Carrier roller operation state abnormity monitoring method and system based on temperature residual error change

Abstract

The invention belongs to the field of preventive maintenance of belt conveyor equipment, and discloses a carrier roller running state abnormity monitoring method and system based on temperature residual variation, and the method comprises the steps: adding a corresponding monitoring module on a support frame of each carrier roller of a belt conveyor line, collecting temperature data, and inputting the temperature data into a model; calculating the acquired temperature data by adopting a similarity measurement residual error calculation method to obtain original data required by abnormal monitoring of the running state of the carrier roller; and carrying out anomaly classification on the obtained data by adopting a support vector machine, inputting a model to obtain a corresponding anomaly monitoring result, and carrying out reporting and early warning processing on the result to realize anomaly monitoring on the running state of the carrier roller. The system has the advantages of intelligent preventive maintenance, real-time performance, man-machine interaction, high environmental interference resistance and high universality.

Description

Technical Field

[0001] The invention belongs to the technical field of preventive maintenance of belt conveyor equipment, and in particular relates to a method and system for monitoring abnormal operation status of rollers based on temperature residual changes. Background Art

[0002] Belt conveyor systems are widely used in various industries due to their simple structure, large transmission capacity and relatively convenient use, and are particularly widely used in mining transportation.

[0003] The belt conveyor has a long conveyor line, which is composed of various connecting supports and a long conveyor belt. If one link of the belt conveyor line is interrupted, the operation of the coal mining face and other workplaces will be interrupted. Therefore, the normal production of the coal mine is directly determined by the operating status of the belt conveyor line.

[0004] The operation and maintenance of traditional belt conveyor lines are entirely dependent on manpower, and the means used are all conventional "knocking and hammering", relying entirely on human experience. This inefficient operation and maintenance method can no longer meet the needs of modern production in coal mines. Therefore, it is necessary to study new technologies or new methods to solve the technical problems of intelligent operation and maintenance of belt conveyor lines.

[0005] With the continuous development of technology, especially the in-depth research and application of AI algorithms and big data analysis technologies, if we can use intelligent algorithms, big data analysis and other technical means to install a series of intelligent sensing devices on the belt conveyor line, collect a series of data representing the operation and maintenance of the belt conveyor line, and use big data analysis technology, AI technology, etc., to monitor the operating status of the important supporting components of the belt conveyor line - rollers in real time, and provide real-time feedback on the results, we can finally realize the function of intelligent operation and inspection of the belt conveyor line.

[0006] In view of the above analysis, the existing technology has the following technical problems that need to be solved urgently:

[0007] The operation and maintenance of traditional belt conveyor lines rely entirely on manpower, and the operation and inspection methods are inefficient. It is necessary to study new technologies or methods to solve the technical difficulties of intelligent operation and inspection of belt conveyor lines. Summary of the invention

[0008] In view of the problems existing in the prior art, the present invention provides a method and system for monitoring the abnormal operation status of rollers based on temperature residual changes.

[0009] The present invention is implemented as follows: a method for monitoring abnormal operation status of a roller based on temperature residual change, the method specifically comprising:

[0010] S1: Add corresponding monitoring modules to the support frame of each roller of the belt conveyor line, collect temperature data, and input them into the model;

[0011] S2: Using the similarity metric residual calculation method, the collected temperature data is calculated to obtain the original data required for abnormal monitoring of the roller operation status;

[0012] S3: Use support vector machine to classify the obtained data, input the model to obtain the corresponding abnormal monitoring results, and report and warn the results to achieve abnormal monitoring of the roller operation status.

[0013] Furthermore, in S1, high-precision temperature data acquisition equipment is installed on both ends of each roller, and real-time data is collected at a sampling frequency of 1 Hz and at a time interval of set time T, and input into the model to calculate the overall operating status analysis of the roller at the current moment.

[0014] Furthermore, in S2, the temperature data of the roller operation status is obtained from the data sensing and acquisition system, and the original data is standardized. The standardized calculation formula is as follows:

[0015]

[0016] Where: X ij Indicates i The jth value of a parameter, i = 1, 2, ..., p, j = 1, 2, ..., m; p Represents the number of data; and S i are the mean and standard deviation respectively. For the P component vectors Z after standardization ij , select its jth value and construct it into a vector Calculate its correlation matrix C:

[0017]

[0018] Then the jth MD value is:

[0019]

[0020] MD is used to process the time domain characteristics of the roller monitoring temperature data, and used together with the previously obtained time domain characteristics to characterize the changes in the health status of the system.

