Fusion gateway-based workshop equipment management method and device, terminal and medium

Abstract

The invention discloses a workshop equipment management method and device based on a fusion gateway, a terminal and a medium, and the method comprises the steps: obtaining workshop equipment operation information, workshop equipment log information and environment information, and carrying out the preprocessing of the workshop equipment operation information and the workshop equipment log information, and obtaining joint representation information; through a preset anomaly recognition model, obtaining an anomaly recognition result according to the joint representation information; a workshop equipment health index is obtained through a preset multi-modal feature fusion model according to the joint representation information and the anomaly recognition result; and through a preset workshop equipment management optimization model, according to the workshop equipment health index and the environment information, obtaining a target optimization management strategy. The invention aims to accurately reduce the fault occurrence frequency of the workshop equipment, and further improve the operation efficiency of the workshop equipment.

Description

Technical Field

[0001] The present application relates to the technical field of workshop equipment management, and in particular, to a method, device, terminal and medium for workshop equipment management based on a fusion gateway. Background Art

[0002] Factory workshop equipment management refers to the process of monitoring, analyzing and optimizing the operating status, performance parameters, energy consumption and maintenance requirements of production equipment in the workshop to ensure the efficient and stable operation of the equipment and to meet the production goals to the greatest extent.

[0003] However, the existing means of workshop equipment management have the following problems: there are a wide variety of equipment in the workshop, and different equipment may adopt different communication protocols or management systems, making it difficult to achieve data interconnection between them. Data isolation leads to the inability to conduct comprehensive analysis and global optimization. At the same time, traditional methods mostly rely on empirical judgment or simple statistical analysis, making it difficult to deeply explore the potential value in the data. For complex equipment status assessment and fault prediction, traditional means often seem inadequate. Summary of the Invention

[0004] The main purpose of the present application is to provide a method, device, terminal and medium for workshop equipment management based on a fusion gateway, aiming to accurately reduce the occurrence frequency of workshop equipment failures and thereby improve the operating efficiency of workshop equipment.

[0005] To achieve the above object, the present application provides a method for workshop equipment management based on a fusion gateway, the method comprising: Obtaining workshop equipment operation information, workshop equipment log information and environmental information, and preprocessing the workshop equipment operation information and the workshop equipment log information to obtain joint characterization information, wherein the workshop equipment operation information is used to characterize the real-time operation status of workshop equipment, the workshop equipment log information is used to characterize the historical operation status of workshop equipment, and the joint characterization information is used to characterize the status of workshop equipment in terms of time series, events and images; Obtaining an anomaly recognition result according to the joint characterization information through a preset anomaly recognition model; Obtaining a workshop equipment health index according to the joint characterization information and the anomaly recognition result through a preset multimodal feature fusion model; Obtaining a target optimization management strategy according to the workshop equipment health index and the environmental information through a preset workshop equipment management optimization model.

[0006] Specifically, the preprocessing of the workshop equipment operation information and the workshop equipment log information to obtain joint characterization information includes: Through the timestamp synchronization algorithm, align the operation information of the workshop equipment and the workshop equipment log information to obtain the synchronized time series signal, the synchronized event text, and the synchronized thermal imaging image; Through the preset multi-modal embedding network model, obtain the joint representation information according to the synchronized time series signal, the synchronized event text, and the synchronized thermal imaging image.

[0007] Specifically, the joint representation information includes a time series feature vector and a thermal imaging image. The preset anomaly recognition model includes an anomaly warning signal model and a thermal state classification model. The anomaly recognition result includes the anomaly warning signal corresponding to the workshop equipment and the thermal state label corresponding to the workshop equipment; The obtaining of the anomaly recognition result according to the joint representation information through the preset anomaly recognition model includes: Through the anomaly warning signal model, obtain the anomaly warning signal according to the time series feature vector; Through the thermal state classification model, obtain the thermal state label according to the thermal imaging image.

[0008] Specifically, the anomaly warning signal model includes an encoder and a decoder. The encoder includes three LSTM layers, and the decoder includes three inverse LSTM layers; The obtaining of the anomaly warning signal according to the time series feature vector through the anomaly warning signal model includes: Through the encoder, compress the time series feature vector into a low-dimensional latent vector; Through the decoder, perform a reconstruction process on the low-dimensional latent vector to obtain a reconstructed time series feature vector; If the mean square error between the reconstructed time series feature vector and the time series feature vector is greater than or equal to a preset dynamic threshold, then obtain the anomaly warning signal based on the reconstructed time series feature vector.

