Cloud-edge collaborative digital twin-based state monitoring method for idler bearings of belt conveyor

Abstract

Disclosed in the present invention is a cloud-edge collaborative digital twin-based state monitoring method for idler bearings of a belt conveyor. The method comprises: acquiring operation data of an idler bearing of an underground belt conveyor; transmitting the operation data to an edge device and a cloud-based central server; constructing a geometric model and a three-dimensional simulation model in the cloud-based central server, constructing a fault diagnosis model on the edge device, and using the geometric model, the three-dimensional simulation model and the fault diagnosis model to form a digital twin; and on the basis of measured data, updating the three-dimensional simulation model in the digital twin in real time, using incremental learning to update the fault diagnosis model in the digital twin in real time, and using the fault diagnosis model to determine an operation state of the idler bearing of the belt conveyor. In the present invention, computing resources are allocated on the basis of a cloud-edge collaborative algorithm to realize the rapid computation of a digital twin, thereby improving the virtual-physical interaction capability of the digital twin, and ultimately realizing the digital twin-based state monitoring, early-warning and visual analysis of idler bearings of a belt conveyor.

Description

Cloud edge collaborative belt conveyor roller bearing digital twin state monitoring method TECHNICAL FIELD

[0001] The present application relates to the technical field of mine equipment fault monitoring, in particular to a cloud edge collaborative belt conveyor roller bearing digital twin state monitoring method. BACKGROUND

[0002] The underground belt conveyor is the main tool for mine transportation, and the roller bearing, as the "joint" in the underground belt conveyor, is numerous, bears heavy load, and works in a harsh environment, so it is prone to failure after a long time of operation, which seriously affects the production efficiency and economic benefits of the mine. Therefore, monitoring the running state of the underground belt conveyor roller bearing and timely processing the running faults have great practical significance for the safe and efficient production of the mine.

[0003] Digital twin refers to constructing a mapping body in virtual space that can reflect the real condition of the mechanical equipment, reflecting the space-time evolution of the key performance parameters of the mechanical equipment based on virtual-real mapping and virtual-real evolution methods. At the technical level, the digital twin technology has the potential to monitor the running state of the roller bearing in real time. However, the digital twin-based belt conveyor roller bearing state monitoring method faces the following challenges: 1) The traditional cloud computing method transmits sensor data to a central cloud server for centralized processing, and the single cloud computing method will cause the underground communication network to be overloaded, causing data transmission delay in the underground, and affecting the real-time data interaction between the roller bearing digital twin virtual body and the physical entity; 2) The operation of the roller bearing has the characteristics of time-varying working conditions and non-constant evolution of performance, and the construction of the roller bearing digital twin body has high requirements for computing resources and real-time computing.

[0004] In summary, under the constraints of communication and computing resources, it is difficult to implement a digital twin-based belt conveyor roller bearing state monitoring method. SUMMARY

[0005] The present application aims to provide a cloud edge collaborative belt conveyor roller bearing digital twin state monitoring method, which realizes real-time data interaction between the digital twin body and the physical entity based on the cloud edge collaborative method, and further constructs a roller bearing digital twin body with iterative updating function, and reflects the whole process of roller bearing state evolution in virtual space.

[0006] Technical solution: The cloud edge collaborative belt conveyor roller bearing digital twin state monitoring method of the application comprises the following steps: Step 1, obtaining the operation data of the underground belt conveyor roller bearing through the sensor arranged on the underground belt conveyor; Step 2, sending the operation data to the router through the ZigBee network by the ZigBee terminal node, transmitting the operation data to the coordinator node by the router through the ZigBee network, sending the operation data to the edge device through the 5G network by the coordinator node, and sending the operation data to the cloud center server through the 5G network by the edge device; Step 3, constructing the geometric model and three-dimensional simulation model of the roller bearing in the cloud center server, constructing the fault diagnosis model on the edge device, and utilizing the geometric model, three-dimensional simulation model and fault diagnosis model of the roller bearing to constitute the digital twin; Step 4, combining the measured data, updating the three-dimensional simulation model in the digital twin in real time, further updating the fault diagnosis model in the digital twin in real time by utilizing the incremental learning, utilizing the fault diagnosis model to determine the operation state of the belt conveyor roller bearing, and performing corresponding early warning and reminding.

