Collaborative Management Method and System Based on Digital Twin for Maglev Power Equipment

Abstract

This disclosure relates to a collaborative management method and system based on digital twins for magnetic levitation power equipment. The method is applied to a collaborative management system. Terminal devices acquire environmental data, equipment operating parameters, and control commands. Based on the control commands, and assuming the terminal device is the one to be digitally twinned, the environmental data and equipment operating parameters are uploaded to the corresponding edge server. The edge server trains a digital twin model based on the environmental data and equipment operating parameters, and sends the local model parameters corresponding to the trained digital twin model to a central cloud server. This digital twin model is used to digitally twin the terminal device to be twinned, resulting in a digital twin corresponding to the terminal device. The central cloud server aggregates the local model parameters sent by at least one edge server to obtain the global model parameters corresponding to the digital twin model. This method combines a collaborative management system with digital twin technology.

Description

TECHNICAL FIELD

[0001] The present disclosure relates to the technical field of computers, and particularly relates to a collaborative management method and system based on digital twinning for magnetic suspension power equipment. BACKGROUND

[0002] Digital technology is a key element of data resources, a main carrier of modern information networks, and an important driving force for the integration of information communication technology applications and full-factor digital transformation, which is promoting profound changes in production methods, lifestyles, and governance methods. Digital technology is an important means to realize the construction of new capabilities of manufacturing enterprises, such as the ability of new business models such as online diagnosis, the ability of product life cycle tracking, and the ability of online off-site collaborative development.

[0003] Magnetic suspension power equipment is a technology industry, which is a general term for a series of magnetic suspension products, mainly including magnetic suspension blowers, magnetic suspension high-speed motors, magnetic suspension bearings, magnetic suspension refrigerant compressors, magnetic suspension high-speed atomizers, magnetic suspension machining spindles, magnetic suspension electronic printing rollers, and magnetic suspension turbines, and other industrial clusters formed by means of magnetic suspension technology. Magnetic suspension power equipment has the characteristics of no contact friction, no energy loss, and high-speed operation, but it needs to maintain a very high speed while working. Once the bearing loses speed, not only will it cause damage to the machine, but also it will bring huge safety hazards. Therefore, the production, testing, and operation and maintenance of magnetic suspension power equipment need a smart, efficient, and safe collaborative management system. SUMMARY

[0004] To overcome the problems in the related art, the present disclosure provides a collaborative management method and system based on digital twinning for magnetic suspension power equipment.

[0005] According to a first aspect of an embodiment of the present disclosure, a collaborative management method based on digital twinning for magnetic suspension power equipment is provided, which is applied to a collaborative management system including a plurality of terminal devices, edge servers corresponding to the terminal devices, and a central cloud server. The terminal devices at least include magnetic suspension power equipment and end devices. The method includes:

[0006] The terminal device acquires corresponding environment data, device operating parameters, and control instructions, and determines whether the terminal device is a terminal device to be twinned according to the control instructions;

[0007] In the case that the terminal device is a terminal device to be twinned, the terminal device uploads the environment data and device operating parameters corresponding to the terminal device to the corresponding edge server;

[0008] The edge server trains a digital twin model in the edge server based on the environment data and the device running parameters corresponding to the terminal device, and sends local model parameters corresponding to the trained digital twin model to the central cloud server, the digital twin model being used for digital twinning of a terminal device to be twinned to obtain a digital twin corresponding to the terminal device.

[0009] The central cloud server aggregates the local model parameters sent by at least one of the edge servers to obtain global model parameters corresponding to the digital twin model, and distributes the global model parameters to at least one of the edge servers, so that the edge server updates the digital twin model based on the received global model parameters.

[0010] In some embodiments, the central cloud server includes a plan management module, and the method further includes:

[0011] The plan management module in the central cloud server manages the terminal device based on order data, the order data including customer relationship data, customer demand data, manufacturing plan data, or supply chain data.

[0012] In some embodiments, the collaborative management system further includes a data acquisition and monitoring system, and the terminal device acquires corresponding environment data, device running parameters, and control instructions, including:

[0013] The central cloud server determines the terminal device to be twinned according to the order data, and sends the control instruction to the data acquisition and monitoring system, the control instruction being used to indicate whether the terminal device is a terminal device to be twinned.

[0014] The data acquisition and monitoring system receives the control instruction, acquires the environment data and the device running parameters corresponding to the terminal device, and sends the corresponding environment data, device running parameters, and control instruction to the terminal device.

[0015] In some embodiments, the data acquisition and monitoring system includes an acquisition and monitoring module, a data processing module, an interface management module, and an architecture module, the data acquisition and monitoring system receiving the control instruction, acquiring the environment data and the device running parameters corresponding to the terminal device, including:

[0016] The acquisition and monitoring module monitors the terminal device and acquires the environment data and the device running parameters corresponding to the terminal device;

[0017] The data processing module receives the control instruction;

[0018] The interface management module determines a target communication protocol;

[0019] The architecture module interacts with other devices according to the target communication protocol.

[0020] In some embodiments, the edge server trains a digital twin model in the edge server based on the environment data and the device running parameters corresponding to the terminal device, including:

[0021] The edge server obtains the digital twin model from the central cloud server at the beginning of the tth iteration training, t is an integer greater than 1;

[0022] The edge server trains the digital twin model based on the environment data and the device running parameters corresponding to the terminal device using a gradient descent algorithm.

[0023] In some embodiments, the digital twin model is represented as:

[0024]

[0025] wherein w i (t) represents the digital twin model in the ith edge server in the tth iteration training, w(t-1) represents the digital twin model trained by the plurality of edge servers in the (t-1)th iteration training, η represents a learning rate, represents training w(t-1) using a gradient descent algorithm.

