of subsequent fault detection.

[0007] An embodiment of the present invention provides an intelligent fault monitoring and diagnosis system for industrial robots, comprising:

[0008] A real-time data acquisition subsystem, used to acquire real-time data of a target sensor of a target industrial robot;

[0009] A fault monitoring subsystem is used to monitor the faults of the target industrial robot based on real-time data;

[0010] The fault diagnosis subsystem is used to diagnose the faults of the target industrial robot based on real-time data based on data analysis and machine learning technology during the fault monitoring process;

[0011] The diagnostic report output subsystem is used to analyze the cause of the diagnostic fault and output a diagnostic report if there is at least one diagnostic fault.

[0012] Preferably, the real-time data acquisition subsystem includes:

[0013] A sensor deployment resource acquisition module is used to acquire sensor deployment resources;

[0014] A movable joint point acquisition module is used to acquire the movable joint points of the target industrial robot;

[0015] A full deployment determination module is used to determine whether full deployment is possible based on sensor deployment resources and movable joints;

[0016] The first real-time data acquisition module is used to, if full deployment is possible, use all movable joints as monitoring locations, deploy corresponding target sensors for each movable joint and acquire real-time data;

[0017] A monitoring part determination module, used to determine the monitoring part in the movable joint point if full deployment is not possible;

[0018] The second real-time data acquisition module is used to deploy corresponding target sensors for the monitored parts and acquire real-time data.

[0019] Preferably, the monitoring site determination module includes:

[0020] The industrial service type acquisition submodule is used to obtain the industrial service type of the target industrial robot;

[0021] A target action sequence determination submodule is used to determine a target action sequence of a target industrial robot according to the industrial service type;

[0022] A second target action determination submodule is used to sequentially traverse each first target action in the target action sequence and use the first target action being traversed as the second target action;

[0023] A rotation joint point determination submodule is used to compare the second target action with the standard posture of the target industrial robot to determine the rotation joint point;

[0024] A rotation count submodule is used to count the number of rotation joints determined after all first target actions are traversed, and determine the number of rotations of each rotation joint;

[0025] The monitoring part determination submodule is used to determine the monitoring parts that can be covered by the sensor deployment resources according to the order of the number of rotations from high to low and the sensor deployment resources.

[0026] An intelligent fault monitoring and diagnosis system for industrial robots provided by an embodiment of the present invention further includes:

[0027] The access cycle setting submodule is used to set the access cycle with the target sensor according to the number of rotations when acquiring real-time data.

[0028] Preferably, the fault monitoring subsystem includes:

[0029] A first operation record set acquisition module, used to acquire a first operation record set of normal operation of a target industrial robot;

[0030] A standard working status data range determination module, used to determine the standard working status data range according to the first operation record set;

[0031] The fault monitoring module is used to perform fault monitoring based on real-time data and standard working status data range.

[0032] Preferably, the fault diagnosis subsystem includes:

[0033] A second operation record set acquisition module, used to acquire a second operation record set of the target industrial robot; the second operation record set includes a plurality of historical fault records;

[0034] A training module, used for training a fault diagnosis model according to the second job record set based on data analysis and machine learning technology;

[0035] The fault diagnosis result acquisition module is used to extract input features according to real-time data, input the input features into the fault diagnosis model, and obtain the fault diagnosis results.

[0036] Preferably, the training module includes:

[0037] The attribution credible value determination submodule is used to obtain the fault attribution type of the historical fault record and determine the attribution credible value of the fault attribution type;

[0038] A third operation record set screening submodule is used to remove historical fault records of fault attribution types that do not meet the attribution credibility value, and classify and summarize the remaining historical fault records in the second operation record set based on the fault attribution type to obtain multiple third operation record sets;

[0039] A training feature extraction submodule, used for extracting training features according to the third operation record set and the fault attribution type;

[0040] The training submodule is used to train the fault diagnosis model according to the training features based on data analysis and machine learning techniques.