[0009] Specifically, the thermal state classification model includes a lightweight ViT module and a classification head; The obtaining of the thermal state label according to the thermal imaging image through the thermal state classification model includes: Through the lightweight ViT module, segment the thermal imaging image into at least one image block, and based on the image block, form an image block sequence; Through the classification head, obtain the classification probability corresponding to the workshop equipment according to the image block sequence; Based on the classification probability and a preset classification probability interval, determine the thermal state label corresponding to the workshop equipment.

[0010] Specifically, the preset multi-modal feature fusion model includes a CNN branch module, a BiGRU branch module, and an attention fusion layer; Obtaining the workshop equipment health index through the preset multi-modal feature fusion model according to the joint representation information and the anomaly recognition result includes: Through the CNN branch module, obtaining a first feature vector according to the anomaly warning signal; Through the BiGRU branch module, obtaining a second feature vector according to the time series feature vector; Through the attention fusion layer, obtaining the workshop equipment health index according to the first feature vector and the second feature vector.

[0011] Specifically, the preset workshop equipment management optimization model includes a first fully connected layer, a second fully connected layer, a third fully connected layer, and a Gaussian policy layer. The ratio of the number of neurons between the first fully connected layer, the second fully connected layer, and the third fully connected layer is 1:2:4. The activation function of the Gaussian policy layer is the Tanh function. The target optimization management strategy includes the optimization control parameters of the workshop equipment; Obtaining the target optimization management strategy through the preset workshop equipment management optimization model according to the workshop equipment health index and the environmental information includes: Through the first fully connected layer, obtaining a first intermediate vector according to the workshop equipment health index and the environmental information; Through the second fully connected layer, obtaining a second intermediate vector according to the first intermediate vector; Through the third fully connected layer, obtaining a preliminary distribution of the control parameters according to the second intermediate vector; Through the Gaussian policy layer, obtaining the optimization control parameters according to the preliminary distribution of the control parameters.

[0012] To achieve the above object, the present application also provides a workshop equipment management device based on a fusion gateway. The device includes: A first unit, configured to obtain workshop equipment operation information, workshop equipment log information, and environmental information, and preprocess the workshop equipment operation information and the workshop equipment log information to obtain joint representation information, where the workshop equipment operation information is used to characterize the real-time operation status of the workshop equipment, the workshop equipment log information is used to characterize the historical operation status of the workshop equipment, and the joint representation information is used to characterize the status of the workshop equipment in terms of time series, events, and images; A second unit, configured to obtain an anomaly recognition result according to the joint representation information through a preset anomaly recognition model; A third unit, configured to obtain a workshop equipment health index according to the joint representation information and the anomaly recognition result through a preset multimodal feature fusion model; A fourth unit, configured to obtain a target optimization management strategy according to the workshop equipment health index and the environmental information through a preset workshop equipment management optimization model.

[0013] To achieve the above object, the present application further provides a terminal, including a memory storing multiple instructions; the processor loads the instructions from the memory to execute the steps in any one of the methods provided by the present application.

[0014] To achieve the above object, the present application further provides a medium, the medium stores multiple instructions, and the instructions are suitable for being loaded by a processor to execute the steps in any one of the methods provided by the present application.

[0015] A workshop equipment management method, device, terminal and medium based on a fusion gateway provided by the present application can first obtain workshop equipment operation information, workshop equipment log information and the environmental information, and preprocess the workshop equipment operation information and the workshop equipment log information to obtain joint representation information, where the workshop equipment operation information is used to represent the real-time operation status of workshop equipment, the workshop equipment log information is used to represent the historical operation status of workshop equipment, and the joint representation information is used to represent the status of workshop equipment in terms of time series, events and images; then, through a preset anomaly recognition model, an anomaly recognition result is obtained according to the joint representation information; then, through a preset multimodal feature fusion model, a workshop equipment health index is obtained according to the joint representation information and the anomaly recognition result; finally, through a preset workshop equipment management optimization model, a target optimization management strategy is obtained according to the workshop equipment health index and the environmental information, and by implementing the target optimization management strategy, the occurrence frequency of workshop equipment failures is finally accurately reduced, thereby improving the operation efficiency of workshop equipment. BRIEF DESCRIPTION OF THE DRAWINGS

[0016] Figure 1 It is a schematic flowchart of the method provided by the embodiment of the present application; Figure 2 It is a schematic structural diagram of the device provided by the embodiment of the present application; Figure 3 It is a schematic structural diagram of the terminal provided by the embodiment of the present application. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0017] Next, the technical solutions in the embodiments of the present application will be clearly and completely described in conjunction with the accompanying drawings in the embodiments of the present application. Obviously, the described embodiments are only a part of the embodiments of the present application, rather than all the embodiments. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without creative efforts belong to the scope of protection of the present application.