[0026] In some embodiments, during the iteration training of the digital twin model, the device energy consumption of the terminal device is:

[0027]

[0028] wherein, represents device energy consumption, α represents an energy consumption coefficient, ξ i represents the number of CPU cycles required to execute one data unit, D i represents the environment data and the device running parameters of the terminal device corresponding to the ith edge server, and the represents the CPU cycle frequency;

[0029] The computing time of the machine interface of the terminal device is:

[0030]

[0031] wherein, represents computing time, ξ i represents the number of CPU cycles required to execute one data unit, D i represents the environment data and the device running parameters of the terminal device corresponding to the ith edge server, and the represents the CPU cycle frequency;

[0032] The transmission time of the terminal device is:

[0033]

[0034] wherein, represents the transmission time, |w i represents the size of the local model parameter, r i represents the transmission speed;

[0035] The transmission energy consumption of the terminal device is:

[0036]

[0037] wherein, represents the transmission energy consumption, β represents the transmission energy consumption coefficient, P i represents the transmission power, |w i represents the size of the local model parameter, r i represents the transmission speed.

[0038] In some embodiments, the collaborative management system further comprises a base station, and the central cloud server aggregates the local model parameters sent by at least one of the edge servers to obtain global model parameters corresponding to the digital twin model, comprising:

[0039] The aggregation of the local model parameters sent by the central cloud server to at least one of the edge servers is represented as:

[0040]

[0041] wherein, w(t) represents the aggregated digital twin model, D g represents the device information of the base station, D i represents the environmental data and device operating parameters of the terminal device corresponding to the i-th edge server, w i represents the digital twin model of the i-th edge server in the t-th iteration training, and N represents the number of edge servers.

[0042] In some embodiments, the method further comprises:

[0043] The central cloud server stores the digital twin corresponding to the terminal device, and different digital twins are virtually connected in communication;

[0044] If the physical communication between different terminal devices fails, the different digital twins corresponding to the terminal devices in the central cloud server are virtually communicated.

[0045] In some embodiments, the method further comprises:

[0046] The edge server models the terminal device based on the device operating principle and the device shape of the terminal device.

[0047] In some embodiments, the method further comprises:

[0048] The edge server sets the terminal device as a node in the blockchain.

[0049] In some embodiments, the magnetic levitation power equipment includes a magnetic levitation device and a production device, and the terminal device includes an Internet of Things device, a sensor, and a camera.

[0050] In some embodiments, 5G wireless communication is used between the edge server and the central cloud server.

[0051] According to a second aspect of the embodiments of the present disclosure, a collaborative management system based on digital twinning for magnetic levitation power equipment is provided, which includes a plurality of terminal devices, edge servers corresponding to the terminal devices, and a central cloud server, the terminal devices at least including a magnetic levitation power equipment and a terminal device:

[0052] The terminal device is configured to acquire corresponding environment data, device operating parameters, and control instructions, and determine whether the terminal device is a terminal device to be twinned according to the control instructions.

[0053] The terminal device is configured to upload the environment data and the device operating parameters corresponding to the terminal device to the corresponding edge server if the terminal device is a terminal device to be twinned.

[0054] The edge server is configured to train a digital twinning model in the edge server based on the environment data and the device operating parameters corresponding to the terminal device, and send local model parameters corresponding to the trained digital twinning model to the central cloud server, the digital twinning model being used to digitally twin the terminal device to be twinned to obtain a digital twin corresponding to the terminal device.

[0055] The central cloud server is configured to aggregate the local model parameters sent by at least one of the edge servers to obtain global model parameters corresponding to the digital twinning model, and distribute the global model parameters to at least one of the edge servers, so that the edge server updates the digital twinning model based on the received global model parameters.

[0056] In some embodiments, the central cloud server comprises a plan management module, the plan management module in the central cloud server is configured to manage the terminal device based on order data, the order data comprising customer relationship data, customer demand data, manufacturing plan data or supply chain data.

[0057] In some embodiments, the collaborative management system further comprises a data acquisition and monitoring system.

[0058] The central cloud server is configured to determine the terminal device to be twinned according to the order data, and send the control instruction to the data acquisition and monitoring system, the control instruction being used to indicate whether the terminal device is a terminal device to be twinned.

[0059] The data acquisition and monitoring system is configured to receive the control instruction, acquire the environment data and the device running parameter corresponding to the terminal device, and send the corresponding environment data, device running parameter and control instruction to the terminal device.

[0060] In some embodiments, the data acquisition and monitoring system comprises an acquisition and monitoring module, a data processing module, an interface management module and an architecture module.

[0061] The acquisition and monitoring module is configured to monitor the terminal device and acquire the environment data and the device running parameter corresponding to the terminal device.

[0062] The data processing module is configured to receive the control instruction.

[0063] The interface management module is configured to determine a target communication protocol.

[0064] The architecture module is configured to interact with other devices according to the target communication protocol.

[0065] In some embodiments, the edge server is configured to acquire the digital twin model from the central cloud server at the beginning of the tth iteration training, t being an integer greater than 1; and train the digital twin model based on the environment data and the device running parameter corresponding to the terminal device using a gradient descent algorithm.

[0066] In some embodiments, the digital twin model is represented as:

[0067]

[0068] wherein w i (t) represents the digital twin model of the tth iteration training in the ith edge server, w(t-1) represents the digital twin model trained by the plurality of edge servers in the (t-1)th iteration training, and η represents a learning rate. represents training by using a gradient descent algorithm on w (t-1).