[0041] Preferably, the attribution credibility value determination submodule includes:

[0042] An attribution credibility value determination basis extraction template acquisition unit, used to acquire an attribution credibility value determination basis extraction template;

[0043] An attribution credible value determination basis extraction unit, used to extract an attribution credible value determination basis set of a fault attribution type based on an attribution credible value determination basis extraction template;

[0044] The attribution credibility value determination unit is used to determine the attribution credibility value according to the attribution credibility value determination basis set.

[0045] Preferably, the attribution credibility value determination unit includes:

[0046] The basis determination subunit is used to compare the attribution credibility value determination basis set with the preset standard determination basis set to determine whether the attribution credibility value determination basis set is the same as any of the standard determination basis sets;

[0047] The attribution credibility value first determination subunit is used to obtain the first determination strategy of the corresponding standard determination basis set if there is an attribution credibility value determination basis set that is the same as the standard determination basis set, and determine the attribution credibility value according to the attribution credibility value determination basis set and the first determination strategy;

[0048] A missing judgment basis determination subunit is used to obtain the standard judgment basis set with the greatest similarity if there is no attribution credible value judgment basis set that is the same as the standard judgment basis set, and determine the missing judgment basis of the attribution credible value judgment basis set corresponding to the standard judgment basis set with the greatest similarity;

[0049] A compensation degree acquisition subunit, used for attempting to acquire compensation basis for missing determination basis, and acquiring a compensation degree of the compensation basis;

[0050] The second attribution credibility value determination subunit is used to determine the second determination strategy based on the first determination strategy and compensation degree of the standard determination basis set with the greatest similarity, and determine the attribution credibility value based on the attribution credibility value determination basis set and the second determination strategy.

[0051] An embodiment of the present invention provides an intelligent fault monitoring and diagnosis method for an industrial robot, comprising:

[0052] Step 1: Acquire real-time data of the target sensor of the target industrial robot;

[0053] Step 2: Perform fault monitoring of the target industrial robot based on real-time data;

[0054] Step 3: During the fault monitoring process, based on data analysis and machine learning technology, fault diagnosis is performed on the target industrial robot according to real-time data;

[0055] Step 4: If there is at least one diagnostic fault, analyze the cause of the diagnostic fault and output a diagnostic report.

[0056] The beneficial effects of the present invention are:

[0057] The present invention deploys a target sensor at a monitoring location of a target industrial robot and obtains real-time data of the target sensor; based on data analysis and machine learning technology, performs fault diagnosis on the target industrial robot according to the real-time data; determines the diagnosed fault and performs cause analysis, outputs a diagnosis report, and specifically determines the monitoring location to collect fault detection data, thereby improving the efficiency of subsequent fault detection.

[0058] Other features and advantages of the present invention will be described in the following description, and partly become apparent from the description, or understood by practicing the present invention. The purpose and other advantages of the present invention can be realized and obtained by the structures specifically pointed out in the present application documents.

[0059] The technical solution of the present invention is further described in detail below through the accompanying drawings and embodiments. BRIEF DESCRIPTION OF THE DRAWINGS

[0060] The accompanying drawings are used to provide a further understanding of the present invention and constitute a part of the specification. Together with the embodiments of the present invention, they are used to explain the present invention and do not constitute a limitation of the present invention. In the accompanying drawings:

[0061] Figure 1 Schematic diagram of an intelligent fault monitoring and diagnosis system for industrial robots in an embodiment of the present invention;

[0062] Figure 2 Schematic diagram of an intelligent fault monitoring and diagnosis method for an industrial robot in an embodiment of the present invention. DETAILED DESCRIPTION

[0063] The preferred embodiments of the present invention are described below in conjunction with the accompanying drawings. It should be understood that the preferred embodiments described herein are only used to illustrate and explain the present invention, and are not used to limit the present invention.

