Industrial robot fault detection system based on Internet

Abstract

The invention relates to the technical field of fault detection, and discloses an internet-based industrial robot fault detection system, which is used for solving the problem that data acquired by a sensor is inaccurate due to the change of the environment where a robot is located during robot fault detection, and comprises the following steps of: acquiring operation data through the sensor; an initial fault index is obtained through evaluation according to the operation data, first-time fault judgment is conducted according to the initial fault index, if the first-time fault judgment robot breaks down, environment influence judgment is conducted according to the real-time environment temperature, and if the environment influence judgment sensor set is influenced by the environment, the operation data is corrected; the actual fault index is obtained through evaluation according to the operation correction data, second-time fault judgment is carried out according to the actual fault index, if the second-time fault judgment shows that the current robot breaks down, an early warning prompt is given out, the influence of environment changes on data collected by a sensor is effectively eliminated, and the accuracy of fault judgment is improved.

Description

Technical Field

[0001] The present invention relates to the technical field of fault detection, and more particularly to an Internet-based industrial robot fault detection system. Background Art

[0002] Industrial robots are mechanical devices used in automated production processes. They can perform various repetitive and high-precision tasks through programming, such as welding, assembly, handling, spraying, and inspection. They usually have robotic arms, sensors, control systems, and drive devices, and can work efficiently and stably in industrial environments, replacing humans to complete complex or dangerous operations. Industrial robots are widely used in manufacturing, the automotive industry, electronic assembly, and other fields, significantly improving production efficiency, product quality, and work safety.

[0003] Industrial robot fault detection is the process of identifying potential faults or performance anomalies by monitoring and analyzing the operating data of each component of the robot. Using sensors, data acquisition systems, and intelligent algorithms to monitor the working state of the robot in real time, predicting and warning possible faults in advance, reducing downtime, extending the equipment life, and ensuring the stable operation of the production line. However, in actual applications, the working environment of industrial robots changes in real time, resulting in inaccurate data collected by sensors and false alarms.

[0004] In view of the above problems, the present invention proposes a solution. Summary of the Invention

[0005] In order to overcome the above-mentioned defects of the prior art, the present invention provides an Internet-based industrial robot fault detection system to solve the problems existing in the above background art.

[0006] To achieve the above object, the present invention provides the following technical solutions: An Internet-based industrial robot fault detection system, the system comprising: a robot data acquisition module for setting a sensor group on the industrial robot, the sensor group including a temperature sensor, a vibration sensor and a force sensor, and collecting the operation data of the industrial robot in real time through the sensor group, the operation data including motor temperature, vibration data and load data, and transmitting the operation data to the first fault analysis module; the first fault analysis module for obtaining an initial fault index according to the operation data and performing a first fault judgment according to the initial fault index; an ambient temperature acquisition module, if the first fault judgment determines that the industrial robot has a fault, collecting the real-time ambient temperature at which the industrial robot works and transmitting the real-time ambient temperature to the environmental impact judgment module; the environmental impact judgment module for performing an environmental impact judgment according to the real-time ambient temperature, and if the environmental impact judgment determines that the sensor group is affected by the environment, transmitting the real-time ambient temperature to the data correction module; the data correction module for correcting the operation data according to the real-time ambient temperature to obtain operation correction data and transmitting the operation correction data to the second fault analysis module; the second fault analysis module for evaluating the operation correction data to obtain an actual fault index and performing a second fault judgment according to the actual fault index, and if the second fault judgment determines that the current industrial robot has a fault, issuing a warning prompt to prompt relevant staff to repair the industrial robot in time.