[0021] Furthermore, the main implementation process of S3 is as follows:

[0022] (1) To collect fault data on site, it is necessary to collect fault data of the roller in three states;

[0023] (2) Use support vector machine to construct a multi-fault classifier. The entire fault diagnosis model consists of two stages: learning stage and fault testing stage. When constructing a multi-classification model, a one-to-one support vector machine classification algorithm is selected according to the actual situation;

[0024] (3) In the learning phase, the data is imported into multiple classification support vector machine models. After the relevant parameters are set, the data that has passed S2 can be trained. The training results can be optimized by the accuracy of the training set.

[0025] (4) In the fault diagnosis stage, it is necessary to select test set data from the fault data. A portion of the data needs to be extracted for testing in each state. The test results can be used to optimize the diagnosis model.

[0026] Another object of the present invention is to provide a system for monitoring the abnormal operation state of a roller based on the change of temperature residual, the system specifically comprising:

[0027] Input module, used to collect sample data of each state;

[0028] The preprocessing module is used to normalize the sample data and divide it into training set and test set;

[0029] The classification module uses support vector machine to classify the obtained data into abnormalities.

[0030] In combination with the above technical solutions and the technical problems solved, the advantages and positive effects of the technical solutions to be protected by the present invention are as follows:

[0031] First, the present invention has the advantages of intelligent and real-time preventive maintenance, human-computer interaction, strong resistance to environmental interference, and strong versatility.

[0032] In traditional belt conveyor systems, it is difficult to effectively monitor the operating status of rollers, especially the temperature changes cannot be captured in real time, which easily leads to the inability to detect and handle abnormal conditions in time. Due to problems such as roller wear, bearing damage or lack of lubrication, the temperature rise of the rollers can cause serious failures and even damage or stop the conveyor belt, thus affecting the continuity and safety of the entire production line. However, the existing technology lacks effective monitoring methods, resulting in no early warning when roller failure occurs.

[0033] The present invention solves the problem in the prior art that roller temperature changes cannot be discovered in time and fault monitoring is inaccurate through a monitoring method based on temperature residual changes. A temperature acquisition module is installed on the support frame of each roller to collect temperature data in real time, and a similarity metric residual calculation method is used to accurately analyze temperature changes and determine whether the roller is abnormal. The support vector machine algorithm is introduced to classify abnormalities in residual data, realizing real-time monitoring of the roller operating status, and being able to issue warnings and report in time before a fault occurs, effectively avoiding production problems caused by roller failures.

[0034] Compared with traditional monitoring methods, the present invention has achieved significant technical progress in terms of accuracy, real-time performance and automation. First, by using high-precision temperature sensors to collect data in real time and classifying the roller operating status through residual analysis and support vector machine models, potential faults can be quickly identified. Second, the system's automated early warning function greatly reduces the need for manual intervention and improves the safety and operating efficiency of the conveyor line. More importantly, the method can adapt to temperature fluctuations under different operating conditions and provide a more reliable fault monitoring solution.

[0035] The present invention has broad application prospects in belt conveyor lines in the fields of mining, logistics, chemical industry, etc. Through real-time monitoring of roller temperature and abnormal classification processing, the system can effectively reduce equipment failure rate, reduce unexpected downtime, and improve production efficiency and equipment life. Since the present invention adopts an automated early warning system, it can not only improve safety, but also reduce maintenance costs, further enhance the application value of industrial automation in actual production, and promote the development of intelligent equipment monitoring technology.

[0036] Second, the present invention solves the common problems of poor real-time performance, low monitoring accuracy, and high equipment maintenance costs in the prior art through a method for monitoring the abnormal operation status of rollers based on changes in temperature residuals. In traditional roller monitoring, regular manual inspections or simple vibration monitoring are usually relied upon, which cannot detect potential equipment failures in a timely manner, resulting in unexpected shutdowns of the conveying system and affecting production efficiency. The present invention collects temperature residual data and combines it with support vector machine (SVM) model analysis to monitor the operation status of rollers in real time and accurately, achieving early warning of abnormalities and effectively reducing shutdown events caused by failures.