[0064] The embodiment of the present invention provides an intelligent fault monitoring and diagnosis system for industrial robots. Figure 1 As shown, including:

[0065] The real-time data acquisition subsystem 1 is used to acquire the real-time data of the target sensor of the target industrial robot; wherein the target industrial robot is an industrial robot that needs to be monitored, such as an industrial robot that performs sorting tasks on an automated production line; the target sensor is a sensor installed on a monitoring part that collects data (such as temperature, pressure, and vibration) on the operating status of the target industrial robot, and the monitoring part is a part of the target industrial robot that needs to be monitored, such as a joint or a drive; the real-time data is data collected by the target sensor;

[0066] The fault monitoring subsystem 2 is used to perform fault monitoring of the target industrial robot according to the real-time data; wherein the fault monitoring is a process of comparing the real-time data with the normal state data to monitor whether the target industrial robot has a fault;

[0067] Fault diagnosis subsystem 3, used for performing fault diagnosis on the target industrial robot based on data analysis and machine learning technology and real-time data during fault monitoring; wherein, performing fault diagnosis on the target industrial robot based on data analysis and machine learning technology and real-time data refers to the process in which AI performs fault diagnosis on the industrial robot based on sensor data by learning artificial intelligence, and performs fault diagnosis on the target industrial robot based on the learning results and real-time data;

[0068] The diagnosis report output subsystem 4 is used to analyze the cause of the diagnosis fault and output a diagnosis report if there is at least one diagnosis fault.

[0069] The working principle and beneficial effects of the above technical solution are:

[0070] The present application deploys target sensors at the monitoring location of the target industrial robot and obtains real-time data of the target sensors; based on data analysis and machine learning technology, fault diagnosis is performed on the target industrial robot according to the real-time data; the diagnosed fault is determined and the cause is analyzed, a diagnosis report is output, and the monitoring location is determined in a targeted manner to collect fault detection data, thereby improving the efficiency of subsequent fault detection.

[0071] The embodiment of the present invention provides an intelligent fault monitoring and diagnosis system for industrial robots, further comprising:

[0072] Predictive maintenance subsystem, used to predict potential failure points of target industrial robots through historical data and trend analysis, and perform preventive maintenance;

[0073] The health assessment subsystem is used to assess the health status of the target industrial robot and provide maintenance and repair priority recommendations;

[0074] The remote monitoring and control subsystem is used to obtain the control operations of the operator when monitoring the target industrial robot based on the remote control interface and perform corresponding control;

[0075] A self-learning optimization subsystem for optimizing the fault diagnosis process based on real-time data and operator feedback;

[0076] The alarm and notification subsystem is used to automatically trigger an alarm and notify relevant maintenance personnel when a diagnostic fault or abnormal behavior of the target industrial robot is detected;

[0077] The historical data analysis subsystem is used to store and analyze the historical operation data of the target industrial robot and make future fault detection and maintenance decisions based on the historical operation data;

[0078] The interactive maintenance guidance subsystem is used to provide maintenance guidance and fault repair suggestions based on the diagnosed faults.

[0079] The working principle and beneficial effects of the above technical solution are:

[0080] Predict potential failure points of the target industrial robot through historical data and trend analysis to achieve preventive maintenance; evaluate the overall health of the target industrial robot and provide maintenance and repair priority recommendations; the operator monitors the status of the target industrial robot through a remote interface and controls it; self-optimizes based on newly collected data and operator feedback to improve the accuracy of fault diagnosis; automatically triggers an alarm and notifies relevant maintenance personnel when a fault or abnormal behavior is detected; stores and analyzes the historical operating data of the target industrial robot to assist in future fault detection and maintenance decisions; provides maintenance guidance and fault repair suggestions to help operators quickly resolve problems.

[0081] In one embodiment, the real-time data acquisition subsystem includes:

[0082] The sensor deployment resource acquisition module is used to acquire sensor deployment resources; wherein the sensor deployment resources are: types and quantity information of sensors that can be deployed;

[0083] The movable joint point acquisition module is used to acquire the movable joint points of the target industrial robot; wherein the movable joint points are: movable joint points of the target industrial robot;

[0084] A full deployment determination module is used to determine whether full deployment is possible based on sensor deployment resources and movable joints; wherein full deployment means that each movable joint can be equipped with the corresponding required sensors;

[0085] The first real-time data acquisition module is used to, if full deployment is possible, use all movable joints as monitoring locations, deploy corresponding target sensors for each movable joint and acquire real-time data;

[0086] A monitoring part determination module is used to determine the monitoring part in the movable joint point if full deployment is not possible; wherein the monitoring part is: the joint point in the movable joint that needs to be monitored first;

[0087] The second real-time data acquisition module is used to deploy corresponding target sensors for the monitored parts and acquire real-time data.