[0007] Preferably, the steps for obtaining the initial fault index are as follows: obtaining the current motor temperature data of the industrial robot through the temperature sensor, obtaining the normal motor temperature of the industrial robot, and calculating a temperature fault coefficient according to the current motor temperature data and the normal motor temperature; setting an acquisition time period, obtaining the vibration data of the industrial robot through the vibration sensor within the acquisition time period, and evaluating a vibration fault coefficient according to the vibration data; obtaining the load data of the industrial robot through the force sensor, and evaluating a load fault coefficient according to the load data; normalizing the temperature fault coefficient, the vibration fault coefficient and the load fault coefficient, and comprehensively evaluating the initial fault index according to the normalized temperature fault coefficient, vibration fault coefficient and load fault coefficient. The specific obtaining steps are as follows: ; In the formula, is expressed as the initial fault index, is expressed as the temperature fault coefficient, is expressed as the vibration fault coefficient, is expressed as the load fault coefficient, , , are expressed as the weight coefficients of the temperature fault coefficient, the vibration fault coefficient and the load fault coefficient.

[0008] Preferably, the step of obtaining the vibration fault coefficient is as follows: Set a collection time period, divide the collection time period evenly into n parts, denoted as sub-collection time periods, and set collection time points within each sub-collection time period; Obtain the vibration data at the collection time points within each sub-collection time period through a vibration sensor, calculate the average value of the vibration data within the sub-collection time period based on the vibration data at the collection time points, and calculate the degree of fluctuation within each sub-collection time period based on the vibration data at the collection time points and the average value of the vibration data; Perform K-means clustering on the degree of fluctuation of each sub-collection time period, and obtain the vibration fault coefficient according to the clustering result.

[0009] Preferably, the step of performing K-means clustering on the degree of fluctuation of each sub-collection time period is as follows: Step 1: Use the degree of fluctuation as the clustering feature, and use the degree of fluctuation of all sub-collection time periods as the data set, and each degree of fluctuation in the data set is a data point; Step 2: Use the silhouette coefficient method to determine the optimal number of clusters K for the data set; Step 3: Randomly select K data points in the data set as the initial cluster centers. For each data point, calculate its Euclidean distance to each initial cluster center. For each data point, traverse the K initial cluster centers and assign it to the cluster corresponding to the nearest initial cluster center; Step 4: After traversing all data points, obtain the initial clusters. For each initial cluster, calculate the mean value of the data points within it to obtain a new cluster center; Step 5: Repeat Step 3 and Step 4 until the cluster centers no longer change, and obtain the final clusters and the final cluster centers.

[0010] Preferably, the step of obtaining the vibration fault coefficient according to the clustering result is as follows: Calculate the ratio of the number of data points in each final cluster to the total number in the data set to obtain the weight of each final cluster; Perform weighted summation of the weight of each final cluster and the final cluster center to obtain the vibration fault coefficient.

[0011] Preferably, the step of obtaining the load fault coefficient is as follows: During the collection time period, obtain the load data of the industrial robot through a force sensor, and calculate the average value of the load data during the collection time period; Calculate the standard deviation of the load data based on the average value of the load data, denoted as the load fluctuation degree; Obtain the load threshold of the industrial robot, and calculate the load fault coefficient based on the load threshold and the load fluctuation degree.

[0012] Preferably, the step of performing the first fault judgment according to the initial fault index is as follows: Set a fault threshold, compare the initial fault index with the fault threshold. If the initial fault index is less than the fault threshold, it is judged that the current industrial robot has not failed, and return to the robot data collection module to continue collecting operation data; If the initial fault index is greater than or equal to the fault threshold, it is judged that the current industrial robot has failed.

[0013] Preferably, the step of judging environmental impact according to the real-time environmental temperature is as follows: obtaining the standard environmental temperature of the industrial robot, calculating the temperature deviation by calculating the difference between the real-time environmental temperature and the standard environmental temperature; calculating the degree of environmental impact according to the temperature deviation, and the specific obtaining steps are as follows: ; In the formula, represents the degree of environmental impact, represents the environmental temperature impact coefficient, represents the temperature deviation; setting an impact threshold, comparing the degree of environmental impact with the impact threshold. If the degree of environmental impact is less than the impact threshold, it is judged that the sensor group is not affected by the environment, and the industrial robot has a current fault, and a warning prompt is given; if the degree of environmental impact is greater than or equal to the impact threshold, it is judged that the sensor group is affected by the environment.