[0037] Another technical problem in industrial applications is that existing roller monitoring technology often has poor adaptability to complex environments, especially in high-temperature and high-dust working environments such as mines and steel. Traditional sensor equipment is easily disturbed by the external environment, resulting in a decrease in the accuracy of monitoring data. The present invention avoids the sensitivity of traditional vibration monitoring technology to environmental noise by installing high-precision temperature acquisition equipment at both ends of the roller and combining it with temperature residual calculation. At the same time, it monitors the internal faults of the roller through temperature changes, ensuring that the equipment can still operate efficiently and accurately under complex working conditions.

[0038] Through the support vector machine classifier, the present invention solves the technical problem that multiple fault types are difficult to effectively distinguish. Traditional roller monitoring systems can usually only target specific types of faults, such as bearing overheating or mechanical wear, and cannot accurately identify multiple faults. By constructing a multi-classification SVM model, the present invention can simultaneously classify roller faults under different states, and through the optimization of the test set data, the accuracy of fault identification is further improved, making early warning of different types of faults possible.

[0039] In actual industrial applications, the roller monitoring method provided by the present invention significantly improves the real-time monitoring capability of the roller operating status and reduces the maintenance cost in the production process. Since the system can operate automatically and output monitoring data in real time, enterprises can reduce their reliance on manual inspections, thereby reducing the consumption of human resources. At the same time, the data-driven analysis model can also help enterprises optimize equipment operation and maintenance strategies, effectively extend the service life of rollers, and improve the operating efficiency of the entire conveying system.

[0040] In addition, the present invention has shown good scalability and adaptability in application, and can adapt to the roller monitoring needs of different types of conveyor lines and under different working conditions. Whether it is a long-distance mine transportation line or a steel production line with a complex process flow, accurate monitoring of the roller operation status can be achieved by flexibly configuring temperature acquisition equipment and algorithm models. This adaptability makes the present invention have a wide range of application prospects in multiple industries and provides strong technical support for intelligent operation and maintenance under the background of Industry 4.0.

[0041] The present invention solves multiple key technical problems in roller operation status monitoring, realizes intelligent, real-time and precise equipment status monitoring, promotes the widespread application of roller abnormality monitoring technology in mining, metallurgy, ports and other fields, and greatly improves the reliability and economic benefits of industrial transportation systems.

[0042] Third, the present invention monitors the operating status of the roller by introducing temperature residual changes, and adopts multiple key parameters, algorithms and mathematical models to effectively solve the problems of real-time, accuracy and multi-fault identification in the prior art. First, the system installs high-precision temperature acquisition equipment at both ends of each roller, sets the time interval and collects temperature data at a sampling frequency of 1Hz. This parameter setting ensures the real-time performance of temperature data during the operation of the equipment, avoids the problem of equipment failures not being discovered in time due to monitoring lag in traditional inspections, and thus improves the response speed of fault warnings.

[0043] In terms of algorithms, the present invention adopts a residual-based similarity metric algorithm to standardize the collected temperature data and calculate the Mahalanobis distance (MD) to extract the characteristics of the roller's operating status. Through this mathematical model, the influence of data noise and errors on the monitoring results can be effectively eliminated, and the system's sensitivity to small temperature fluctuations is enhanced. Compared with traditional vibration monitoring methods, this processing method is more resistant to interference and is particularly suitable for fault monitoring in complex environments such as high temperature and high dust, solving the problem of the existing technology's strong dependence on the environment.

[0044] In order to further improve the accuracy of the system, the present invention uses support vector machine (SVM) to construct a multi-fault classification model, which divides the roller operation status into three types: normal, minor fault, and major fault. Compared with the traditional single fault monitoring method, the SVM model can classify and predict multiple faults based on the roller temperature residual data. Through model training and test set optimization processing in the learning stage, the algorithm improves the recognition accuracy of different types of faults, effectively solving the problem that the existing technology cannot accurately distinguish multiple faults.