[0007] Further, the monitoring method further comprises the following steps: Step 5, real-time monitoring the operation state of the belt conveyor roller bearing group through the remote monitoring platform.

[0008] Further, in step 1, the sensor comprises a temperature sensor, a vibration sensor, a bearing force sensor, a rotating speed sensor and a load sensor; and the operation data comprises roller bearing temperature data, roller bearing vibration data, pressure data of the roller bearing stress surface, roller rotating speed and load.

[0009] Further, before step 2, it further comprises: assembling a plurality of tree-shaped topological ZigBee networks, each ZigBee network comprising a coordinator node, a plurality of routers and a plurality of ZigBee terminal nodes, and each ZigBee terminal node comprising a single-chip microcomputer and a ZigBee communication module.

[0010] Further, in step 2, the process of sending the operation data to the router through the ZigBee network by the ZigBee terminal node comprises: converting the operation data from a physical quantity into an electrical analog signal by the sensor, sending the electrical analog signal to the single-chip microcomputer, converting the electrical analog signal into a digital signal by the single-chip microcomputer, and transmitting the digital signal to the router through the ZigBee communication module.

[0011] Further, in step 3, the three-dimensional simulation model of the roller bearing is constructed by the roller bearing geometric model and the operation data, and the three-dimensional simulation model comprises a temperature field model, a vibration modal model and a stress distribution model.

[0012] Further, in step 3, after constructing the fault diagnosis model on the edge device, further comprising: training the fault diagnosis model; the process of training is: running the three-dimensional simulation model in the cloud center server in real time, inputting the roller bearing speed information and load information of the belt conveyor in real time into the three-dimensional simulation model, generating virtual temperature signals, force signals and vibration signals under different speeds and different loads by using the three-dimensional simulation model, establishing a simulation signal library D1, and storing it in the cloud center server; training the fault diagnosis model by using the simulation signal library D1, and the mathematical expression of the fault diagnosis model is: f=F(x, θ i ), wherein f is the fault diagnosis result, x is the roller bearing temperature, vibration, roller speed and load running data sample, θ i is the fault diagnosis model parameter, and F is the fault diagnosis model.

[0013] Further, the number of edge devices is 2, which are respectively placed at the edge side of the belt conveyor, and are respectively the first edge device and the second edge device, and the two edge devices are used to calculate the tasks generated by the fault diagnosis model.

[0014] Further, the calculation of the tasks generated by the fault diagnosis model by using the two edge devices comprises: firstly, according to the computing resources on the first edge device, all or part of the tasks are unloaded to the device for calculation, if there are remaining calculation tasks, the remaining calculation tasks are unloaded to the second edge device through the 5G network, and the results are returned to the first edge device after the calculation is completed.

[0015] Secondly, if the first edge device does not receive the calculation results returned by the second edge device after completing the calculation task, the remaining calculation task part in the second edge device is unloaded to the first edge device, so that the first edge device and the second edge device complete all the remaining calculation tasks at the same time.

[0016] Further, in step 4, real-time updating the three-dimensional simulation model and the fault diagnosis model in the digital twin comprises: according to the actual measurement results of the roller bearing temperature, vibration, force, roller speed and load, etc., the simulation signal library D1 is updated in real time in the cloud center server by using the three-dimensional simulation model; collecting the bearing entity temperature signal, vibration signal and force signal under the normal running state of the belt conveyor, establishing an entity signal library D2, and storing it to the cloud center server; calculating the similarity of the corresponding signals of the simulation signal library D1 and the entity signal library D2 in the cloud center server by using the Pearson correlation coefficient, and transmitting the similarity calculation result to the edge device through 5G; if the similarity is greater than the threshold value, using the fault diagnosis model to judge the fault, otherwise judging that a new type of fault occurs.