[0069] In some embodiments, during one iteration of training of the digital twin model, the device energy consumption of the terminal device is:

[0070]

[0071] wherein, represents device energy consumption, a represents an energy consumption coefficient, and i represents the number of CPU cycles required to execute one data unit, D i represents the environment data and device running parameters of the terminal device corresponding to the i-th edge server, and the represents the CPU cycle frequency;

[0072] The computing time of the machine interface of the terminal device is:

[0073]

[0074] wherein, represents computing time, and i represents the number of CPU cycles required to execute one data unit, D i represents the environment data and device running parameters of the terminal device corresponding to the i-th edge server, and the represents the CPU cycle frequency;

[0075] The transmission time of the terminal device is:

[0076]

[0077] wherein, represents transmission time, and i represents the size of the local model parameters, and r i represents the transmission speed;

[0078] The transmission energy consumption of the terminal device is:

[0079]

[0080] wherein, represents transmission energy consumption, and i represents the transmission power, and i represents the size of the local model parameters, and r i represents the transmission speed.

[0081] In some embodiments, the collaborative management system further comprises a base station, the central cloud server, and the central cloud server is configured to aggregate the local model parameters sent by at least one of the edge servers to represent a global model parameter as follows:

[0082]

[0083] wherein w(t) represents the aggregated digital twin model, D g represents the device information of the base station, D i represents the environment data and the device operation parameters of the terminal device corresponding to the i-th edge server, w i (t) represents the digital twin model of the i-th edge server in the t-th iteration training, and N represents the number of edge servers.

[0084] In some embodiments, the central cloud server is configured to store the digital twins corresponding to the terminal devices, and the digital twins of different terminal devices are virtually connected; and when the physical communication between different terminal devices fails, the digital twins corresponding to the different terminal devices in the central cloud server are virtually communicated.

[0085] In some embodiments, the edge server is configured to model the terminal device based on the device operation principle and the device shape of the terminal device.

[0086] In some embodiments, the edge server is configured to set the terminal device as a node in a blockchain.

[0087] In some embodiments, the magnetic levitation power equipment comprises a magnetic levitation device and a production device, and the terminal device comprises an Internet of Things device, a sensor, and a camera.

[0088] In some embodiments, the edge server and the central cloud server adopt 5G wireless communication.

[0089] According to a third aspect of the embodiments of the present disclosure, a non-transitory computer-readable storage medium is provided, when instructions in the storage medium are executed by a processor of an electronic device, the electronic device is enabled to perform the collaborative management method according to any one of the first aspect of the embodiments of the present disclosure.

[0090] The above method of the present disclosure has the following beneficial effects:

[0091] The method provided by the embodiment of the present disclosure is a collaborative management method for magnetic levitation power equipment based on digital twinning. A terminal device acquires environment data, device operation parameters, and control instructions, and according to the control instructions, in a case where the terminal device is a terminal device to be twinned, the environment data and the device operation parameters are uploaded to a corresponding edge server. The edge server trains a digital twinning model in the edge server based on the environment data and the device operation parameters, and sends local model parameters corresponding to the trained digital twinning model to a central cloud server. The digital twinning model is used to perform digital twinning on the terminal device to be twinned, to obtain a digital twin corresponding to the terminal device. The central cloud server aggregates the local model parameters sent by at least one edge server to obtain global model parameters corresponding to the digital twinning model. The method combines a collaborative management system and digital twinning technology, and uses the terminal device, the edge server, and the central cloud server to cooperatively perform virtual-real mapping on the terminal device, that is, to perform digital twinning on the terminal device, thereby reducing the pressure of information transmission, processing, and storage.

[0092] It should be understood that the above general description and the following detailed description are only exemplary and explanatory, and cannot limit the present disclosure. BRIEF DESCRIPTION OF DRAWINGS

[0093] The accompanying drawings, which are incorporated into and form part of the specification, illustrate an embodiment consistent with the present disclosure and, together with the specification, serve to explain the principles of the present disclosure.

[0094] Figure 1 FIG. 1 is a schematic diagram of a collaborative management system for magnetic levitation power equipment based on digital twinning according to an exemplary embodiment;

[0095] Figure 2 FIG. 2 is a flowchart of a collaborative management method for magnetic levitation power equipment based on digital twinning according to an exemplary embodiment;

[0096] Figure 3 FIG. 3 is a flowchart of a collaborative management method for magnetic levitation power equipment based on digital twinning according to an exemplary embodiment;

[0097] Figure 4 FIG. 4 is a schematic diagram of a system management system for magnetic levitation power equipment based on digital twinning according to an exemplary embodiment. DETAILED DESCRIPTION

[0098] The exemplary embodiments will be described in detail herein with reference to the attached drawings. In the following description, like reference numerals refer to like elements, unless the context clearly dictates otherwise. The following description of exemplary embodiments is not representative of all possible embodiments consistent with the present application. Instead, it is merely an example of apparatus and methods consistent with some aspects of the present application as detailed in the appended claims.

[0099] Figure 1 is a schematic diagram of a collaborative management system for magnetic levitation power equipment based on digital twinning according to an exemplary embodiment, see Figure 1 The collaborative management system includes a plurality of terminal devices, edge servers corresponding to the terminal devices, and a central cloud server. The terminal devices include at least the magnetic levitation power equipment and the end devices.

[0100] In the embodiments of the present disclosure, the terminal device is configured to obtain corresponding environment data, device operating parameters, and control instructions, and determine whether the terminal device is a to-be-twinning terminal device according to the control instructions. The terminal device is further configured to upload the environment data and the device operating parameters corresponding to the terminal device to the corresponding edge server in the case that the terminal device is the to-be-twinning terminal device. The edge server is configured to train a digital twinning model in the edge server based on the environment data and the device operating parameters corresponding to the terminal device, and send local model parameters corresponding to the trained digital twinning model to the central cloud server. The digital twinning model is configured to perform digital twinning on the to-be-twinning terminal device to obtain a digital twin corresponding to the terminal device. The central cloud server is configured to aggregate the local model parameters sent by at least one edge server to obtain global model parameters corresponding to the digital twinning model, and distribute the global model parameters to at least one edge server, so that the edge server updates the digital twinning model based on the received global model parameters.