[0045] The present invention has made significant technical progress in temperature data collection, residual calculation, and multi-fault classification algorithms, which not only improves the real-time and accuracy of roller abnormality monitoring, but also realizes accurate classification and early warning of multiple faults through intelligent mathematical models. These technological advances make the system widely applicable in different industrial transportation scenarios and promote the intelligence and automation of roller operation status monitoring. BRIEF DESCRIPTION OF THE DRAWINGS

[0046] Figure 1 It is a flow chart of a method for monitoring abnormal operation status of a roller based on temperature residual change provided by an embodiment of the present invention;

[0047] Figure 2 It is a flowchart of the implementation process of the method for monitoring the abnormal operation state of the roller based on the temperature residual change provided by the embodiment of the present invention;

[0048] Figure 3 is a feature data set test diagram provided by an embodiment of the present invention;

[0049] Figure 4 It is a module diagram of a roller operation state abnormality monitoring system based on temperature residual change provided by an embodiment of the present invention;

[0050] Figure 5 It is a schematic diagram of the installation of the sensing device provided in an embodiment of the present invention. DETAILED DESCRIPTION

[0051] In order to make the purpose, technical solution and advantages of the present invention more clearly understood, the present invention is further described in detail below in conjunction with the embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not used to limit the present invention.

[0052] like Figure 1 , Figure 2 As shown, an embodiment of the present invention provides a method for monitoring the abnormal operation state of a roller based on the change of temperature residual, and the method specifically includes:

[0053] S1: Add corresponding monitoring modules to the support frame of each roller of the belt conveyor line, collect temperature data, and input them into the model;

[0054] S2: Using the similarity metric residual calculation method, the collected temperature data is calculated to obtain the original data required for abnormal monitoring of the roller operation status;

[0055] S3: Use support vector machine to classify the obtained data, and finally input the model to obtain the corresponding abnormal monitoring results, and report and warn the results, finally achieving the goal of abnormal monitoring of the roller running status.

[0056] Monitoring modules are installed on each roller support frame of the belt conveyor line. These monitoring modules collect the operating temperature data of each roller in real time through high-precision temperature sensors. These data are transmitted to the monitoring system through wireless or wired transmission systems. The temperature data of each roller is input into the monitoring model as an input feature, forming the basic basis for operating status evaluation. This step ensures real-time temperature monitoring of all rollers, so that subsequent analysis can be based on the latest temperature changes.

[0057] The collected temperature data is processed by similarity metric residual calculation method. The temperature residual is calculated by comparing the real-time temperature data of each roller with the historical temperature data under normal operating conditions. The residual reflects the abnormal fluctuation of the roller temperature. A large residual value indicates that the roller has abnormal operation. This residual value, as a key indicator, constitutes the original data for abnormal monitoring of the roller operation status, laying the foundation for subsequent abnormal classification and analysis.

[0058] Support vector machine (SVM) is used to classify the calculated temperature residual data for abnormality. SVM is a machine learning algorithm that can effectively distinguish between normal and abnormal roller temperature states. During the model training phase, the system generates classification boundaries by learning a large number of normal and abnormal roller data samples. In actual operation, the model can accurately classify which rollers are in normal state and which rollers are abnormal based on the input real-time temperature residual data. This process realizes real-time monitoring of roller operation status based on temperature residual changes.

[0059] When the support vector machine model detects an abnormal operating temperature of the roller, the system will immediately generate an alarm signal. The signal is reported through the monitoring platform to notify the operator to promptly check and maintain the abnormal roller to avoid equipment failure and downtime. At the same time, the system will also record detailed data of the abnormal event, including the specific location of the roller, temperature changes, and the time when the abnormality occurred, providing a basis for subsequent fault analysis and equipment optimization. Through this early warning mechanism, the safety and operating efficiency of the conveying system can be effectively improved.

[0060] In the S1, high-precision temperature data acquisition equipment is installed on both ends of each roller, with a sampling frequency of 1 Hz and a time interval of set time T to collect real-time data, input the data into the model, and calculate the overall operation status analysis of the roller at the current moment.