[0017] Further, when a new type of fault occurs, the collected new type of fault signal is stored in the new category database D new , D new Stored in the edge device, and the entity signal library D2 and the old category data in the simulation signal library D1 as an example set, using the iCaRL (Incremental Classifier and Representation Learning) incremental learning method, based on the new category database D new The category incremental update of the fault diagnosis model is carried out, and the updated fault diagnosis model F new Is calculated; The loss function of incremental update is as follows: L=(1-λ))L new +λL KD , wherein L is the loss function during training, the value range of the hyperparameter λ is [0, 1], L new It is the loss function of the original fault diagnosis model F and the updated fault diagnosis model F new On the new category database D new , L KD It is the distillation loss of the original fault diagnosis model F and the updated fault diagnosis model F new On the example set.

[0018] Further, the running state of the belt conveyor roller bearing is judged by using the fault diagnosis model, and corresponding early warning is carried out, including: when the fault diagnosis model in the edge device diagnoses the fault, the fault early warning is issued; When a new type of fault is identified, an incremental warning of a new type of fault is issued.

[0019] Advantages: compared with the prior art, the present application has the following advantages: 1, the present application constructs the digital twin of the belt conveyor roller bearing, including geometric model, three-dimensional simulation model and fault diagnosis model, realizes the remote real-time health state monitoring of all the belt conveyor roller bearings by using the digital twin, improves the intelligent degree of the roller bearing state monitoring, and ensures that the health state of the underground belt conveyor roller bearing is timely, accurately and comprehensively presented; 2, the present application forms a ZigBee and 5G combined communication network, reduces the underground communication network load, reduces the underground data transmission delay, realizes the real-time data interaction between the digital twin virtual body and the physical entity of the belt conveyor roller bearing; 3, the present application uses cloud edge collaborative computing technology, reasonably allocates the computing capacity of the cloud center server and the edge device, arranges the fault diagnosis model on the edge device close to the roller bearing, and arranges the three-dimensional simulation model on the cloud center server, improves the belt conveyor operation and maintenance efficiency; 4, the present application uses incremental learning technology, updates the digital twin of the belt conveyor roller bearing in real time, improves the iterative updating performance of the digital twin, and improves the fault diagnosis accuracy. BRIEF DESCRIPTION OF DRAWINGS

[0020] Fig. 1 is a flow chart of the cloud-edge collaborative belt conveyor roller bearing digital twin state monitoring method in the embodiment; Fig. 2 is a schematic diagram of a communication network fusing 5G and ZigBee in the embodiment; Fig. 3 is a flow chart of cloud-edge collaborative computing of the digital twin in the embodiment; and Fig. 4 is a flow chart of digital twin updating iteration in the embodiment. DETAILED DESCRIPTION

[0021] In order to make the purposes, technical solutions and advantages of the present application clearer, the present application is further described in detail below with reference to the accompanying drawings and embodiments.

[0022] The cloud-edge collaborative belt conveyor roller bearing digital twin state monitoring method described in the embodiment has the flow chart as shown in Fig. 1, and includes the following steps: Step 1, obtaining the operation data of the underground belt conveyor roller bearing through the sensors arranged on the underground belt conveyor; Step 2, sending the operation data to the router through the ZigBee network by the ZigBee terminal node, transmitting the operation data to the coordinator node by the ZigBee network by the router, sending the operation data to the edge device by the 5G network by the coordinator node, and sending the operation data to the cloud center server by the 5G network by the edge device; Step 3, constructing the geometric model and the three-dimensional simulation model of the roller bearing in the cloud center server, constructing the fault diagnosis model on the edge device, and using the geometric model, the three-dimensional simulation model and the fault diagnosis model of the roller bearing to constitute the digital twin; Step 4, updating the three-dimensional simulation model in the digital twin in real time in combination with the measured data, further updating the fault diagnosis model in the digital twin in real time by using incremental learning, using the fault diagnosis model to determine the operation state of the belt conveyor roller bearing, and performing corresponding early warning and reminding.