[0088] The working principle and beneficial effects of the above technical solution are:

[0089] The present invention determines whether the current resources can support the deployment of corresponding required sensors at each movable joint point according to the acquired sensor deployment resources and the movable joint points of the target industrial robot. If so, all movable joint points are used as monitoring parts, and corresponding target sensors are deployed for each movable joint point to obtain real-time data. If full deployment is not possible, the movable joint points (monitoring parts) that need to be monitored first are determined, and then the sensors are deployed and real-time data is obtained. In the case of limited sensor resources, the deployment efficiency of the sensors is improved.

[0090] In one embodiment, the monitoring part determination module includes:

[0091] The industrial service type acquisition submodule is used to acquire the industrial service type of the target industrial robot; the industrial service type is: the type of task undertaken by the target industrial robot in industrial applications, such as sorting, assembly and handling;

[0092] The target action sequence determination submodule is used to determine the target action sequence of the target industrial robot according to the industrial service type; wherein the target action sequence is: an action sequence including a sequential action sequence generated by the target industrial robot when performing a service of the industrial service type;

[0093] A second target action determination submodule is used to sequentially traverse each first target action in the target action sequence and use the first target action being traversed as the second target action;

[0094] The rotation joint point determination submodule is used to compare the second target action with the standard posture of the target industrial robot to determine the rotation joint point; wherein the standard posture is: the posture of the target industrial robot in the standby state;

[0095] A rotation count submodule is used to count the number of rotation joints determined after all first target actions are traversed, and determine the number of rotations of each rotation joint;

[0096] The monitoring part determination submodule is used to determine the monitoring parts that can be covered by the sensor deployment resources according to the order of the number of rotations from high to low and the sensor deployment resources. Among them, according to the order of the number of rotations from high to low and the sensor deployment resources, the monitoring parts that can be covered by the sensor deployment resources are determined as follows: the sensor deployment resources are allocated to the corresponding rotation joints according to the order of the number of rotations from high to low, and when the allocation is completed, the rotation joints to which the sensor deployment resources have been allocated are used as monitoring parts.

[0097] The working principle and beneficial effects of the above technical solution are:

[0098] The present invention introduces the industrial service type of the target industrial robot, determines the target action sequence corresponding to the industrial service type; sequentially traverses the first target action in the target action sequence, determines the second target action being traversed; compares the second target action with the standard posture of the target industrial robot, and determines the rotation joint point where the activity occurs; after all the first target actions are traversed, counts the number of times the rotation joint points are determined, and allocates limited sensor deployment resources accordingly in descending order of the number of rotations, thereby improving the allocation accuracy of the sensor deployment resources.

[0099] The embodiment of the present invention provides an intelligent fault monitoring and diagnosis system for industrial robots, further comprising:

[0100] The access cycle setting submodule is used to set the access cycle with the target sensor according to the number of rotations when acquiring real-time data. When setting the access cycle of the target sensor, the higher the number of rotations, the shorter the corresponding access cycle. The specific ratio between the number of rotations and the access cycle is set manually.

[0101] The working principle and beneficial effects of the above technical solution are:

[0102] When acquiring real-time data, the present invention sets an access cycle for accessing a target sensor according to the number of rotations, thereby improving access efficiency.

[0103] In one embodiment, the fault monitoring subsystem includes:

[0104] A first operation record set acquisition module, used to acquire a first operation record set of a target industrial robot in normal operation; wherein the first operation record set includes operation process records of multiple target industrial robots in normal operation, including detection data of sensors at various parts;

[0105] A standard working state data range determination module is used to determine the standard working state data range according to the first operation record set; wherein the standard working state data range is: the range of detection data of each sensor when the target industrial robot is operating normally;

[0106] The fault monitoring module is used to perform fault monitoring according to the real-time data and the standard working state data range. When performing fault monitoring according to the real-time data and the standard working state data range, it is monitored whether the real-time data falls within the standard working state data range. If it falls within the standard working state data range, the working state is normal, otherwise it is a fault.