[0018] Since the existing workshop equipment management means have the following problems: there are a wide variety of equipment in the workshop, different equipment may adopt different communication protocols or management systems, and it is difficult to achieve data intercommunication between them. Data isolation leads to the inability to conduct comprehensive analysis and global optimization; at the same time, traditional methods mostly rely on empirical judgment or simple statistical analysis, and it is difficult to deeply explore the potential value in the data. For complex equipment status assessment and fault prediction, traditional means often seem inadequate.

[0019] Therefore, the embodiments of the present application provide a workshop equipment management method, device, terminal and medium based on a fusion gateway to solve practical technical problems.

[0020] In some embodiments, the device can be specifically integrated in an electronic device, and the electronic device can be devices such as a terminal, a server, etc.

[0021] In some embodiments, the server can also be implemented in the form of a terminal.

[0022] Among them, the server can be an independent physical server, or a server cluster or distributed system composed of multiple physical servers, or a cloud server that provides basic cloud computing services such as cloud services, cloud information libraries, cloud computing, cloud functions, cloud storage, network services, cloud communications, middleware services, domain name services, security services, CDN (Content Delivery Network), and large information and artificial intelligence platforms.

[0023] Among them, the terminal can be a smart phone, a tablet computer, a notebook computer, a desktop computer, a smart speaker, a smart watch, etc., but is not limited thereto. The terminal and the server can be directly or indirectly connected through wired or wireless communication methods, and the present application does not make any restrictions here.

[0024] The following will be described in detail respectively. It should be noted that the serial numbers of the following embodiments do not limit the preferred order of the embodiments.

[0025] An embodiment of the present application provides a workshop equipment management method based on a fusion gateway, which is applied to a workshop equipment management system. The workshop equipment management system includes a fusion gateway and a server terminal. The fusion gateway is used to receive workshop equipment operation information, workshop equipment log information, and environmental information of the workshop where the workshop equipment is located, and is used to upload the workshop equipment operation information, workshop equipment log information, and environmental information to the server terminal.

[0026] Specifically, a fusion gateway is a device or system that plays a key role in the enterprise and government network environment and has an important position in the above-mentioned workshop equipment management method based on the fusion gateway. The factory workshop equipment management of the fusion gateway is a modern equipment management mode. By using the fusion gateway as the core data collection and processing node, it realizes real-time monitoring, data analysis, fault prediction, and optimization scheduling of the operation status of various devices in the workshop.

[0027] The fusion gateway can support multiple communication protocols (such as Modbus, OPC UA, MQTT, etc.) and collect multi-source heterogeneous data from different types of devices. After the data is preprocessed, it is transmitted to the server terminal through a security protocol for further analysis.

[0028] The fusion gateway can integrate a PON optical module (such as GPON ONU) and directly connect to the operator's OLT through optical fiber to achieve high-speed fiber access. PON (Passive Optical Network), an optical fiber-based access technology, realizes point-to-multipoint communication through an optical splitter without active devices. Enterprise-level functions are superimposed on the basis of PON access, such as: Security: Support for intrusion prevention (IPS), URL filtering, virus protection; Intelligent routing: Load balancing based on applications / links; Branch management: Zero-configuration deployment, centralized monitoring.

[0029] The fusion gateway can also be a multi-functional network device for enterprise / government users, providing functions such as Internet access, security protection, and traffic management for enterprise-government gateways. Its core functions include: NAT conversion, firewall, VPN (IPsec / SSL), QoS, multi-WAN port load balancing, and branch interconnection. Its application scenarios include: enterprise headquarters / branch network exits, government affairs private network access, and cloud service connections.

[0030] The workshop equipment management system can utilize big data analysis and artificial intelligence technologies (such as deep learning, reinforcement learning, etc.) to evaluate the operation status of equipment, predict potential faults, and optimize equipment scheduling strategies. The management system displays the analysis results through a visual interface and allows managers to send control instructions to workshop equipment through the fusion gateway to form a closed-loop feedback mechanism. It supports adaptive adjustment in a dynamic environment, meets the requirements of complex production scenarios, and improves the utilization rate and production efficiency of workshop equipment.