[0023] In one example, the cloud-edge collaborative belt conveyor roller bearing digital twin state monitoring method further includes the following step: Step 5, monitoring the operation state of the group of belt conveyor roller bearings in real time through the remote monitoring platform.

[0024] Specifically, in Step 1, the sensors include temperature sensors, vibration sensors, bearing force sensors and rotational speed sensors; the operation data includes temperature data, vibration data, force data, rotational speed and load data, and the roller bearing temperature data, the roller bearing vibration data, the pressure data of the roller bearing force bearing surface, the roller rotational speed and the load data during the operation of the underground belt conveyor are obtained through the temperature sensors, the vibration sensors, the bearing force sensors, the rotational speed sensors and the load sensors, respectively.

[0025] The cloud-edge collaborative belt conveyor roller bearing digital twin state monitoring method described in the embodiment further comprises, before step 2: organizing multiple tree topology ZigBee networks, each ZigBee network comprising a coordinator node, multiple routers and multiple ZigBee terminal nodes, each ZigBee terminal node comprising a CC2530 single-chip microcomputer and a ZigBee communication module.

[0026] As shown in FIG. 2, the sensor sends the operation data of the downhole belt conveyor roller bearing to the single-chip microcomputer, which sends the operation data to the router through the ZigBee communication module, 1 router is connected with multiple terminal nodes (in this example, 1 router corresponds to 3 terminal nodes is taken as an example for illustration), the router is connected with the coordinator node through the 5G intelligent gateway, and the router sends the operation data to the coordinator node through the 5G intelligent gateway under the ZigBee network. The coordinator node sends to the edge device through the 5G network, and the edge device sends to the cloud center server through the 5G network. Each coordinator node is connected with a corresponding edge device, each edge device is connected through the 5G network, and all edge devices are connected with the same cloud center server. Through the fusion networking of the tree networking ZigBee network and the 5G network, the data acquired by the ZigBee terminal node is transmitted to the edge device, and the communication between the edge device and the cloud center server is maintained.

[0027] In step 2, the process of sending the operation data to the router through the ZigBee network by the ZigBee terminal node comprises: converting the operation data from a physical quantity into an electrical analog signal by the sensor, and sending the electrical analog signal to the single-chip microcomputer; the single-chip microcomputer converts the electrical analog signal into a digital signal and transmits it to the router through the ZigBee communication module.

[0028] In step 3, the three-dimensional simulation model of the roller bearing is constructed by the roller bearing geometric model and the operation data, and the three-dimensional simulation model comprises a temperature field model, a vibration modal model and a stress distribution model.

[0029] Specifically, the geometric model G of the roller bearing is constructed in the cloud center server by using a geometric three-dimensional scanner and solidworks v , and further according to the acquired other operation information, a three-dimensional simulation model R of the downhole belt conveyor roller bearing is constructed v , and the three-dimensional simulation model comprises a temperature field model, a vibration modal model and a stress distribution model.

[0030] The geometric model, the three-dimensional simulation model and the fault diagnosis model constitute a digital twin, as shown in the following formula: VE=(G v , R v, F), wherein VE is a digital twin, VE includes G v , R v , and F model, F is a fault diagnosis model. In an example, the fault diagnosis model F can be selected as a ResNet32 network. By using the three-dimensional simulation model R v integrated in the cloud center server and the fault diagnosis model F at the edge device end, data interaction between the cloud and the edge can be realized, the running state of the belt conveyor roller bearing underground can be mapped in real time, including the temperature field, stress distribution, vibration mode and fault information, and the digital twin of the belt conveyor roller bearing underground can be constructed. The cloud center server is the computing center of the whole system, which is used for dynamic simulation of the roller bearing, and remote health state monitoring of all roller bearings of the belt conveyor underground is realized through a visualization platform. The edge device is close to the roller bearing, the signal communication delay is low, and the signals generated by the roller bearing can be diagnosed in time, and the fault diagnosis model can be incrementally learned when a new type of fault occurs.