[0107] The working principle and beneficial effects of the above technical solution are:

[0108] The present invention introduces a first operation record set of normal operation of a target industrial robot, determines a standard operating state data range corresponding to the first operation record set and performs subsequent fault monitoring, and the process of determining the standard operating state data range is more reasonable.

[0109] In one embodiment, the fault diagnosis subsystem includes:

[0110] A second operation record set acquisition module is used to acquire a second operation record set of the target industrial robot; the second operation record set includes a plurality of historical fault records; wherein the historical fault record is: a process record of a fault occurring in the target industrial robot in history, including indicators of each sensor at the time of the fault and a record of the type of fault, etc.;

[0111] A training module, for training a fault diagnosis model according to the second operation record set based on data analysis and machine learning technology; wherein the fault diagnosis model is an AI model that learns the relationship between the fault data features recorded in the historical fault records in the second operation record set and the corresponding fault types;

[0112] The fault diagnosis result acquisition module is used to extract input features based on real-time data, input the input features into the fault diagnosis model, and obtain the fault diagnosis results. The output features are: the specific values ​​of the sensor data in the real-time data;

[0113] The working principle and beneficial effects of the above technical solution are:

[0114] The present invention introduces data analysis technology and machine learning technology, trains a fault diagnosis model according to the acquired second operation record set; extracts input features of real-time data, inputs the input features into the fault diagnosis model, obtains fault diagnosis results, monitors the health status of the target industrial robot in real time, quickly diagnoses faults and facilitates subsequent fault processing, thereby reducing downtime and improving production efficiency.

[0115] In one embodiment, the training module includes:

[0116] The attribution credibility value determination submodule is used to obtain the fault attribution type of the historical fault record and determine the attribution credibility value of the fault attribution type; wherein the fault attribution type is: the fault type of the industrial robot recorded in the historical fault record; the attribution credibility value indicates the credibility of the fault attribution type, and the larger the attribution credibility value, the more credibility;

[0117] The third operation record set screening submodule is used to remove historical fault records of the fault attribution type that does not meet the attribution credibility value, and classify and summarize the remaining historical fault records in the second operation record set based on the fault attribution type to obtain multiple third operation record sets; wherein the third operation record set is: a set obtained by removing historical fault records of unreliable attribution and summarizing historical fault records of the same reliable fault attribution type;

[0118] The training feature extraction submodule is used to extract training features according to the third operation record set and the fault attribution type; wherein the training feature is: which sensor data feature corresponds to which fault attribution type;

[0119] The training submodule is used to train the fault diagnosis model according to the training features based on data analysis and machine learning techniques.

[0120] The working principle and beneficial effects of the above technical solution are:

[0121] The present invention determines the attribution credibility value of the fault attribution type of the historical fault record, eliminates the historical fault records of the fault attribution type that does not meet the attribution credibility value, and classifies and summarizes the historical fault records corresponding to the remaining credible fault attribution types, extracts sensor data features and corresponding fault attribution types; introduces data analysis and machine learning technology to learn training features to obtain a fault diagnosis model, thereby improving the training quality.

[0122] In one embodiment, the attribution credibility value determination submodule includes:

[0123] The attribution trustworthy value determination basis extraction template acquisition unit is used to acquire the attribution trustworthy value determination basis extraction template; wherein the attribution trustworthy value determination basis extraction template is used to extract the attribution trustworthy value determination basis by comparing the attribution record of the fault attribution type, for example: the attribution trustworthy value determination basis is the attribution personnel information (for example: personnel position, experience), then the attribution trustworthy value determination basis extraction template correspondingly executes the instruction of acquiring the attribution personnel information;

[0124] The attribution credibility value determination basis extraction unit is used to extract the attribution credibility value determination basis set of the fault attribution type based on the attribution credibility value determination basis extraction template; wherein the attribution credibility value determination basis set is: all attribution credibility value determination bases extracted from the attribution record of the attribution type of which the attribution credibility value is being calculated;

[0125] The attribution credibility value determination unit is used to determine the attribution credibility value according to the attribution credibility value determination basis set.