[0031] As Figure 1 , the specific process of the method can be as follows: S110. Obtain the workshop equipment operation information, workshop equipment log information, and the environmental information, and preprocess the workshop equipment operation information and the workshop equipment log information to obtain joint characterization information, where the workshop equipment operation information is used to characterize the real-time operation status of workshop equipment, the workshop equipment log information is used to characterize the historical operation status of workshop equipment, and the joint characterization information is used to characterize the status of workshop equipment in terms of time series, events, and images.

[0032] In some embodiments, the fusion gateway can parse different device protocols through a dynamic protocol adaptation engine (supporting Modbus, OPC UA, etc.), extract the workshop equipment operation information and workshop equipment log information of workshop equipment, and obtain standardized data.

[0033] In some embodiments, the preprocessing of the workshop equipment operation information and the workshop equipment log information to obtain joint characterization information includes the step contents from A1 to A2 as shown below: A1. Through a timestamp synchronization algorithm, perform alignment processing on the workshop equipment operation information and the workshop equipment log information to obtain synchronized time series signals, synchronized event texts, and synchronized thermal imaging images.

[0034] A2. Through a preset multi-modal embedding network model, obtain the joint characterization information according to the synchronized time series signals, synchronized event texts, and synchronized thermal imaging images.

[0035] Specifically, the workshop equipment operation information and the workshop equipment log information can be input into a multi-modal embedding network (MMEN), and a joint characterization vector (with a configurable dimension, such as 256 dimensions) for characterizing time series - event - image in a unified feature space is output, that is, the joint characterization information.

[0036] S120. Through a preset anomaly recognition model, obtain an anomaly recognition result according to the joint characterization information.

[0037] In some embodiments, the joint characterization information includes a time series feature vector and a thermal imaging image, the preset anomaly recognition model includes an anomaly warning signal model and a thermal state classification model, and the anomaly recognition result includes an anomaly warning signal corresponding to the workshop equipment and a thermal state label corresponding to the workshop equipment.

[0038] Specifically, the obtaining of the anomaly recognition result according to the joint characterization information through the preset anomaly recognition model includes the step contents from B1 to B2 as shown below: B1. Using the abnormal warning signal model, based on the timing feature vector, obtain the abnormal warning signal.

[0039] Continuing the description based on the above embodiment, the abnormal warning signal model includes an encoder and a decoder. The encoder includes three LSTM layers, and the decoder includes three reverse LSTM layers.

[0040] The step of using the abnormal warning signal model to obtain the abnormal warning signal based on the timing feature vector includes the following steps B11 to B13: B11. Using the encoder, compress the timing feature vector into a low-dimensional latent vector.

[0041] B12. Using the decoder, perform a reconstruction process on the low-dimensional latent vector to obtain a reconstructed timing feature vector.

[0042] B13. If the mean square error between the reconstructed timing feature vector and the timing feature vector is greater than or equal to a preset dynamic threshold, then based on the reconstructed timing feature vector, obtain the abnormal warning signal.

[0043] Continuing the description based on the above embodiment, the timing feature vector can be vibration spectrum data, which contains timing signal features.

[0044] The LSTM layer (Long Short-Term Memory) is a special RNN (Recurrent Neural Network) structure, specifically designed to solve the problem of gradient vanishing or gradient explosion that traditional RNNs are prone to when processing long sequence data. LSTM controls the flow of information by introducing a "gating mechanism", thereby being able to effectively capture long-term dependencies in time series.

[0045] The reverse LSTM layer is a variant of the long short-term memory network (LSTM). It is mainly used for reverse modeling of time series data. Different from the standard forward LSTM layer, the reverse LSTM layer calculates in the reverse order of the time series (from back to front), thereby capturing reverse dependencies in the input sequence.

[0046] In some embodiments, the reverse LSTM layer is used in combination with the LSTM layer to form a bidirectional LSTM (Bidirectional LSTM). The bidirectional LSTM utilizes both forward and backward information, thereby understanding time series data more comprehensively.

[0047] B2. Using the thermal state classification model, obtain the thermal state label according to the thermal imaging image.

[0048] Continuing the description based on the above embodiments, the thermal state classification model includes a lightweight ViT module and a classification head.

[0049] Specifically, using the thermal state classification model to obtain the thermal state label according to the thermal imaging image includes the step contents from B21 to B23 as follows: B21. Using the lightweight ViT module, segment the thermal imaging image into at least one image patch, and based on the image patches, form an image patch sequence.

[0050] B22. Using the classification head, obtain the classification probability corresponding to the workshop equipment according to the image patch sequence.

[0051] B23. Based on the classification probability and a preset classification probability interval, determine the thermal state label corresponding to the workshop equipment.

[0052] In some embodiments, the thermal state classification model can be a MobileViT model.