[0014] Preferably, the step of performing a second fault judgment according to the actual fault index is as follows: comparing the actual fault index with the fault threshold. If the actual fault index is less than the fault threshold, it is judged that the current industrial robot has no fault, and the robot data acquisition module is returned to continue collecting operation data; if the actual fault index is greater than or equal to the fault threshold, it is judged that the current industrial robot has a fault.

[0015] The technical effects and advantages of the present invention: By collecting operation data through sensors, obtaining an initial fault index according to the evaluation of the operation data, and performing a first fault judgment according to the initial fault index. If the first fault judgment determines that the industrial robot has a fault, then judge the environmental impact according to the real-time environmental temperature. If the environmental impact judgment determines that the sensor group is affected by the environment, then correct the operation data to obtain operation corrected data, obtain the actual fault index according to the evaluation of the operation corrected data, and perform a second fault judgment according to the actual fault index. If the second fault judgment determines that the current industrial robot has a fault, a warning prompt is issued, effectively eliminating the influence of environmental changes on the data collected by the sensors and improving the accuracy of fault judgment. Description of the Drawings

[0016] Figure 1 is the overall structure diagram of the present invention. Detailed Embodiments

[0017] The technical solutions in the present invention will be clearly and completely described below in conjunction with the drawings in the present invention. In addition, the forms of each structure described in the following embodiments are only examples, and an industrial robot fault detection system based on the Internet involved in the present invention is not limited to the structures described in the following embodiments. All other embodiments obtained by those of ordinary skill in the art without creative efforts fall within the scope of protection of the present invention.

[0018] The present invention provides an Internet-based industrial robot fault detection system, such as Figure 1 As shown, the system includes: The robot data acquisition module is used to set a sensor group on the industrial robot. The sensor group includes a temperature sensor, a vibration sensor and a force sensor. The operating data of the industrial robot is collected in real time through the sensor group. The operating data includes motor temperature, vibration data and load data. For example, the arm of the industrial robot may generate vibration data of different frequencies when working. When the industrial robot carries heavy objects, the load data recorded by the load sensor is the weight data of the carried object, and the operating data is transmitted to the first fault analysis module. A sensor group consisting of temperature sensors, vibration sensors and force sensors is installed on the industrial robot to collect motor temperature, vibration data and load data in real time, which can comprehensively monitor the operating status of the industrial robot and detect potential faults or abnormalities in advance. This approach helps to timely identify problems such as motor overheating, component wear or excessive load, reduce downtime, and improve production efficiency. At the same time, through data analysis, the maintenance plan of the industrial robot is optimized, the service life of the equipment is extended, and the efficient and stable operation of the production line is ensured.

[0019] The first fault analysis module is used to obtain an initial fault index based on the operating data evaluation and make a first fault judgment based on the initial fault index; By evaluating the initial fault index based on the operating data and making the first fault judgment based on the initial fault index, early fault warning can be achieved, which helps to timely discover the potential failure risks of industrial robots. By quantifying the fault index, the current health status of the industrial robot can be accurately evaluated, and targeted maintenance or repair measures can be taken to avoid equipment damage and production interruptions caused by failure to detect the fault in time. This approach helps to improve the accuracy of fault detection and reduce unnecessary downtime, thereby improving overall production efficiency and equipment reliability.