[0031] In step 3, after constructing the fault diagnosis model on the edge device, the fault diagnosis model is further trained. The training process is as follows: the three-dimensional simulation model is run in real time on the cloud center server, the roller speed information and load information of the belt conveyor during real-time operation are input into the three-dimensional simulation model, the three-dimensional simulation model is used to generate virtual temperature signals, force signals and vibration signals under different speeds and loads, a simulation signal library D1 is established, and stored in the cloud center server; the fault diagnosis model is trained by using the simulation signal library D1, and the mathematical expression of the fault diagnosis model is: f=F(x, θ i ), wherein f is a fault diagnosis result, x is a roller bearing temperature, vibration, and roller speed and load running data sample, θ i is a fault diagnosis model parameter, and F is a fault diagnosis model.

[0032] The number of edge devices is 2, which are placed at the edge side of the belt conveyor, respectively, as a first edge device and a second edge device, and the two edge devices are used to calculate the tasks generated by the fault diagnosis model.

[0033] As shown in FIGS. 3-4, the real-time updating of the three-dimensional simulation model and the fault diagnosis model in the digital twin in step 4 includes: according to the actual measurement results of the roller bearing temperature, vibration, force, and roller rotational speed and load, etc., the simulation signal library D1 is updated in real time in the cloud center server by using the three-dimensional simulation model; the bearing entity temperature signal, force signal, and vibration signal under the normal operation state of the belt conveyor are collected, the entity signal library D2 is established, and stored to the cloud center server; the similarity of the corresponding signals in the simulation signal library D1 and the entity signal library D2 is calculated in the cloud center server by using the Pearson correlation coefficient, and the similarity calculation result is transmitted to the edge device through 5G; if it is greater than the threshold value, the fault diagnosis model is used for fault judgment, otherwise a new type of fault is determined. In the example, the threshold value can be set to 0.6.

[0034] Further, when a new type of fault occurs, the collected new type of fault signal is stored in the new category database D new , D new is stored in the edge device. The old category data in the entity signal library D2 and the simulation signal library D1 are used as the example set, and the iCaRL (Incremental Classifier and Representation Learning) incremental learning method is used to incrementally update the fault diagnosis model based on the new category database D new , and the updated fault diagnosis model F new is calculated; the loss function of the incremental update is as follows: L = (1-λ))L new +λL KD , wherein L is the loss function during training, the value range of the hyperparameter λ is [0, 1], L new is the loss function of the updated fault diagnosis model F new on the new category database D new , and L KD is the distillation loss of the original fault diagnosis model F and the updated fault diagnosis model F new on the example set.

[0035] Further, the edge computing has the characteristics of real-time and low latency, which can ensure the real-time of the fault diagnosis and life prediction of the belt conveyor, and the tasks generated by the fault diagnosis model are calculated by using two edge devices, including: first, according to the computing resources on the first edge device I1, all or part of the tasks are offloaded to the device for calculation, if there are remaining calculation tasks, the remaining calculation tasks are offloaded to the second edge device I2 through the 5G network, and the result is returned to the first edge device I1 after the calculation is completed.

[0036] Secondly, if the first edge device I1 does not receive the calculation result returned by the second edge device I2 after completing the calculation task, the remaining calculation task in the second edge device I2 is partially offloaded to the first edge device I1, so that the first edge device I1 and the second edge device I2 complete all the remaining calculation tasks at the same time.

[0037] Further, the running state of the belt conveyor roller bearing is judged by using the fault diagnosis model, and a corresponding early warning is given, including: when the fault diagnosis model in the edge device diagnoses a fault, a fault early warning is given; when a new type of fault is identified, a new type of fault incremental early warning is given.