[0053] The lightweight ViT module (Vision Transformer) is an improved version of the Vision Transformer architecture, aiming to reduce the computational complexity and the number of parameters to make it more suitable for resource-constrained scenarios (such as mobile devices, edge computing, etc.), while maintaining high performance.

[0054] The classification head is a key module in a deep learning model for completing classification tasks, located at the end of the neural network. The classification head maps the features extracted by the network to specific class labels, thereby realizing the classification of the input data. Specifically, the classification head can be a fully connected layer for obtaining the classification probability corresponding to the workshop equipment. For example, thermal state labels: normal (probability > 0.8), overheat (0.5 ≤ probability ≤ 0.8), local high temperature (probability < 0.5).

[0055] S130. Using a preset multi-modal feature fusion model, obtain the health index of the workshop equipment according to the joint representation information and the anomaly recognition result.

[0056] Continuing the description based on the above embodiments, the preset multi-modal feature fusion model includes a CNN branch module, a BiGRU branch module, and an attention fusion layer.

[0057] Specifically, by means of the preset multi-modal feature fusion model, based on the joint representation information and the anomaly recognition result, the workshop equipment health index is obtained, including the step contents from S131 to S133 as shown below: S131. Through the CNN branch module, according to the anomaly warning signal, a first feature vector is obtained.

[0058] S132. Through the BiGRU branch module, according to the time series feature vector, a second feature vector is obtained.

[0059] S133. Through the attention fusion layer, according to the first feature vector and the second feature vector, the workshop equipment health index is obtained.

[0060] In some embodiments, the CNN branch module maps the anomaly warning signal (such as whether the equipment has vibration / temperature anomalies) to a low-dimensional feature vector to capture spatial patterns (such as specific abnormal waveforms in the spectrum). The anomaly warning signal is the input binary anomaly signal (for example, [1, 0, 1] indicates that the 1st and 3rd devices are abnormal). The CNN branch module may include: an embedding layer: converting the anomaly signal into a continuous vector (such as [0.1, 0.3, 0.5]); a convolutional layer: extracting local features through 1D convolution (such as high-frequency anomaly patterns in the spectrum).

[0061] The output of the CNN branch module, that is, the first feature vector (such as [0.2, 0.4, 0.6]), reflects the distribution pattern of the anomaly signal.

[0062] Continuing the above embodiments, the BiGRU branch module processes the time series feature vector (such as vibration spectrum, temperature curve) to extract time-dependent features (such as the equipment degradation trend).

[0063] The input of the BiGRU branch module is the time series feature vector (such as a vibration FFT with a length of 512) The BiGRU branch module may include: a bidirectional GRU: capturing both forward and backward time series dependencies simultaneously (such as abnormal signs before a fault and the recovery trend after a fault); a fully connected layer: compressing high-dimensional features into a low-dimensional representation (such as reducing from 256 dimensions to 64 dimensions).

[0064] The output of the BiGRU branch module is the second feature vector (such as [0.05, 0.18,..., 0.92]), reflecting the dynamic changes in the equipment operation state.

[0065] Continuing the above embodiments, the attention fusion layer dynamically weights and fuses the feature vectors of the CNN and BiGRU to generate the equipment health index (HI).

[0066] The inputs of the attention fusion layer include: the first feature vector (the abnormal pattern features output by the CNN), and the second feature vector (the time series features output by the BiGRU).

[0067] The attention fusion layer may include: concatenated features: combining the features of the two branches (e.g., [0.2, 0.4] + [0.05, 0.18] = [0.25, 0.58]); attention mechanism: calculating the weights of the features of the two parts (e.g., the abnormal signal accounts for 30% and the time series features account for 70%); weighted fusion: generating comprehensive features and mapping them to a health index (HI ∈ [0, 1]).

[0068] The output of the attention fusion layer is the health index of the workshop equipment (e.g., HI = 0.76 indicates that the equipment is in good condition).

[0069] S140. Through a preset workshop equipment management optimization model, obtain a target optimization management strategy according to the health index of the workshop equipment and the environmental information Continuing the above embodiment, the preset workshop equipment management optimization model includes a first fully connected layer, a second fully connected layer, a third fully connected layer, and a Gaussian policy layer. The ratio of the number of neurons between the first fully connected layer, the second fully connected layer, and the third fully connected layer is 1:2:4. The activation function of the Gaussian policy layer is the Tanh function. The target optimization management strategy includes the optimization control parameters of the workshop equipment.