[0101] In some embodiments, the central cloud server includes a plan management module. The plan management module in the central cloud server is configured to manage the terminal devices based on order data, which includes customer relationship data, customer demand data, manufacturing plan data, or supply chain data.

[0102] In some embodiments, the collaborative management system further includes a data acquisition and monitoring system. The central cloud server is configured to determine the to-be-twinning terminal device according to the order data, and send a control instruction to the data acquisition and monitoring system. The control instruction is used to indicate whether the terminal device is the to-be-twinning terminal device. The data acquisition and monitoring system is configured to receive the control instruction, acquire the environment data and the device operating parameters corresponding to the terminal device, and send the corresponding environment data, device operating parameters, and control instructions to the terminal device.

[0103] In some embodiments, the data acquisition and monitoring system comprises an acquisition and monitoring module, a data processing module, an interface management module and an architecture module; the acquisition and monitoring module is used for monitoring the terminal device and acquiring the environmental data and the device running parameter corresponding to the terminal device; the data processing module is used for receiving the control instruction; the interface management module is used for determining the target communication protocol; and the architecture module is used for interacting with other devices according to the target communication protocol.

[0104] In some embodiments, the edge server is used for acquiring the digital twin model from the central cloud server at the beginning of the tth iteration training, t is an integer greater than 1; and the digital twin model is trained by using the gradient descent algorithm based on the environmental data and the device running parameter corresponding to the terminal device.

[0105] In some embodiments, the digital twin model is expressed as:

[0106]

[0107] wherein, w i (t) represents the digital twin model of the tth iteration training in the ith edge server, w(t-1) represents the digital twin model trained by the plurality of edge servers in the (t-1)th iteration training, η represents the learning rate, represents that the gradient descent algorithm is used to train w(t-1).

[0108] In some embodiments, during the iteration training of the digital twin model, the device energy consumption of the terminal device is:

[0109]

[0110] wherein, represents the device energy consumption, α represents the energy consumption coefficient, ξ i represents the number of CPU cycles required for executing one data unit, D i represents the environmental data and the device running parameter of the terminal device corresponding to the ith edge server, represents the CPU cycle frequency;

[0111] The calculation time of the machine interface of the terminal device is:

[0112]

[0113] wherein, represents the calculation time, ξ i represents the number of CPU cycles required for executing one data unit, D i represents the environmental data and the device running parameter of the terminal device corresponding to the ith edge server, represents the CPU cycle frequency;

[0114] The transmission time of the terminal device is:

[0115]

[0116] wherein, denotes the transmission time, |w i denotes the size of the local model parameter, r i denotes the transmission speed;

[0117] The transmission energy consumption of the terminal device is:

[0118]

[0119] wherein, denotes the transmission energy consumption, β denotes the transmission energy consumption coefficient, P i denotes the transmission power, |w i denotes the size of the local model parameter, r i denotes the transmission speed.

[0120] In some embodiments, the collaborative management system further comprises a base station, a central cloud server, and an edge server for aggregating the local model parameters sent by the at least one edge server, denoted as:

[0121]

[0122] wherein, w(t) denotes the aggregated digital twin model, D g denotes the device information of the base station, D i denotes the environmental data and device operating parameters of the terminal device corresponding to the i-th edge server, w i denotes the digital twin model of the i-th edge server in the t-th iteration training, and N denotes the number of edge servers.

[0123] In some embodiments, the central cloud server is used to store the digital twins corresponding to the terminal devices, and the different digital twins are virtually connected in communication; if the physical communication between different terminal devices fails, the different digital twins corresponding to the terminal devices in the central cloud server are virtually communicated.

[0124] In some embodiments, the edge server is used to model the terminal device based on the device operating principle and the device shape of the terminal device.

[0125] In some embodiments, the edge server is used to set the terminal device as a node in the blockchain.

[0126] In some embodiments, the magnetic levitation power equipment comprises a magnetic levitation device and a production device, and the terminal device comprises an Internet of Things device, a sensor, and a camera.

[0127] In some embodiments, 5G wireless communication is adopted between the edge server and the central cloud server.

[0128] The collaborative management system for the magnetic levitation power equipment based on digital twinning provided by the embodiments of the present disclosure is described below in detail.

[0129] The above Figure 1 The above Figure 2 The above Figure 3 The above

[0130] Figure 2 is a flowchart of a collaborative management method for magnetic levitation power equipment based on digital twinning according to an exemplary embodiment, referring to Figure 2 The method is applied to a collaborative management system, which includes a plurality of terminal devices, edge servers corresponding to the terminal devices, and a central cloud server. The terminal devices at least include magnetic levitation power equipment and end devices. The method includes the following steps:

[0131] In step S201, the terminal device acquires corresponding environmental data, device operating parameters, and control instructions, and determines whether the terminal device is a terminal device to be twinned according to the control instructions.

[0132] The environment data is data describing an environment where the terminal device is located, for example, the environment data is temperature, humidity and the like in the current environment. The device running parameter is data describing a running condition of the terminal device, for example, the device running parameter is a displacement amount of a rotor, a current computing capability, a current order situation and the like. The control instruction is an instruction indicating whether the terminal device is a twin terminal device. The environment data, the device running parameter and the control instruction can be obtained by other devices and sent to the terminal device.