[0038] Further, a remote monitoring platform of the underground belt conveyor roller bearing is established in the cloud center server to monitor the running state of the roller group in real time.

[0039] Through the remote monitoring platform, the network and real-time running state of each bearing in the underground belt conveyor can be viewed; the historical fault information of each bearing and the historical fault information of the entire underground belt conveyor can be viewed; the digital twin simulation model and simulation data of various fault types, and various health state data in the entity signal library can be viewed, and the signal data can be manually added, deleted and modified.

Claims

1. A cloud-edge collaborative belt conveyor idler bearing digital twin condition monitoring method, characterized in that, The method comprises the following steps: Step 1: obtaining operation data of the belt conveyor roller bearing through sensors arranged on the downhole belt conveyor; Step 2: sending the operation data to the router through the ZigBee network by using the ZigBee terminal node, transmitting the operation data to the coordinator node by the router through the ZigBee network, sending the operation data to the edge device through the 5G network by the coordinator node, and sending the operation data to the cloud center server through the 5G network by the edge device; Step 3: constructing a geometric model and a three-dimensional simulation model of the roller bearing in the cloud center server, constructing a fault diagnosis model on the edge device, and using the geometric model, the three-dimensional simulation model and the fault diagnosis model of the roller bearing to form a digital twin; Step 4: combining the measured data, updating the three-dimensional simulation model in the digital twin in real time, further updating the fault diagnosis model in the digital twin in real time by using incremental learning, using the fault diagnosis model to determine the operation state of the belt conveyor roller bearing, and performing corresponding early warning and reminding.

2. The cloud-edge collaborative belt conveyor roller bearing digital twin condition monitoring method of claim 1, wherein, Further comprising the following steps: Step 5: real-time monitoring the operation state of the belt conveyor roller bearing group through the remote monitoring platform.

3. The cloud-edge collaborative belt conveyor roller bearing digital twin condition monitoring method of claim 1, wherein, In step 1, the sensors include temperature sensors, vibration sensors, bearing force sensors, speed sensors and load sensors; the operation data includes roller bearing temperature data, roller bearing vibration data, roller bearing stress surface pressure data, roller speed and load.

4. The cloud-edge collaborative belt conveyor roller bearing digital twin condition monitoring method of claim 3, wherein, Before step 2, further comprising: assembling a plurality of tree-shaped topological ZigBee networks, each ZigBee network comprising a coordinator node, a plurality of routers and a plurality of ZigBee terminal nodes, and each ZigBee terminal node comprising a single-chip microcomputer and a ZigBee communication module.

5. The cloud-edge collaborative belt conveyor roller bearing digital twin condition monitoring method of claim 4, wherein, In step 2, the process of sending the operation data to the router through the ZigBee network by using the ZigBee terminal node comprises: Using the sensors to convert the operation data from physical quantities into electrical analog signals, and sending the electrical analog signals to the single-chip microcomputer, the single-chip microcomputer converts the electrical analog signals into digital signals, and transmits the digital signals to the router through the ZigBee communication module.

6. The cloud-edge collaborative belt conveyor roller bearing digital twin condition monitoring method of claim 5, wherein, In step 3, the three-dimensional simulation model of the roller bearing is constructed by the geometric model of the roller bearing and the operation data, and the three-dimensional simulation model comprises a temperature field model, a vibration modal model and a stress distribution model.

7. The cloud-edge collaborative belt conveyor idler bearing digital twin condition monitoring method of claim 6, wherein, In step 3, after constructing the fault diagnosis model on the edge device, further comprising: training the fault diagnosis model; The training process is as follows: Running the three-dimensional simulation model in the cloud center server, inputting the roller speed information and load information of the belt conveyor during real-time operation into the three-dimensional simulation model, generating virtual temperature signals, force signals and vibration signals under different speeds and different loads by using the three-dimensional simulation model, establishing a simulation signal library D1, and storing the simulation signal library D1 in the cloud center server; Training the fault diagnosis model by using the simulation signal library D1, and the mathematical expression of the fault diagnosis model is: f = F(x, θ i ), Wherein, f is the fault diagnosis result, x is the roller bearing temperature, vibration, and roller speed and load running data samples, θ i is the fault diagnosis model parameter, and F is the fault diagnosis model.