[0070] Specifically, the step of obtaining a target optimization management strategy through the preset workshop equipment management optimization model according to the health index of the workshop equipment and the environmental information includes the following steps S141 to S144:[[]] S141. Through the first fully connected layer, obtain a first intermediate vector according to the health index of the workshop equipment and the environmental information.

[0071] S142. Through the second fully connected layer, obtain a second intermediate vector according to the first intermediate vector.

[0072] S143. Through the third fully connected layer, obtain a preliminary distribution of the control parameters according to the second intermediate vector.

[0073] S144. Through the Gaussian policy layer, obtain the optimization control parameters according to the preliminary distribution of the control parameters.

[0074] Continuing the above embodiment, the inputs of the first fully connected layer include: the health index of the workshop equipment (HI, range 0 - 1), and environmental parameters (temperature, humidity, order priority, etc., input after standardization) The structure of the first fully connected layer may include: Number of neurons: Assuming the input dimension is D, the output dimension is 2D (designed with a ratio of 1:2). Activation function: ReLU.

[0075] The first fully connected layer is used to extract the nonlinear feature interaction between the health index and environmental parameters (such as the coupling effect between HI and ambient temperature).

[0076] Continuing with the above embodiment, the input of the second fully connected layer is the intermediate vector (e.g., 256-dimensional) output by the first layer. The structure of the second fully connected layer may include: Number of neurons: The number of neurons in the second layer is twice that of the first layer (e.g., 512 neurons). Activation function: ReLU.

[0077] The second fully connected layer is used to further mine high-order features (such as the correlation between the changing trend of HI over time and energy consumption).

[0078] Continuing with the above embodiment, the input of the third fully connected layer may be an intermediate vector (eg, 512-dimensional) output by the second layer.

[0079] The structure of the third fully connected layer may include: Number of neurons: the number of neurons in the third layer is twice that of the second layer (eg, 1024 neurons). Activation function: ReLU.

[0080] The third fully connected layer is used to integrate global features and generate a preliminary distribution of control parameters (such as mean μ and variance σ).

[0081] Continuing with the above-mentioned embodiment, the Gaussian strategy layer maps the preliminary distribution into Gaussian distribution parameters (μ, σ), and outputs a range through Tanh constraints (such as speed control within ±10%), and finally obtains the optimized control parameters.

[0082] For example, the optimization control parameters may include: speed optimization parameters: μ=95% (reference speed), σ=5% (±5% fluctuation allowed). Cooling water flow: μ=100% (reference value), σ=10% (dynamic adjustment). The target optimization management strategy is determined in this way.

[0083] In summary, the present application provides a workshop equipment management method based on a fusion gateway. Through a workshop equipment management system based on a fusion gateway, the frequency of workshop equipment failures can be accurately reduced, thereby improving the operating efficiency of the workshop equipment.

[0084] To better implement the above method, an embodiment of the present application further provides a workshop equipment management device based on a fusion gateway. This device can be specifically integrated into an electronic device, which can be a device such as a terminal or a server. Among them, the terminal can be a device such as a mobile phone, a tablet computer, a smart Bluetooth device, a laptop computer, or a personal computer; the server can be a single server or a server cluster composed of multiple servers.

[0085] For example, in this embodiment, taking the workshop equipment management device based on the fusion gateway being specifically integrated into the terminal as an example, the method of the embodiment of the present application will be described in detail.

[0086] For instance, as Figure 2 shown, the workshop equipment management device 200 based on the fusion gateway may include a first unit 201, a second unit 202, a third unit 203, and a fourth unit 204. The device includes: The first unit 201 is configured to obtain workshop equipment operation information, workshop equipment log information, and the environmental information, and preprocess the workshop equipment operation information and the workshop equipment log information to obtain combined characterization information. Among them, the workshop equipment operation information is used to characterize the real-time operation status of the workshop equipment, the workshop equipment log information is used to characterize the historical operation status of the workshop equipment, and the combined characterization information is used to characterize the status of the workshop equipment in terms of time series, events, and images; The second unit 202 is configured to obtain an anomaly recognition result according to the combined characterization information through a preset anomaly recognition model; The third unit 203 is configured to obtain a workshop equipment health index according to the combined characterization information and the anomaly recognition result through a preset multimodal feature fusion model; The fourth unit 204 is configured to obtain a target optimization management strategy according to the workshop equipment health index and the environmental information through a preset workshop equipment management optimization model.

[0087] In specific implementation, the above-mentioned each unit can be implemented as an independent entity, or can be combined arbitrarily to be implemented as the same or several entities. For the specific implementation of the above-mentioned each unit, reference can be made to the method embodiment above, which will not be elaborated here.