[0133] In step S202, when the terminal device is a twin terminal device, the terminal device uploads environment data and device running parameters corresponding to the terminal device to a corresponding edge server.

[0134] In the embodiments of the present disclosure, in order to alleviate the computing pressure and reduce the communication delay, an edge server is provided for each terminal device for data updating. Therefore, after it is determined that the terminal device is a twin terminal device, the environment data and the device running parameter corresponding to the terminal device are uploaded to the corresponding edge server.

[0135] In step S203, the edge server trains a digital twin model in the edge server based on the environment data and the device running parameter corresponding to the terminal device, and sends local model parameters corresponding to the trained digital twin model to a central cloud server. The digital twin model is used for digital twinning of the terminal device to be twinned, and a digital twin corresponding to the terminal device is obtained.

[0136] The edge server stores a digital twin model. The edge server trains the digital twin model based on the received environment data and device running parameter, and sends local model parameters corresponding to the trained digital twin model to the central cloud server. The digital twin model is used for digital twinning of the terminal device to be twinned. The digital twinning refers to a virtual device, which is equivalent to a backup of data of the terminal device.

[0137] In step S204, the central cloud server aggregates the local model parameters sent by the at least one edge server to obtain global model parameters corresponding to the digital twin model, and distributes the global model parameters to the at least one edge server, so that the edge server updates the digital twin model based on the received global model parameters.

[0138] Since the embodiments of the present disclosure involve multiple edge servers, the environment parameters and device running parameters used for training the digital twin model of each edge server are corresponding to the terminal device, and cannot use the environment parameters and device running parameters of other terminal devices for training. Therefore, after the edge server obtains the trained digital twin model, the edge server sends the local model parameters corresponding to the trained digital twin model to the center cloud server, so that the center cloud server comprehensively considers the training results of multiple edge servers to obtain global model parameters, and then the center cloud server distributes the global model parameters to the edge server. The edge server can perform next iteration training on the digital twin model based on the obtained global model parameters.

[0139] It should be noted that the embodiments of the present disclosure only illustrate the process of training the digital twin model. In another embodiment, after the digital twin model is trained, the environment data and device running parameters corresponding to the terminal device are processed based on the trained digital twin model to obtain the digital twin corresponding to the terminal device.

[0140] The method for cooperative management of magnetic levitation power equipment based on digital twinning provided by the embodiments of the present disclosure, the terminal device acquires environment data, device running parameters and control instructions, and according to the control instructions, in the case of determining that the terminal device is a terminal device to be twinned, the environment data and device running parameters are uploaded to the corresponding edge server; the edge server trains the digital twin model in the edge server based on the environment data and device running parameters, and sends the local model parameters corresponding to the trained digital twin model to the center cloud server, the digital twin model is used for digital twinning of the terminal device to be twinned, and the digital twin corresponding to the terminal device is obtained; the center cloud server aggregates the local model parameters sent by at least one edge server to obtain global model parameters corresponding to the digital twin model. The method combines the cooperative management system and the digital twinning technology, and uses the terminal device, the edge server and the center cloud server to mutually cooperate to perform virtual-real mapping on the terminal device, that is, to perform digital twinning on the terminal device, thereby reducing the pressure of information transmission, processing and storage.

[0141] Figure 3 is a flow chart of a method for cooperative management of magnetic levitation power equipment based on digital twinning according to an exemplary embodiment, see Figure 3 The method is applied to a cooperative management system, and the cooperative management system includes multiple terminal devices, edge servers corresponding to the terminal devices, a center cloud server and a data acquisition and monitoring system. The terminal device at least includes a magnetic levitation power equipment and a terminal device. The magnetic levitation power equipment includes a magnetic levitation device and a production device. The terminal device includes an Internet of Things device, a sensor and a camera, etc. The method includes the following steps:

[0142] In step S301, the center cloud server determines a terminal device to be twinned according to order data, and sends a control instruction to a data acquisition and monitoring system, the control instruction being used to indicate whether the terminal device is the terminal device to be twinned.

[0143] The order data includes customer relationship data, customer demand data, manufacturing plan data, or supply chain data, wherein the customer relationship data refers to customer relationship management (CRM), the customer demand data refers to customer demand, and the customer relationship data and the customer demand data can be data input by a user; the manufacturing plan data and the supply chain data can be automatically generated according to a rotor displacement of the terminal device, a current computing capability, a current order arrangement situation, and the like, and the supply chain data can also be referred to as supply chain management (SCM), and the manufacturing plan data can also be referred to as enterprise resource planning (ERP). Alternatively, an ERP system is used to build a whole-process production and manufacturing management mode covering customer demand, manufacturing plan, manufacturing process, and material supply.

[0144] In some embodiments, the center cloud server determines a terminal device with a large current customer demand and a large current order arrangement as the terminal device to be twinned. Other manners can also be used to determine the terminal device to be twinned, and the specific implementation of determining the terminal device to be twinned is not limited in the embodiments of the present disclosure.

[0145] In addition, in the embodiments of the present disclosure, after the terminal device to be twinned is determined, a control instruction needs to be sent to a data acquisition and monitoring system (SCADA) so that the data acquisition and monitoring system can acquire corresponding environment data and device running parameters of the terminal device according to the control instruction.

[0146] In step S302, the data acquisition and monitoring system receives the control instruction, acquires the environment data and the device running parameters of the terminal device, and sends corresponding environment data, device running parameters, and the control instruction to the terminal device.

[0147] The data acquisition and monitoring system is a system used to acquire data, and the data acquisition and monitoring system can include a sensor and other components used to acquire data.

[0148] In some embodiments, the data acquisition and monitoring system is connected to the terminal device through a WiFi or LoRa link protocol, and the environment data, the device running parameters, and the control instruction acquired by the data acquisition and monitoring system are sent to the terminal device through the WiFi or LoRa link protocol.