[0126] The working principle and beneficial effects of the above technical solution are:

[0127] The present invention introduces an attribution credible value determination basis extraction template, extracts an attribution credible value determination basis set of a fault attribution type, calculates an attribution credible value based on the attribution credible value determination basis, and improves the accuracy of attribution credible value calculation.

[0128] In one embodiment, the attribution credibility value determination unit includes:

[0129] The basis determination subunit is used to compare the attribution credibility value determination basis set with the preset standard determination basis set to determine whether there is an attribution credibility value determination basis set that is the same as any standard determination basis set; wherein the standard determination basis set is: a combination of bases that can determine the attribution credibility value together, such as: attribution personnel information and attribution result recording personnel information;

[0130] The attribution credibility value first determination subunit is used for obtaining the first determination strategy of the corresponding standard determination basis set if there is an attribution credibility value determination basis set that is the same as the standard determination basis set, and determining the attribution credibility value according to the attribution credibility value determination basis set and the first determination strategy; wherein the first determination strategy is: a method for calculating the attribution credibility value according to the basis value of the basis item in the basis combination corresponding to the basis combination, for example: the standard determination basis set includes: attribution personnel information and attribution result recording personnel information, then the corresponding first determination strategy is: first determine the target value corresponding to the attribution personnel information (the working years of the attribution personnel divided by the average staff years multiplied by a preset coefficient (100)), then verify whether the attribution result recording personnel information is credible, and if it is credible, output the target value as the attribution credibility value;

[0131] The missing judgment basis determination subunit is used to obtain the standard judgment basis set with the greatest similarity if there is no attribution credible value judgment basis set that is the same as the standard judgment basis set, and determine the missing judgment basis of the attribution credible value judgment basis set corresponding to the standard judgment basis set with the greatest similarity; wherein the missing judgment basis is: the different judgment basis obtained by comparing the attribution credible value judgment basis set with the closest standard judgment basis set;

[0132] The compensation degree acquisition subunit is used to try to obtain the compensation basis of the missing judgment basis and obtain the compensation degree of the compensation basis; wherein the compensation basis is: the correlation basis of the missing judgment basis, for example: if the missing judgment basis is the attribution personnel information, then the compensation basis is the personnel information of the attribution personnel's co-workers; the compensation degree is: the correlation degree between the missing judgment basis and the compensation basis;

[0133] The second determination subunit of the attribution credibility value is used to determine the second determination strategy according to the first determination strategy and compensation degree of the standard determination basis set with the greatest similarity, and determine the attribution credibility value according to the attribution credibility value determination basis set and the second determination strategy. Among them, the second determination strategy is: a strategy obtained by adjusting the first determination strategy based on the compensation degree, for example: when the compensation basis is the personnel information of the staff of the same period of the attribution personnel, the second determination strategy is: for example: first determine the target value corresponding to the personnel information of the staff of the same period (the working years of the staff of the same period divided by the average staff years multiplied by the preset coefficient (100)), and then verify that the personnel information recorded in the attribution result is credible, and then output the product of the target value and the compensation degree (for example: 0.9) as the attribution credibility value.

[0134] The working principle and beneficial effects of the above technical solution are:

[0135] The present invention introduces a standard judgment basis set, compares the attribution credible value judgment basis set with the standard judgment basis set, and determines whether there is any similarity; if there is any similarity, obtains the first judgment strategy of the corresponding standard judgment basis set, and determines the attribution credible value based on the first judgment strategy and the attribution credible value judgment basis set with the same existence; if there is no attribution credible value judgment basis set that is the same as the standard judgment basis set, determines the standard judgment basis set with the greatest similarity, and determines the missing judgment basis with the standard judgment basis set with the greatest similarity; obtains the compensation basis and compensation degree of the missing judgment basis; according to the compensation basis and compensation degree, adjusts the first judgment strategy of the standard judgment basis set with the greatest similarity to obtain the second judgment strategy, and determines the attribution credible value according to the attribution credible value judgment basis set and the second judgment strategy, thereby further improving the judgment accuracy of the attribution credible value.