[0020] In this embodiment, it should be specifically explained that the steps for obtaining the initial fault index are: The current motor temperature data of the industrial robot is obtained through the temperature sensor, and the normal motor temperature of the industrial robot is obtained. The temperature fault coefficient is calculated based on the current motor temperature data and the normal motor temperature. The specific acquisition steps are as follows: ; In the formula, Expressed as the temperature failure coefficient, Represents the current motor temperature data, Indicates normal motor temperature; Set the acquisition time period, and obtain the vibration data of the industrial robot through a vibration sensor within the acquisition time period, and evaluate the vibration fault coefficient according to the vibration data; Obtain the load data of the industrial robot through a force sensor, and evaluate the load fault coefficient according to the load data; Normalize the temperature fault coefficient, vibration fault coefficient, and load fault coefficient, and comprehensively evaluate the initial fault index according to the normalized temperature fault coefficient, vibration fault coefficient, and load fault coefficient. The specific acquisition steps are as follows: ; In the formula, represents the initial fault index, represents the temperature fault coefficient. The degree to which the temperature of the motor deviates from the normal operating temperature directly affects the assessment of the fault risk. When the motor temperature rises abnormally, the temperature fault coefficient increases, thereby driving the rise of the initial fault index, indicating an increase in the likelihood of a fault. This relationship shows that temperature change is an important early warning signal for industrial robot faults. By monitoring the temperature fault coefficient, potential fault risks can be detected in advance and corresponding preventive measures can be taken to ensure the safe operation of the equipment and the continuity of production, represents the vibration fault coefficient. When the industrial robot has abnormal vibrations, the vibration fault coefficient increases, indicating that there may be problems such as abnormal mechanical wear, looseness, and bearing faults inside the machine. These problems will lead to a decline in machine performance or an increase in the likelihood of a fault. By monitoring the vibration data and calculating the vibration fault coefficient, the health status of the industrial robot can be evaluated in real time, and potential fault hazards can be detected in a timely manner. The increase in the vibration fault coefficient not only indicates the abnormal operation of the equipment but also provides an important basis for fault prediction, represents the load fault coefficient. When the load exceeds the designed bearing range of the industrial robot or has abnormal fluctuations, the load fault coefficient will increase, indicating that the working load of the machine is too large or uneven, which may lead to motor overload, damage to mechanical components, or a decline in system performance. This relationship shows that by real-time monitoring of the load data and calculating the load fault coefficient, abnormal load conditions of the industrial robot during operation can be identified in a timely manner, and potential fault problems can be detected early. The increase in the load fault coefficient usually means that the equipment may face overload, excessive wear, or an impending fault. Therefore, timely optimization and adjustment of the load can help avoid equipment damage and production interruption, 、 、 represent the weight coefficients of the temperature fault coefficient, vibration fault coefficient, and load fault coefficient, and , 、 、 The specific values are determined by professionals according to the actual situation. For example , , It can be 0.4, 0.4, 0.2.

[0021] In this embodiment, it should be specifically noted that the steps for obtaining the vibration fault coefficient are as follows: Set the acquisition time period, divide the acquisition time period evenly into n parts, denoted as sub - acquisition time periods, and set the acquisition time points within each sub - acquisition time period; Obtain the vibration data of the acquisition time points within each sub - acquisition time period through a vibration sensor, calculate the average value of the vibration data within the sub - acquisition time period based on the vibration data of the acquisition time points, and calculate the degree of fluctuation within each sub - acquisition time period based on the vibration data of the acquisition time points and the average value of the vibration data. The specific acquisition steps are as follows: ; In the formula, represents the degree of fluctuation, m represents the number of acquisition time points within the sub - acquisition time period, represents the vibration data of the j - th acquisition time point within the sub - acquisition time period, represents the average value of the vibration data within the sub - acquisition time period; Perform K - means clustering on the degree of fluctuation of each sub - acquisition time period, and obtain the vibration fault coefficient according to the clustering result.

[0022] By performing K - means clustering on the degree of fluctuation of each sub - acquisition time period, the vibration data in different time periods can be divided into multiple clusters, and each cluster represents a specific vibration mode or behavior. The advantage of this method is that it can automatically identify the potential patterns in the data and classify the vibration data, thus avoiding relying solely on manually setting thresholds for judgment. Through clustering, the system can automatically evaluate the vibration fault coefficient according to the characteristics of different clusters and assign a suitable fault coefficient to each sub - acquisition time period, making the fault judgment more accurate and dynamic, better adapting to the changes of industrial robots in different working states, and improving the accuracy and reliability of fault prediction.