[0149] In some embodiments, the data acquisition and monitoring system comprises an acquisition and monitoring module, a data processing module, an interface management module and an architecture module, the data acquisition and monitoring system receives a control instruction, acquires environmental data and device operation parameters corresponding to a terminal device, comprising: the acquisition and monitoring module monitors the terminal device and acquires environmental data and device operation parameters corresponding to the terminal device; the data processing module receives the control instruction; the interface management module determines a target communication protocol; and the architecture module interacts with other devices according to the target communication protocol.

[0150] Among them, the data acquisition and monitoring system acquires data of Internet of Things devices, sensors, industrial cameras, PLC / DSC control systems in the magnetic levitation power equipment intelligent manufacturing, monitors variable configuration and management, and multi-level monitoring of devices / units / workshops. The target communication protocol can be Modbus / TCP, Distributed Network Protocol 3 (DNP3) and Ethernet / IP communication protocol. The architecture module uses OPC UA unified architecture to perform information modeling, service packaging and object interaction, supports the development of various production and operation modules and plugs into the platform to run. The architecture module can realize the discovery, registration, interaction and security mechanism of the model, realize the dynamic loading, plug and play, request-response mode communication, publish-subscribe mode communication and event handling mechanism, realize the comprehensive fusion of production data and the seamless integration of various control functions, and support phased construction and continuous improvement. The data processing module uses spark related processing technology, uses Mysql, Orcale and other databases, and uses information middleware kafka to efficiently transmit instructions and information.

[0151] Step S303, the terminal device receives the corresponding environmental data, device operation parameters and control instruction, and determines whether the terminal device is a twin-to-be terminal device according to the control instruction. If the terminal device is a twin-to-be terminal device, the terminal device uploads the corresponding environmental data and device operation parameters to the corresponding edge server.

[0152] In the embodiments of the present disclosure, in order to ensure the safety of environmental data and device operation parameters, and save the local computing amount of the terminal device, a corresponding edge server is set for each terminal device, and the edge server is used to process the data corresponding to the terminal device.

[0153] In some embodiments, the edge server comprises a digital twin module, and the terminal device uploads the corresponding environmental data and device operation parameters to the digital twin module in the corresponding edge server, and the digital twin module is used for digital twinning of the terminal device.

[0154] In step S304, the edge server trains the digital twin model in the edge server based on the environment data and the device running parameter corresponding to the terminal device, and sends local model parameters corresponding to the trained digital twin model to the central cloud server.

[0155] The digital twin model is used for digital twinning of the terminal device to be twinned to obtain a digital twin corresponding to the terminal device.

[0156] In some embodiments, the edge server trains the digital twin model in the edge server based on the environment data and the device running parameter corresponding to the terminal device, including: the edge server obtains the digital twin model from the central cloud server at the beginning of the tth iteration training, t is an integer greater than 1; the edge server trains the digital twin model based on the environment data and the device running parameter corresponding to the terminal device using a gradient descent algorithm.

[0157] In some embodiments, the digital twin model is represented as:

[0158]

[0159] wherein w i (t) represents the digital twin model in the ith edge server in the tth iteration training, w(t-1) represents the digital twin model trained by the plurality of edge servers in the (t-1)th iteration training, η represents a learning rate, which means training w(t-1) using a gradient descent algorithm.

[0160] In some embodiments, during one iteration training of the digital twin model, the device energy consumption of the terminal device is:

[0161]

[0162] wherein, represents the device energy consumption, α represents an energy consumption coefficient, ξ i represents the number of CPU cycles required to execute one data unit, D i represents the environment data and the device running parameter of the terminal device corresponding to the ith edge server, represents the CPU cycle frequency;

[0163] The calculation time of the machine interface of the terminal device is:

[0164]

[0165] wherein, represents the calculation time, ξ i represents the number of CPU cycles required to execute one data unit, D ienvironment data and device running parameters of a terminal device corresponding to the i-th edge server, CPU cycle frequency;

[0166] The transmission time of the terminal device is:

[0167]

[0168] wherein, transmission time, |w i size of the local model parameters, r i transmission speed;

[0169] The transmission energy consumption of the terminal device is:

[0170]

[0171] wherein, transmission energy consumption, β represents a transmission energy consumption coefficient, P i transmission power, |w i size of the local model parameters, r i transmission speed.

[0172] In some embodiments, the digital twin model is trained using a gradient descent algorithm, and the gradient is For a positive number L, it is uniformly Lipschitz continuous, that is:

[0173]

[0174] wherein, w t+1 the digital twin model trained in the t+1-th iteration, w t the digital twin model trained in the t-th iteration.

[0175] In addition, considering that the objective function F(w) has strong convexity with a parameter μ, and is twice continuously differentiable, the following can be obtained:

[0176]

[0177] Since F(w) is strongly convex, for any w, the following can also be obtained:

[0178]

[0179] In the case of w=w t

[0180]

[0181] By from and ​Subtract F(w) from both sides * From this, we can obtain E[F(w(t+1))-F(w)] * )], and E[F(w(t+1))-F(w * The following relationship must be satisfied:

[0182]

[0183] Therefore, federated learning algorithms converge to the optimal global model during training, even if the digital twin model trained on the edge server reaches its optimal state.

[0184] In some embodiments, the edge server models the terminal device based on its operating principle and physical appearance. The model results in a virtual device that is similar to the terminal device in terms of operating principle and physical appearance. After creating a digital twin of the terminal device, the data of the terminal device is backed up to the virtual device, thus obtaining the digital twin of the terminal device.