[0136] The embodiment of the present invention provides an intelligent fault monitoring and diagnosis method for industrial robots. Figure 2 As shown, including:

[0137] Step 1: Acquire real-time data of the target sensor of the target industrial robot;

[0138] Step 2: Perform fault monitoring of the target industrial robot based on real-time data;

[0139] Step 3: During the fault monitoring process, based on data analysis and machine learning technology, fault diagnosis is performed on the target industrial robot according to real-time data;

[0140] Step 4: If there is at least one diagnostic fault, analyze the cause of the diagnostic fault and output a diagnostic report.

[0141] Obviously, those skilled in the art can make various changes and modifications to the present invention without departing from the spirit and scope of the present invention. Thus, if these modifications and variations of the present invention fall within the scope of the claims of the present invention and their equivalents, the present invention is also intended to include these modifications and variations.

Claims

1. An intelligent fault monitoring and diagnosis system for industrial robots, characterized in that: include: A real-time data acquisition subsystem, used to acquire real-time data of a target sensor of a target industrial robot; A fault monitoring subsystem is used to monitor the faults of the target industrial robot based on real-time data; The fault diagnosis subsystem is used to diagnose the faults of the target industrial robot based on real-time data based on data analysis and machine learning technology during the fault monitoring process; A diagnostic report output subsystem, for analyzing the cause of the diagnostic fault and outputting a diagnostic report if at least one diagnostic fault exists; The real-time data acquisition subsystem comprises: A sensor deployment resource acquisition module is used to acquire sensor deployment resources; A movable joint point acquisition module is used to acquire the movable joint points of the target industrial robot; A full deployment determination module is used to determine whether full deployment is possible based on sensor deployment resources and movable joints; The first real-time data acquisition module is used to, if fully deployed, use all movable joints as monitoring locations, deploy corresponding target sensors for each movable joint and acquire real-time data; A monitoring part determination module, used to determine the monitoring part in the movable joint point if full deployment is not possible; A second real-time data acquisition module is used to deploy corresponding target sensors for the monitored parts and acquire real-time data; The monitoring part determination module comprises: The industrial service type acquisition submodule is used to obtain the industrial service type of the target industrial robot; A target action sequence determination submodule is used to determine a target action sequence of a target industrial robot according to the industrial service type; A second target action determination submodule is used to sequentially traverse each first target action in the target action sequence and use the first target action being traversed as the second target action; A rotation joint point determination submodule is used to compare the second target action with the standard posture of the target industrial robot to determine the rotation joint point; A rotation count submodule is used to count the number of rotation joints determined after all first target actions are traversed, and determine the number of rotations of each rotation joint; The monitoring part determination submodule is used to determine the monitoring parts that can be covered by the sensor deployment resources according to the order of the number of rotations from high to low and the sensor deployment resources.

2. The intelligent fault monitoring and diagnosis system for industrial robots according to claim 1, characterized in that: Also includes: The access cycle setting submodule is used to set the access cycle with the target sensor according to the number of rotations when acquiring real-time data.

3. The intelligent fault monitoring and diagnosis system for industrial robots according to claim 1, characterized in that: Fault monitoring subsystem, including: A first operation record set acquisition module, used to acquire a first operation record set of normal operation of a target industrial robot; A standard working status data range determination module, used to determine the standard working status data range according to the first operation record set; The fault monitoring module is used to perform fault monitoring based on real-time data and standard working status data range.

4. The intelligent fault monitoring and diagnosis system for industrial robots according to claim 1, characterized in that: Fault diagnosis subsystem, including: A second operation record set acquisition module, used to acquire a second operation record set of the target industrial robot; the second operation record set includes a plurality of historical fault records; A training module, used for training a fault diagnosis model according to the second job record set based on data analysis and machine learning technology; The fault diagnosis result acquisition module is used to extract input features according to real-time data, input the input features into the fault diagnosis model, and obtain the fault diagnosis results.