[0061] The steps of S2 are as follows:

[0062] The residual processing of the time series of time domain features considers each indicator separately, while the distance measurement of these time domain features considers the decline of the system health status from an overall perspective.

[0063] The temperature data of the roller operating status is obtained from the data sensing and acquisition system, and the original data is standardized. The standardized calculation formula is as follows:

[0064]

[0065] Where: X ij Indicates i The jth value of a parameter, i = 1, 2, ..., p, j = 1, 2, ..., m; p Represents the number of data; and S i are the mean and standard deviation respectively. ij , select its jth value and construct it into a vector Calculate its correlation matrix C:

[0066]

[0067] Then the jth MD value is:

[0068]

[0069] MD can be transformed into the sum of squares of multiple independent standard normal distributions, i.e., chi-square distribution. In this way, the correlation interference between variables can be eliminated, so it can measure the distance between the observed sample and the healthy sample set very well. Therefore, this paper uses MD to process the time domain characteristics of the roller monitoring temperature data, and uses it together with the previously obtained time domain characteristics to characterize the changes in the health status of the system.

[0070] This can obtain the characteristic data set of the abnormal state of the roller, such as Figure 3 shown.

[0071] The main implementation process of S3 is as follows:

[0072] (1) To collect fault data on site, it is necessary to collect fault data of the roller in three states;

[0073] (2) Use support vector machine to construct a multi-fault classifier, which can classify fault features, that is, the classification of three state labels. The entire fault diagnosis model consists of two stages, namely the learning stage and the fault testing stage. When constructing a multi-classification model, according to the actual situation, a one-to-one support vector machine classification algorithm is selected. The basic principle is as follows:

[0074] Select all samples that satisfy y t = t and y s =s sample points, where 1≤s≤k, 1≤t≤k, s≠t, the decision function is as follows:

[0075]

[0076] (3) In the learning phase, the data is imported into multiple classification support vector machine models. After the relevant parameters are set, the data that has passed S2 can be trained. The training results can be optimized by the accuracy of the training set.

[0077] (4) In the fault diagnosis stage, it is necessary to select test set data from the fault data, and a part of the data needs to be extracted for testing in each state. The test results can be used to optimize the diagnosis model.

[0078] like Figure 4 As shown, an embodiment of the present invention provides a roller operation state abnormality monitoring system based on temperature residual change, which specifically includes:

[0079] Input module, used to collect sample data of each state;

[0080] The preprocessing module is used to normalize the sample data and divide it into training set and test set;

[0081] The classification module uses support vector machine to classify the obtained data into abnormalities.

[0082] The schematic diagram of the installation of the sensing device of the present invention is as follows Figure 5 shown.

[0083] Example 1: Monitoring the operating status of rollers in a mine belt conveyor line

[0084] On the belt conveyor line of the mine, the rollers are usually in a high-load working state for a long time. Due to the harsh environment, the rollers are prone to problems such as bearing overheating and wear. The method for monitoring the abnormal operation status of rollers based on temperature residual changes of the present invention installs a high-precision temperature sensor on each roller support frame of the mine belt conveyor line. The sensor collects temperature data in real time at a frequency of 1Hz. Through the similarity metric residual calculation method, the real-time temperature is compared with the historical data during normal operation to monitor the temperature fluctuation of the roller. When the temperature residual of the roller increases abnormally, the support vector machine algorithm will determine it as an abnormal state, and the system will immediately issue an early warning to prompt the maintenance personnel to deal with it in time to prevent damage or shutdown of the conveyor belt due to roller failure. This real-time monitoring system effectively improves the operating safety and maintenance efficiency of mine conveying equipment.

[0085] Example 2: Roller Failure Warning System in the Logistics Industry

[0086] In logistics centers, belt conveyor lines are used to transport large quantities of goods, and conveying efficiency is crucial. In order to ensure the stable operation of the conveying system, the method of the present invention is applied to the temperature monitoring of rollers. High-precision temperature sensors are installed at the support frame of each roller to collect data in real time and analyze the temperature changes of the rollers through the similarity metric residual method. The system uses a support vector machine to analyze and classify the residual data to determine whether there is an abnormal state. Once an abnormal temperature change is detected, the system immediately issues a warning signal and notifies maintenance personnel to inspect and repair to avoid logistics delays caused by abnormal roller temperatures. The system improves the stability of equipment operation in the logistics center, reduces the problem of production line stagnation caused by roller failures, and reduces equipment maintenance costs.