[0023] In this embodiment, it should be specifically noted that the steps for performing K - means clustering on the degree of fluctuation of each sub - acquisition time period are as follows: Step 1: Take the degree of fluctuation as the clustering feature, take the degree of fluctuation of all sub - acquisition time periods as the data set, and each degree of fluctuation in the data set is a data point; Step 2: Use the silhouette coefficient method to determine the optimal number of clusters K of the data set; The silhouette coefficient method is a method used to evaluate and select the number of clusters K. It determines the quality of clustering by measuring the compactness of data points with other points within their respective clusters and the separation from the nearest neighboring clusters. The silhouette coefficient of each data point reflects the difference between the average distance of the point from other points in the same cluster (intra-cluster compactness) and the average distance from the nearest cluster (inter-cluster separation). The silhouette coefficient value ranges from -1 to 1, and a larger value indicates a better clustering effect. The silhouette coefficient method calculates the average silhouette coefficient for different values of K and selects the number of clusters that maximizes the average silhouette coefficient as the optimal number of clusters; Step 3: Randomly select K data points in the dataset as the initial cluster centers. For each data point, calculate its Euclidean distance to each initial cluster center. The specific acquisition steps are as follows: , where represents the Euclidean distance from the data point to the initial cluster center, where, represents the data point, represents the initial cluster center. For each data point, traverse the K initial cluster centers and assign it to the cluster corresponding to the nearest initial cluster center; In the K-means clustering method, the Euclidean distance is a common method for measuring the difference between two data points. It represents the straight-line distance between two points in space. Simply put, the Euclidean distance can be regarded as the shortest path connecting two points, reflecting their relative positions in space. By calculating the Euclidean distance between data points and cluster centers, K-means clustering can assign each data point to the nearest cluster center, thus achieving data grouping.

[0024] Step 4: After traversing all data points, obtain the initial clusters. For each initial cluster, calculate the mean of the data points within it to obtain new cluster centers; Step 5: Repeat Step 3 and Step 4 until the cluster centers no longer change, obtaining the final clusters and final cluster centers.

[0025] In this embodiment, it should be specifically noted that the steps to obtain the vibration fault coefficient according to the clustering result are as follows: Calculate the ratio of the number of data points in each final cluster to the total number in the dataset to obtain the weight of each final cluster; Perform a weighted sum of the weight of each final cluster and the final cluster center to obtain the vibration fault coefficient. The specific acquisition steps are as follows: , where, represents the vibration fault coefficient, represents the weight of the i-th final cluster, represents the i-th final cluster center, and K is the optimal number of clusters in the dataset.

[0026] In this embodiment, it should be specifically noted that the steps for obtaining the load fault coefficient are as follows: During the acquisition time period, load data of the industrial robot is obtained through a force sensor, and the average value of the load data during the acquisition time period is calculated; The standard deviation of the load data is calculated based on the average value of the load data, denoted as the load fluctuation degree; The load threshold of the industrial robot is obtained. Usually, based on the maximum load capacity of the industrial robot, the load fault coefficient is calculated according to the load threshold and the load fluctuation degree. The specific acquisition steps are as follows: ; In the formula, represents the load fault coefficient, represents the load fluctuation degree, represents the load threshold.

[0027] In this embodiment, it should be specifically noted that the steps for the first fault judgment according to the initial fault index are as follows: A fault threshold is set, and the initial fault index is compared with the fault threshold. If the initial fault index is less than the fault threshold, it is judged that the current industrial robot has not failed, and the robot data acquisition module is returned to continue running data acquisition; if the initial fault index is greater than or equal to the fault threshold, it is judged that the current industrial robot has failed.