[0088] As can be seen from the above, the embodiment of the present application can accurately reduce the occurrence frequency of workshop equipment failures, and thus improve the operation efficiency of workshop equipment.

[0089] Embodiments of the present application further provide an electronic device, which may be a device such as a terminal or a server. Among them, the terminal may be a mobile phone, a tablet computer, a smart Bluetooth device, a laptop computer, a personal computer, etc.; the server may be a single server or a server cluster composed of multiple servers, etc.

[0090] In some embodiments, the product processing device may also be integrated in multiple electronic devices. For example, the product processing device may be integrated in multiple servers, and multiple servers are used to implement the workshop equipment management method based on the fusion gateway of the present application.

[0091] In this embodiment, the electronic device of this embodiment will be taken as an example of a terminal for detailed description. For example, as Figure 3 shown, it shows a schematic structural diagram of the terminal 300 involved in the embodiments of the present application. Specifically: The terminal 300 may include components such as a processor 301 with one or more processing cores, a memory 302 with one or more media, a power supply 303, an input module 304, and a communication module 305. Those skilled in the art can understand that Figure 3 the structure of the terminal 300 shown in does not constitute a limitation on the terminal 300, and it may include more or fewer components than shown, or combine certain components, or have different component arrangements. Among them: The processor 301 is the control center of the terminal 300, connecting various parts of the entire terminal 300 through various interfaces and lines. By running or executing software programs and / or modules stored in the memory 302, and by calling information stored in the memory 302, it executes various functions of the terminal 300 and processes information, thereby monitoring the terminal 300 as a whole. In some embodiments, the processor 301 may include one or more processing cores; in some embodiments, the processor 301 may integrate an application processor and a modem processor. Among them, the application processor mainly processes the operating system, user interface, and application programs, etc., and the modem processor mainly processes wireless communication. It can be understood that the above modem processor may not be integrated into the processor 301 either.

[0092] The memory 302 can be used to store software programs and modules. The processor 301 executes various functional applications and information processing by running the software programs and modules stored in the memory 302. The memory 302 mainly includes a program storage area and an information storage area. Among them, the program storage area can store an operating system, application programs required for at least one function (such as a sound playback function, an image playback function, etc.); the information storage area can store information created according to the use of the terminal 300. In addition, the memory 302 can include high-speed random access memory and can also include non-volatile memory, such as at least one magnetic disk storage device, a flash memory device, or other volatile solid-state storage devices. Correspondingly, the memory 302 can also include a memory controller to provide the processor 301 with access to the memory 302.

[0093] The terminal 300 further includes a power supply 303 for powering each component. In some embodiments, the power supply 303 can be logically connected to the processor 301 through a power management system, so as to implement functions such as management of charging, discharging, and power consumption management through the power management system. The power supply 303 can also include any components such as one or more DC or AC power supplies, a recharge system, a power failure detection circuit, a power converter or inverter, and a power status indicator.

[0094] The terminal 300 may further include an input module 304, which can be used to receive input digital or character information, and generate keyboard, mouse, joystick, optical or trackball signal inputs related to user settings and function controls.

[0095] The terminal 300 may further include a communication module 305. In some embodiments, the communication module 305 can include a wireless module. The terminal 300 can perform short-distance wireless transmission through the wireless module of the communication module 305, thereby providing users with wireless broadband Internet access. For example, the communication module 305 can be used to help users send and receive emails, browse web pages, and access streaming media, etc.

[0096] Although not shown, the terminal 300 may further include a display unit, etc., which will not be elaborated here. Specifically, in this embodiment, the processor 301 in the terminal 300 will load the executable files corresponding to the processes of one or more application programs into the memory 302 according to the following instructions, and the processor 301 will run the application programs stored in the memory 302 to implement various functions as follows: Obtain the operation information of workshop equipment, the log information of workshop equipment, and the environmental information, and preprocess the operation information of workshop equipment and the log information of workshop equipment to obtain combined characterization information, where the operation information of workshop equipment is used to characterize the real-time operation status of workshop equipment, the log information of workshop equipment is used to characterize the historical operation status of workshop equipment, and the combined characterization information is used to characterize the status of workshop equipment in terms of time series, events, and images; Obtain an anomaly recognition result according to the combined characterization information through a preset anomaly recognition model; Obtain the health index of workshop equipment according to the combined characterization information and the anomaly recognition result through a preset multi-modal feature fusion model; Obtain a target optimization management strategy according to the health index of workshop equipment and the environmental information through a preset workshop equipment management optimization model.