[0087] The following are two specific application examples of the method for monitoring the abnormal operation status of the roller based on the change of temperature residual:

[0088] Application Example 1:

[0089] Roller monitoring system deployment and anomaly detection

[0090] 1. Data collection equipment installation: On a belt conveyor line, according to the implementation plan, high-precision temperature collectors are installed at both ends of each roller. The sampling frequency of these temperature collectors is 1Hz, ensuring that the temperature data of the rollers can be collected in real time.

[0091] 2. Data standardization and residual calculation: The system obtains temperature data in real time through data acquisition equipment, standardizes the data, and uses formulas to calculate the mean and standard deviation of the data. Then, the standardized data is processed using the Mahalanobis distance (MD) to obtain the time domain temperature characteristics of the roller.

[0092] 3. Support vector machine classification model: A multi-fault classifier is constructed by inputting the residual data into the support vector machine (SVM) classification model. First, the roller data is trained in the learning phase using a one-to-one SVM, and then the classifier is optimized and debugged using the test set data under different states.

[0093] 4. Abnormal monitoring and early warning: The system determines the operating status of the roller based on the monitored temperature residual changes and the output results of the classification model. If an abnormality is detected, the system will automatically generate an early warning signal and report it to the operator to ensure that the problem is discovered in time and maintenance measures are taken.

[0094] Application Example 2:

[0095] Large-scale conveyor line roller monitoring and fault prediction

[0096] 1. System installation and debugging: On a mine conveyor line, this monitoring method is used to monitor 200 rollers. A temperature collector with wireless transmission function is installed on the bracket of each roller. The collector collects temperature data once a second and transmits the data to the central control system through the wireless network.

[0097] 2. Temperature data processing and abnormal analysis: Through step S2, a large amount of roller temperature data is obtained, and the data is standardized to construct a data vector matrix. Residual analysis and Mahalanobis distance are used to calculate the operating status characteristics of each roller, and the system monitors the changes in temperature residuals in real time.

[0098] 3. Fault classification and optimization: The system is trained through support vector machines to distinguish three types of roller operation states: normal operation, minor fault, and major fault. Each state is trained with a large amount of historical data for classification models, and in the test phase, part of the data from each fault data is selected for model testing to optimize the accuracy of the classifier.

[0099] 4. Online monitoring and fault prediction: When a significant change in temperature residual is detected or the support vector machine classification result is abnormal, the system immediately sends an early warning signal to the operator. Combined with historical data, the system can also predict the time of failure of certain rollers and arrange maintenance in advance to avoid downtime caused by roller damage.

[0100] These two application examples demonstrate the actual operation of the present invention in different application scenarios, focusing on real-time monitoring of small and medium-sized rollers and anomaly detection and fault prediction in large-scale conveying systems.

[0101] It should be noted that the embodiments of the present invention can be implemented by hardware, software, or a combination of software and hardware. The hardware part can be implemented using dedicated logic; the software part can be stored in a memory and executed by an appropriate instruction execution system, such as a microprocessor or dedicated design hardware. It can be understood by a person of ordinary skill in the art that the above-mentioned devices and methods can be implemented using computer executable instructions and / or contained in a processor control code, such as a carrier medium such as a disk, CD or DVD-ROM, a programmable memory such as a read-only memory (firmware), or a data carrier such as an optical or electronic signal carrier. Such code is provided on the carrier medium. The device and its modules of the present invention can be implemented by hardware circuits such as very large-scale integrated circuits or gate arrays, semiconductors such as logic chips, transistors, etc., or programmable hardware devices such as field programmable gate arrays, programmable logic devices, etc., can also be implemented by software executed by various types of processors, and can also be implemented by a combination of the above-mentioned hardware circuits and software, such as firmware.

[0102] The above description is only a specific implementation mode of the present invention, but the protection scope of the present invention is not limited thereto. Any modifications, equivalent substitutions and improvements made by any technician familiar with the technical field within the technical scope disclosed by the present invention and within the spirit and principle of the present invention should be covered by the protection scope of the present invention.