[0028] The ambient temperature acquisition module. If it is judged in the first fault judgment that the industrial robot has failed, the real-time ambient temperature at which the industrial robot works is acquired and transmitted to the environmental impact judgment module; In the first fault judgment, if the industrial robot has failed, acquiring the real-time ambient temperature at which the industrial robot works helps to accurately analyze the cause of the fault. The ambient temperature may be an important factor leading to the fault. For example, too high temperature may cause the motor to overheat or the control system to malfunction. By real-time monitoring the ambient temperature, it can be confirmed whether the fault is caused by temperature fluctuation, and then the working environment can be adjusted or the robot system can be optimized accordingly to avoid repeated faults caused by environmental factors. This approach helps to perform more accurate fault diagnosis, improve the accuracy of fault prediction, and ensure the long-term stable operation of the industrial robot.

[0029] The environmental impact judgment module is used to perform environmental impact judgment according to the real-time ambient temperature. If the environmental impact judgment sensor group is affected by the environment, the real-time ambient temperature is transmitted to the data correction module; In this embodiment, it should be specifically noted that the steps for environmental impact judgment according to the real-time ambient temperature are as follows: The standard ambient temperature of the industrial robot is obtained, and the temperature deviation is calculated by taking the difference between the real-time ambient temperature and the standard ambient temperature; The degree of environmental impact is calculated based on the temperature deviation, and the specific acquisition steps are as follows: ; In the formula, represents the degree of environmental impact, represents the environmental temperature impact coefficient, which is used to represent the impact intensity of temperature change on the industrial robot and is usually obtained through experiments. represents the temperature deviation; Set an impact threshold, and compare the degree of environmental impact with the impact threshold. If the degree of environmental impact is less than the impact threshold, it is judged that the sensor group is not affected by the environment, and the industrial robot has a current fault, and a warning prompt is given to prompt the relevant staff to repair the industrial robot in time; if the degree of environmental impact is greater than or equal to the impact threshold, it is judged that the sensor group is affected by the environment.

[0030] In this embodiment, it should be specifically noted that the steps for obtaining the environmental temperature impact coefficient are as follows: Under the controlled environmental temperature, run the industrial robot and record the performance data in real time. Increase or decrease the environmental temperature and repeat the experiment, and record the performance data of the industrial robot under each temperature condition; Collect the environmental temperature under different experimental conditions, the machine performance data at different temperatures, and the performance data under the standard temperature condition; According to the experimental data, establish a mathematical relationship between the environmental temperature change and the machine performance change through linear regression or other fitting methods; Through regression analysis of the experimental data, calculate the environmental temperature impact coefficient.

[0031] The data correction module is used to correct the operation data according to the real-time environmental temperature to obtain the operation correction data, and transmit the operation correction data to the second fault analysis module; In this embodiment, it should be specifically noted that the steps for correcting the operation data according to the real-time environmental temperature to obtain the operation correction data are as follows: Calculate the corrected motor temperature according to the current motor temperature data and the temperature deviation. The specific acquisition steps are as follows: ; In the formula, represents the corrected motor temperature, represents the current motor temperature data, represents the motor temperature correction coefficient, which is obtained through experiments and represents the motor temperature change caused by each unit temperature change. represents the temperature deviation; Obtain the current vibration data, and calculate the corrected vibration data according to the current vibration data and the temperature deviation. The specific acquisition steps are as follows: ; In the formula, is represented as the corrected vibration data, is represented as the current vibration data, is represented as the vibration data correction coefficient, indicating the change in vibration data caused by each unit temperature change; Obtain the current load data, and calculate the corrected load data based on the current load data and the temperature deviation. The specific obtaining steps are as follows: ; In the formula, is represented as the corrected load data, is represented as the current load data, is represented as the load data correction coefficient, indicating the change in load data caused by each unit ambient temperature change; Integrate the corrected motor temperature, corrected vibration data, and corrected load data to obtain the operation correction data.

[0032] The second fault analysis module is used to evaluate the operation correction data to obtain the actual fault index, and perform a second fault judgment based on the actual fault index. If the second fault judgment determines that the current industrial robot has a fault, a warning prompt is issued to prompt the relevant staff to repair the industrial robot in time.