8. The cloud-edge collaborative belt conveyor idler bearing digital twin condition monitoring method of claim 7, wherein, The number of edge devices is 2, which are placed at the edge side of the belt conveyor, respectively, and are respectively a first edge device and a second edge device, and the two edge devices are used to calculate the tasks generated by the fault diagnosis model.

9. The cloud-edge collaborative belt conveyor idler bearing digital twin condition monitoring method of claim 8, wherein, The computing of the tasks generated by the fault diagnosis model by the two edge devices comprises: First, according to the computing resources on the first edge device, all or part of the tasks are offloaded to the device for computing, if there are remaining computing tasks, the remaining computing tasks are offloaded to the second edge device through the 5G network, and the results are returned to the first edge device after the computing is completed; Second, if the first edge device does not receive the computing results returned by the second edge device after completing the computing task, the remaining computing tasks in the second edge device are partially offloaded to the first edge device, so that the first edge device and the second edge device complete all the remaining computing tasks at the same time.

10. The cloud-edge collaborative belt conveyor idler bearing digital twin condition monitoring method of claim 9, wherein, The real-time updating of the three-dimensional simulation model and the fault diagnosis model in the digital twin in step 4 comprises: According to the actual measurement results of the roller bearing temperature, vibration, force, and roller rotational speed and load, the three-dimensional simulation model is used to update the simulation signal library D1 in the cloud center server in real time; The bearing entity temperature signal, vibration signal and force signal under the normal operation state of the belt conveyor are collected, the entity signal library D2 is established, and stored to the cloud center server; The similarity of the corresponding signals in the simulation signal library D1 and the entity signal library D2 is calculated in the cloud center server by using the Pearson correlation coefficient, and the similarity calculation result is transmitted to the edge device through the 5G network; if the similarity is greater than the threshold value, the fault diagnosis model is used for fault discrimination, and if the similarity is less than the threshold value, a new type of fault is discriminated.

11. The cloud-edge collaborative belt conveyor idler bearing digital twin condition monitoring method of claim 10, wherein, When a new type of fault occurs, the collected new type of fault signal is stored in the new category database D new , and the entity signal library D2 and the old category data in the simulation signal library D1 are used as an example set, and the iCaRL incremental learning method is used to update the fault diagnosis model based on the new category database D new , and the old category data in the simulation signal library D1 are used as an example set, and the iCaRL incremental learning method is used to update the fault diagnosis model based on the new category database D new , and the old category data in the simulation signal library D1 are used as an example set, and the iCaRL incremental learning method is used to update the fault diagnosis model based on the new category database D new , and the old category data in the simulation signal library D1 are used as an example set, and the iCaRL incremental learning method is used to update the fault diagnosis model based on the new category database D KD , and the old category data in the simulation signal library D1 are used as an example set, and the iCaRL incremental learning method is used to update the fault diagnosis model based on the new category database D Wherein, L is the loss function during training, the value range of hyperparameter λ is [0, 1], L new is the updated fault diagnosis model F new The loss function L new on the new class database D KD is the original fault diagnosis model F and the updated fault diagnosis model F new The distillation loss on the example set.

12. The cloud-edge collaborative belt conveyor idler bearing digital twin condition monitoring method of claim 11, wherein, The use of the fault diagnosis model to discriminate the running state of the belt conveyor roller bearing and to give corresponding early warning comprises: When the fault diagnosis model in the edge device diagnoses a fault, a fault early warning is given; when a new type of fault is discriminated, a new type of fault incremental early warning is given.

Images