[0185] In some embodiments, 5G wireless communication is used between the edge server and the central cloud server. Of course, other communication methods can also be used, and this disclosure does not limit this.

[0186] In some embodiments, the edge server also determines the update frequency of the digital twin model and sets the terminal device as a node in the blockchain to prevent the data or model of the terminal device from being leaked or tampered with.

[0187] It should be noted that, in some embodiments, the steps performed by the edge server described above can be performed by the digital twin module in the edge server.

[0188] In this embodiment of the disclosure, the edge server undertakes scheduling, control, diagnosis, and prediction tasks with high latency requirements.

[0189] In step S305, the central cloud server aggregates the local model parameters sent by at least one edge server to obtain the global model parameters corresponding to the digital twin model, and sends the global model parameters to at least one edge server so that the edge server updates the digital twin model based on the received global model parameters.

[0190] In some embodiments, the collaborative management system further includes a base station, and the central cloud server aggregates the local model parameters sent by at least one edge server to obtain global model parameters corresponding to the digital twin model, including:

[0191] The central cloud server aggregates the local model parameters sent by at least one edge server and represents them as follows:

[0192]

[0193] wherein w(t) represents the aggregated digital twin model, D g represents the device information of the base station, D i represents the environmental data and device operating parameters of the terminal device corresponding to the i-th edge server, w i (t) represents the digital twin model of the i-th edge server in the t-th iteration training, and N represents the number of edge servers.

[0194] The above process is repeated until the minimized local model parameters w(t) satisfy the following relationship:

[0195]

[0196] wherein F g (w) represents the optimal objective function, F i (w) represents the objective function corresponding to the i-th edge server.

[0197] In some embodiments, the central cloud server includes a plan management module, and the plan management module in the central cloud server manages the terminal device based on order data, the order data including customer relationship data, customer demand data, manufacturing plan data, or supply chain data. Wherein, managing the terminal device refers to generating control instructions, training the digital twin model, and determining whether to perform digital twinning on the terminal device.

[0198] In step S306, the edge server performs digital twinning on the plurality of terminal devices based on the trained digital twin model to obtain a digital twin corresponding to each terminal device, and sends the obtained digital twin to the central cloud server, and the central cloud server stores the digital twin corresponding to the terminal device, and the different digital twins are virtually connected; if the physical communication between different terminal devices fails, the central cloud server performs virtual communication based on the digital twin corresponding to the different terminal devices.

[0199] In some embodiments, the terminal device performs wired / wireless communication between the objects (P2P Communications). The digital twins in the central cloud server can be connected through virtual-to-virtual communication (V2V Communications), and build a digital twin network between each other to complete the self-optimization and evolution of the maglev dynamic equipment intelligent manufacturing.

[0200] In some embodiments, WISE-PaaS provides a variety of programming language development tools, stores them into a database in the central cloud server, and is ready to use the application running on the terminal device as a data source at any time. Hadoop is used as a big data base, spark deltaLake is used to build a database, clickhouse is used to realize a data mart, data of Mysql, Orcale and other databases are obtained, and spark related processing technology is used for processing.

[0201] In some embodiments, the central cloud server also provides data modeling tools, object model service tools, computing engines, planning engines, rule engines, etc., for central scheduling, control, diagnosis and prediction work.

[0202] In addition, referring to Figure 4 The collaborative management system includes cloud, edge and terminal, where the cloud refers to the central cloud server (cloud server), the edge refers to the edge server, and the terminal refers to the terminal device (magnetic levitation device). The cloud, edge and terminal are collaboratively processed based on the steps of the above embodiments, realizing digital twin and cloud-edge-terminal collaborative management of the magnetic levitation power equipment intelligent manufacturing scene, from Figure 4 It can be seen that each terminal device can communicate with each other, the terminal device and the edge server can communicate with each other, and the edge server and the central cloud server can communicate with each other.

[0203] The collaborative management method for the magnetic levitation power equipment based on digital twin provided by the embodiments of the present disclosure, the terminal device obtains environment data, device running parameters and control instructions, and according to the control instructions, in the case of determining that the terminal device is a to-be-twin terminal device, the environment data and the device running parameters are uploaded to the corresponding edge server; the edge server trains a digital twin model in the edge server based on the environment data and the device running parameters, and sends local model parameters corresponding to the trained digital twin model to the central cloud server, the digital twin model is used for digital twin of the to-be-twin terminal device, and a digital twin corresponding to the terminal device is obtained; the central cloud server aggregates the local model parameters sent by at least one edge server to obtain global model parameters corresponding to the digital twin model. The method combines the collaborative management system and the digital twin technology, and uses the terminal device, the edge server and the central cloud server to collaboratively map the virtual and real terminal device, i.e., to perform digital twin on the terminal device, thereby reducing the pressure of information transmission, processing and storage.

[0204] And, the production and service management and control of the intelligent maglev power equipment are carried out by using a cloud edge architecture, and the equipment that needs to be digitally twinned is selected by issuing an instruction through a cloud server in the face of planning, necessary virtual-real mapping is carried out, virtual diagnosis, verification and self-evolution are realized, the pressure of information transmission, processing and storage is reduced, and the safety of the maglev equipment data is ensured and the requirement of decentralized decision is met by combining the block chain and federated learning technology.

[0205] The embodiment of the present disclosure further provides a non-transitory computer-readable storage medium, when instructions in the storage medium are executed by a processor of an electronic device, the electronic device can execute the collaborative management method based on digital twinning of the maglev power equipment in the above-mentioned embodiments.

[0206] Other embodiments of the application will be apparent to those skilled in the art from consideration of the specification and practice of the application disclosed herein. This application is intended to cover any variations, uses or adaptations of the application following the principles of the application and including such departures from the present disclosure as come within known use or custom in the art. It is intended to include all such variations and modifications in keeping with the scope of the application as claimed and the principle and true spirit of the disclosure.