[0033] Performing a second fault judgment based on the corrected operation data can effectively eliminate the influence of ambient temperature on sensor data, making the fault judgment more accurate. By identifying potential problems during the first fault judgment and then using the corrected data to eliminate the deviation caused by temperature, false triggering or missed fault judgments can be avoided, thereby improving the reliability and accuracy of the fault detection system. The corrected data more truly reflects the state of the industrial robot in the actual working environment, making the fault warning and maintenance decision-making more scientific and effective.

[0034] In this embodiment, it should be specifically noted that the steps for performing a second fault judgment based on the actual fault index are as follows: Compare the actual fault index with the fault threshold. If the actual fault index is less than the fault threshold, it is determined that the current industrial robot has no fault, and the robot data acquisition module is returned to continue collecting operation data; if the actual fault index is greater than or equal to the fault threshold, it is determined that the current industrial robot has a fault.

[0035] Finally: The above are only the preferred embodiments of the present invention and are not used to limit the present invention. Any modifications, equivalent replacements, improvements, etc. made within the spirit and principle of the present invention shall be included within the protection scope of the present invention.

[0036] As described above, it is only the specific implementation manner of the present application, but the protection scope of the present application is not limited thereto. Any person skilled in the art within the technical scope disclosed by the present application can easily think of changes or substitutions, which should all be covered within the protection scope of the present application. Therefore, the protection scope of the present application shall be subject to the protection scope of the claims described above.

Claims

1. An Internet-based industrial robot fault detection system, characterized in that, The system includes: A robot data acquisition module, which is used to set a sensor group on an industrial robot. The sensor group includes a temperature sensor, a vibration sensor, and a force sensor. The operation data of the industrial robot is collected in real time through the sensor group. The operation data includes motor temperature, vibration data, and load data, and the operation data is transmitted to the first fault analysis module; The first fault analysis module is used to evaluate an initial fault index based on the operation data and perform a first fault judgment according to the initial fault index; An ambient temperature acquisition module. If the first fault judgment determines that the industrial robot has a fault, it acquires the real-time ambient temperature at which the industrial robot works and transmits the real-time ambient temperature to the environmental impact judgment module; The environmental impact judgment module is used to perform an environmental impact judgment based on the real-time ambient temperature. If the environmental impact judgment determines that the sensor group is affected by the environment, it transmits the real-time ambient temperature to the data correction module; The data correction module is used to correct the operation data according to the real-time ambient temperature to obtain operation corrected data, and transmit the operation corrected data to the second fault analysis module; The second fault analysis module is used to evaluate the operation corrected data to obtain an actual fault index, and perform a second fault judgment according to the actual fault index. If the second fault judgment determines that the current industrial robot has a fault, it issues a warning prompt to prompt relevant staff to repair the industrial robot in time.

2. The industrial robot fault detection system based on the Internet according to claim 1, characterized in that: The steps for obtaining the initial fault index are as follows: Obtain the current motor temperature data of the industrial robot through the temperature sensor, obtain the normal motor temperature of the industrial robot, and calculate the temperature fault coefficient according to the current motor temperature data and the normal motor temperature; Set a collection time period, and obtain the vibration data of the industrial robot through the vibration sensor within the collection time period, and evaluate the vibration fault coefficient according to the vibration data; Obtain the load data of the industrial robot through the force sensor, and evaluate the load fault coefficient according to the load data; Normalize the temperature fault coefficient, vibration fault coefficient, and load fault coefficient, and comprehensively evaluate the initial fault index according to the normalized temperature fault coefficient, vibration fault coefficient, and load fault coefficient. The specific obtaining steps are as follows: ； Wherein, is expressed as the initial failure index, is expressed as the temperature failure coefficient, is expressed as the vibration failure coefficient, is expressed as the load failure coefficient, , , are expressed as the weight coefficients of the temperature failure coefficient, the vibration failure coefficient, and the load failure coefficient.