5. The intelligent fault monitoring and diagnosis system for industrial robots according to claim 4, characterized in that: Training modules include: The attribution credible value determination submodule is used to obtain the fault attribution type of the historical fault record and determine the attribution credible value of the fault attribution type; A third operation record set screening submodule is used to remove historical fault records of fault attribution types that do not meet the attribution credibility value, and classify and summarize the remaining historical fault records in the second operation record set based on the fault attribution type to obtain multiple third operation record sets; A training feature extraction submodule, used for extracting training features according to the third operation record set and the fault attribution type; The training submodule is used to train the fault diagnosis model according to the training features based on data analysis and machine learning techniques.

6. The intelligent fault monitoring and diagnosis system for industrial robots according to claim 5, characterized in that: The attribution trust value determination submodule includes: An attribution credibility value determination basis extraction template acquisition unit, used to acquire an attribution credibility value determination basis extraction template; An attribution credible value determination basis extraction unit, used to extract an attribution credible value determination basis set of a fault attribution type based on an attribution credible value determination basis extraction template; The attribution credibility value determination unit is used to determine the attribution credibility value according to the attribution credibility value determination basis set.

7. An intelligent fault monitoring and diagnosis system for industrial robots as claimed in claim 6, characterized in that: The attribution credibility value determination unit includes: The basis determination subunit is used to compare the attribution credibility value determination basis set with the preset standard determination basis set to determine whether the attribution credibility value determination basis set is the same as any of the standard determination basis sets; The attribution credibility value first determination subunit is used to obtain the first determination strategy of the corresponding standard determination basis set if there is an attribution credibility value determination basis set that is the same as the standard determination basis set, and determine the attribution credibility value according to the attribution credibility value determination basis set and the first determination strategy; A missing judgment basis determination subunit is used to obtain the standard judgment basis set with the greatest similarity if there is no attribution credible value judgment basis set that is the same as the standard judgment basis set, and determine the missing judgment basis of the attribution credible value judgment basis set corresponding to the standard judgment basis set with the greatest similarity; A compensation degree acquisition subunit, used for attempting to acquire compensation basis for missing determination basis, and acquiring a compensation degree of the compensation basis; The second attribution credibility value determination subunit is used to determine the second determination strategy based on the first determination strategy and compensation degree of the standard determination basis set with the greatest similarity, and determine the attribution credibility value based on the attribution credibility value determination basis set and the second determination strategy.

8. An intelligent fault monitoring and diagnosis method for industrial robots, characterized in that: include: Step 1: Acquire real-time data of the target sensor of the target industrial robot; Step 2: Perform fault monitoring of the target industrial robot based on real-time data; Step 3: During the fault monitoring process, based on data analysis and machine learning technology, fault diagnosis is performed on the target industrial robot according to real-time data; Step 4: If there is at least one diagnostic fault, analyze the cause of the diagnostic fault and output a diagnostic report; The step 1: obtaining real-time data of a target sensor of a target industrial robot includes: Get sensor deployment resources; Obtain the movable joint points of the target industrial robot; Determine whether full deployment is possible based on sensor deployment resources and movable joints; If it is possible to deploy it comprehensively, all movable joints will be used as monitoring parts, and corresponding target sensors will be deployed for each movable joint to obtain real-time data; If full deployment is not possible, determine the monitoring locations in the movable joints; Deploy corresponding target sensors for monitoring locations and obtain real-time data; If the deployment cannot be fully completed, the monitoring parts in the movable joints are determined, including: Get the industrial service type of the target industrial robot; Determine the target action sequence of the target industrial robot according to the industrial service type; Traversing each first target action in the target action sequence in turn, and taking the first target action being traversed as the second target action; Compare the second target action with the standard posture of the target industrial robot to determine the rotation joint point; After all the first target actions are traversed, the number of times the rotation joints are determined is counted to determine the number of rotations of each rotation joint; According to the order of the number of rotations from high to low and the sensor deployment resources, the monitoring locations that can be covered by the sensor deployment resources are determined.

Images