[0207] It should be understood that the application is not limited to the precise construction that has been described above and shown in the accompanying drawings, and that various modifications and changes can be made by those skilled in the art without departing from the scope of the application. The scope of the application is limited only by the appended claims.

Claims

1. A method for collaborative management based on digital twinning for magnetic levitation power equipment, characterized in that, The method is applied to a collaborative management system, the collaborative management system comprising a plurality of terminal devices, edge servers corresponding to the terminal devices, and a central cloud server, the terminal devices at least comprising magnetic levitation power equipment and end devices, the method comprising: The terminal device acquires corresponding environment data, device operation parameters, and control instructions, and determines whether the terminal device is a twin-to-be terminal device according to the control instructions; In the case that the terminal device is a twin-to-be terminal device, the terminal device uploads the environment data and the device operation parameters corresponding to the terminal device to the corresponding edge server; The edge server trains a digital twin model in the edge server based on the environment data and the device operation parameters corresponding to the terminal device, and sends local model parameters corresponding to the trained digital twin model to the central cloud server, the digital twin model being used for digital twinning of the twin-to-be terminal device to obtain a digital twin corresponding to the terminal device; The central cloud server aggregates the local model parameters sent by at least one edge server to obtain global model parameters corresponding to the digital twin model, and distributes the global model parameters to at least one edge server, so that the edge server updates the digital twin model based on the received global model parameters; The central cloud server stores the digital twins corresponding to the terminal devices, and the different digital twins are virtually connected in communication; If physical communication between different terminal devices fails, the digital twins corresponding to different terminal devices in the central cloud server are virtually communicated.

2. The collaborative management method of claim 1, wherein, The central cloud server comprises a plan management module, and the method further comprises: The plan management module in the central cloud server manages the terminal devices based on order data, the order data comprising customer relationship data, customer demand data, manufacturing plan data, or supply chain data.

3. The collaborative management method of claim 2, wherein, The collaborative management system further comprises a data acquisition and monitoring system, the terminal device acquires corresponding environment data, device operation parameters, and control instructions, comprising: The central cloud server determines the twin-to-be terminal devices according to the order data, and sends the control instructions to the data acquisition and monitoring system, the control instructions being used to indicate whether the terminal device is a twin-to-be terminal device; The data acquisition and monitoring system receives the control instructions, acquires the environment data and the device operation parameters corresponding to the terminal device, and sends the corresponding environment data, device operation parameters, and control instructions to the terminal device.

4. The collaborative management method of claim 3, wherein, The data acquisition and monitoring system comprises an acquisition and monitoring module, a data processing module, an interface management module, and an architecture module, the data acquisition and monitoring system receiving the control instructions, acquiring the environment data and the device operation parameters corresponding to the terminal device, comprising: The acquisition and monitoring module monitors the terminal device, and acquires the environment data and the device operation parameters corresponding to the terminal device; The data processing module receives the control instructions; The interface management module determines a target communication protocol; The architecture module interacts with other devices according to the target communication protocol.

5. The collaborative management method of claim 1, wherein, The edge server trains a digital twin model in the edge server based on the environment data and the device running parameters corresponding to the terminal device, including: The edge server obtains the digital twin model from the center cloud server at the beginning of the tth iteration training, t being an integer greater than 1; The edge server trains the digital twin model based on the environment data and the device running parameters corresponding to the terminal device using a gradient descent algorithm.

6. The collaborative management method of claim 1, wherein, The collaborative management system further comprises a base station, and the center cloud server aggregates the local model parameters sent by at least one edge server to obtain global model parameters corresponding to the digital twin model, including: The center cloud server aggregates the local model parameters sent by at least one edge server, which is expressed as: wherein, denotes the digital twin model after the aggregation, denotes the device information of the base station, denotes the environment data and the device running parameters of the terminal device corresponding to the i-th edge server, denotes the digital twin model of the i-th edge server after the t-th iteration training, denotes the number of edge servers.

7. The collaborative management method of claim 1, wherein, The method further comprises: The edge server sets the terminal device as a node in the blockchain.

8. The collaborative management method of claim 1, wherein, The magnetic levitation power equipment comprises a magnetic levitation device and a production device, and the end device comprises an Internet of Things device, a sensor and a camera.

9. A collaborative management system based on digital twinning for magnetic levitation powered equipment, characterized by, The collaborative management system comprises a plurality of terminal devices, edge servers corresponding to the terminal devices and a center cloud server, and the terminal devices at least comprise magnetic levitation power equipment and end devices: The terminal device is configured to acquire corresponding environment data, device running parameters and control instructions, and determine whether the terminal device is a terminal device to be twinned according to the control instructions; The terminal device is configured to upload the environment data and the device running parameters corresponding to the terminal device to the corresponding edge server if the terminal device is a terminal device to be twinned; The edge server is configured to train a digital twin model in the edge server based on the environment data and the device running parameters corresponding to the terminal device, and send local model parameters corresponding to the trained digital twin model to the center cloud server, the digital twin model being used to digitally twin the terminal device to be twinned to obtain a digital twin corresponding to the terminal device; The center cloud server is configured to aggregate the local model parameters sent by at least one edge server to obtain global model parameters corresponding to the digital twin model, and distribute the global model parameters to at least one edge server, so that the edge server updates the digital twin model based on the received global model parameters; The center cloud server stores the digital twins corresponding to the terminal devices, and the different digital twins are virtually connected in communication; If physical communication between different terminal devices fails, the digital twins corresponding to the different terminal devices in the center cloud server are virtually communicated.

Images