3. An Internet-based industrial robot fault detection system according to claim 2, characterized in that: The steps for obtaining the vibration fault coefficient are as follows: Set a collection time period, divide the collection time period evenly into n parts, denoted as sub-collection time periods, and set collection time points within each sub-collection time period; Obtain the vibration data at the collection time points within each sub-collection time period through the vibration sensor, calculate the mean value of the vibration data within the sub-collection time period according to the vibration data at the collection time points, and calculate the fluctuation degree of each sub-collection time period according to the vibration data at the collection time points and the mean value of the vibration data; Perform K-means clustering on the fluctuation degree of each sub-collection time period, and obtain the vibration fault coefficient according to the clustering result.

4. An Internet-based industrial robot fault detection system according to claim 3, characterized in that: The steps for performing K-means clustering on the fluctuation degree of each sub-collection time period are as follows: Step 1: Use the fluctuation degree as the clustering feature, use the fluctuation degrees of all sub-collection time periods as the data set, and each fluctuation degree in the data set is a data point; Step 2: Use the silhouette coefficient method to determine the optimal number of clusters K of the data set; Step 3: Randomly select K data points in the dataset as the initial cluster centers. For each data point, calculate its Euclidean distance to each initial cluster center. For each data point, traverse the K initial cluster centers and assign it to the cluster corresponding to the nearest initial cluster center; Step 4: After traversing all data points, obtain the initial clusters. For each initial cluster, calculate the mean of the data points within it to obtain the new cluster centers; Step 5: Repeat Step 3 and Step 4 until the cluster centers no longer change, obtaining the final clusters and the final cluster centers.

5. An Internet-based industrial robot fault detection system according to claim 4, characterized in that: The steps for obtaining the vibration fault coefficient according to the clustering result are as follows: Calculate the ratio of the number of data points in each final cluster to the total number of data points in the dataset to obtain the weight of each final cluster; Perform a weighted sum of the weights of each final cluster and the final cluster centers to obtain the vibration fault coefficient.

6. The industrial robot fault detection system based on the Internet according to claim 2, wherein: The steps for obtaining the load fault coefficient are as follows: During the acquisition time period, obtain the load data of the industrial robot through a force sensor and calculate the average value of the load data during the acquisition time period; Calculate the standard deviation of the load data based on the average value of the load data, denoted as the load fluctuation degree; Obtain the load threshold of the industrial robot and calculate the load fault coefficient based on the load threshold and the load fluctuation degree.

7. An Internet-based industrial robot fault detection system according to claim 1, characterized in that: The steps for the first fault judgment according to the initial fault index are as follows: Set the fault threshold, compare the initial fault index with the fault threshold. If the initial fault index is less than the fault threshold, it is judged that the current industrial robot has not failed, and return to the robot data acquisition module to continue collecting operation data; if the initial fault index is greater than or equal to the fault threshold, it is judged that the current industrial robot has failed.

8. An Internet-based industrial robot fault detection system according to claim 1, characterized in that: The steps for the environmental impact judgment according to the real-time environmental temperature are as follows: Obtain the standard environmental temperature of the industrial robot, calculate the difference between the real-time environmental temperature and the standard environmental temperature to obtain the temperature deviation; Calculate the environmental impact degree based on the temperature deviation. The specific acquisition steps are as follows: ； In the formula, represents the degree of environmental impact, represents the environmental temperature influence coefficient, represents the temperature deviation; Set the impact threshold, compare the environmental impact degree with the impact threshold. If the environmental impact degree is less than the impact threshold, it is judged that the sensor group is not affected by the environment, the current industrial robot has failed, and a warning prompt is given; If the environmental impact degree is greater than or equal to the impact threshold, it is judged that the sensor group is affected by the environment.

9. An Internet-based industrial robot fault detection system according to claim 7, characterized in that: The steps for the second fault judgment according to the actual fault index are as follows: Compare the actual fault index with the fault threshold. If the actual fault index is less than the fault threshold, it is judged that the current industrial robot has not failed, and return to the robot data acquisition module to continue collecting operation data; If the actual fault index is greater than or equal to the fault threshold, it is judged that the current industrial robot has failed.

Images