Claims

1. A method for monitoring abnormal operation status of rollers based on temperature residual changes, characterized in that: The method specifically includes: S1: Add corresponding monitoring modules to the support frame of each roller of the belt conveyor line, collect temperature data, and input them into the model; S2: Using the similarity metric residual calculation method, the collected temperature data is calculated to obtain the original data required for abnormal monitoring of the roller operation status; S3: Use support vector machine to classify the obtained data, input the model to obtain the corresponding abnormal monitoring results, and report and warn the results to achieve abnormal monitoring of the roller operation status.

2. The method for monitoring abnormal operation status of rollers based on temperature residual change according to claim 1, characterized in that: In the S1, high-precision temperature data acquisition equipment is installed on both ends of each roller, with a sampling frequency of 1 Hz and a time interval of set time T to collect real-time data, input the data into the model, and calculate the overall operation status analysis of the roller at the current moment.

3. The method for monitoring abnormal operation status of rollers based on temperature residual change according to claim 1, characterized in that: S2, obtains the temperature data of the roller running status from the data sensing and acquisition system, and standardizes the original data. The standardization calculation formula is as follows: Where: X ij Indicates i The j-th value of a parameter, i = 1, 2, ..., p, j = 1, 2, ..., m; p represents the number of data; and S i are the mean and standard deviation respectively. For the P component vectors Z after standardization ij , select its jth value and construct it into a vector Calculate its correlation matrix C: Then the jth MD value is: MD is used to process the time domain characteristics of the roller monitoring temperature data, and used together with the previously obtained time domain characteristics to characterize the changes in the health status of the system.

4. The method for monitoring abnormal operation status of rollers based on temperature residual change according to claim 1, characterized in that: The main implementation process of S3 is as follows: (1) To collect fault data on site, it is necessary to collect fault data of the roller in three states; (2) Use support vector machine to construct a multi-fault classifier. The entire fault diagnosis model consists of two stages: learning stage and fault testing stage. When constructing a multi-classification model, a one-to-one support vector machine classification algorithm is selected according to the actual situation; (3) In the learning phase, the data is imported into multiple classification support vector machine models. After the relevant parameters are set, the data that has passed S2 can be trained. The training results can be optimized by the accuracy of the training set. (4) In the fault diagnosis stage, it is necessary to select test set data from the fault data. A portion of the data needs to be extracted for testing in each state. The test results can be used to optimize the diagnosis model.

5. A roller operation status abnormality monitoring system based on temperature residual change as described in claims 1-4, characterized in that: The system specifically includes: The temperature data acquisition module is installed at the support frame of each roller and is used to collect the temperature data of the roller in real time at a sampling frequency of 1 Hz; The data processing module is used to standardize the collected temperature data and obtain the original data for abnormal monitoring of the roller operation status through the similarity metric residual calculation method; Anomaly classification module, used to use support vector machine to perform anomaly classification on the standardized temperature data and generate anomaly monitoring results; The early warning and reporting module is used to receive the abnormal monitoring results generated by the abnormal classification module, trigger the alarm and perform reporting processing.

6. The roller operation state abnormality monitoring system based on temperature residual change according to claim 5 is characterized in that: The temperature data acquisition module includes two high-precision temperature sensors, which are respectively installed at the two ends of each roller. The temperature sensors are connected to the data processing module to collect the temperature data of the roller at a set time interval T and transmit the data to the data processing module.

7. The roller operation state abnormality monitoring system based on temperature residual change according to claim 5 is characterized in that: The data processing module is used to standardize the collected temperature data according to the following standardized calculation formula, and the obtained standardized data is used to construct a multidimensional vector and further generate a correlation matrix, and calculate the Mahalanobis distance (MD) to characterize the health status of the roller.

8. The roller operation state abnormality monitoring system based on temperature residual change according to claim 5 is characterized in that: The abnormal classification module adopts a one-to-one multi-classification support vector machine model. In the model training stage, it is trained by inputting data of normal and various fault states. In the fault diagnosis stage, a test set is extracted from the collected fault data, the state of the roller is classified, and the corresponding abnormal monitoring results are generated.

Images