[0097] As can be seen from the above, the embodiments of the present application can accurately reduce the occurrence frequency of workshop equipment failures, and thus improve the operation efficiency of workshop equipment.

[0098] Those of ordinary skill in the art can understand that all or part of the steps in the above methods of the embodiments can be completed by instructions, or by controlling related hardware through instructions. The instructions can be stored in a medium and loaded and executed by a processor.

[0099] Therefore, the embodiments of the present application provide a medium in which multiple instructions are stored. The instructions can be loaded by a processor to execute the steps in any of the workshop equipment management methods based on a fusion gateway provided by the embodiments of the present application. For example, the instructions can execute the following steps: Obtain the operation information of workshop equipment, the log information of workshop equipment, and the environmental information, and preprocess the operation information of workshop equipment and the log information of workshop equipment to obtain combined characterization information, where the operation information of workshop equipment is used to characterize the real-time operation status of workshop equipment, the log information of workshop equipment is used to characterize the historical operation status of workshop equipment, and the combined characterization information is used to characterize the status of workshop equipment in terms of time series, events, and images; Obtain an anomaly recognition result according to the combined characterization information through a preset anomaly recognition model; Obtain the health index of workshop equipment according to the combined characterization information and the anomaly recognition result through a preset multi-modal feature fusion model; Obtain a target optimization management strategy according to the health index of workshop equipment and the environmental information through a preset workshop equipment management optimization model.

[0100] Among them, the medium may include: read-only memory (ROM, Read Only Memory), random access memory (RAM, Random Access Memory), magnetic disk or optical disc, etc.

[0101] According to one aspect of the present application, there is provided a computer program product or a computer program. The computer program product or the computer program includes computer instructions, and the computer instructions are stored in a medium. The processor of the computer device reads the computer instructions from the medium, and the processor executes the computer instructions, so that the computer device executes the methods provided in the various alternative implementations in the above embodiments.

[0102] Since the instructions stored in the medium can execute the steps in any of the workshop equipment management methods based on the fusion gateway provided in the embodiments of the present application, therefore, the beneficial effects that can be achieved by any of the workshop equipment management methods based on the fusion gateway provided in the embodiments of the present application can be realized. For details, see the previous embodiments and will not be elaborated here.

[0103] The above has introduced in detail a workshop equipment management method, device, terminal and medium based on a fusion gateway provided in the embodiments of the present application. Specific examples are used in this article to elaborate on the principle and implementation manner of the present application. The description of the above embodiments is only used to help understand the method and its core idea of the present application; at the same time, for those skilled in the art, according to the idea of the present application, there will be changes in the specific implementation manner and application scope. In summary, the content of this specification should not be construed as a limitation to the present application.

Claims

1. A workshop equipment management method based on a fusion gateway, characterized in that: The method comprises: Acquire workshop equipment operation information, workshop equipment log information and environmental information, and pre-process the workshop equipment operation information and workshop equipment log information to obtain joint representation information, wherein the workshop equipment operation information is used to represent the real-time operation status of the workshop equipment, the workshop equipment log information is used to represent the historical operation status of the workshop equipment, and the joint representation information is used to represent the status of the workshop equipment at the time sequence, event and image levels; By presetting an anomaly recognition model, an anomaly recognition result is obtained according to the joint representation information; By presetting a multimodal feature fusion model, a workshop equipment health index is obtained according to the joint characterization information and the abnormality recognition result; By presetting the workshop equipment management optimization model, a target optimization management strategy is obtained according to the workshop equipment health index and the environmental information.

2. The method according to claim 1, characterized in that The preprocessing of the workshop equipment operation information and the workshop equipment log information to obtain joint representation information includes: The workshop equipment operation information and the workshop equipment log information are aligned by a timestamp synchronization algorithm to obtain a synchronized timing signal, a synchronized event text, and a synchronized thermal imaging image; By presetting a multimodal embedding network model, the joint representation information is obtained according to the synchronized time series signal, the synchronized event text and the synchronized thermal imaging image.

3. The method according to claim 1, characterized in that The joint characterization information includes a time series feature vector and a thermal imaging image, the preset abnormality recognition model includes an abnormality warning signal model and a thermal state classification model, and the abnormality recognition result includes an abnormality warning signal corresponding to the workshop equipment and a thermal state label corresponding to the workshop equipment; The abnormality recognition result is obtained by presetting the abnormality recognition model according to the joint representation information, including: Obtaining the abnormal warning signal according to the time series feature vector through the abnormal warning signal model; The thermal state label is obtained according to the thermal imaging image through the thermal state classification model.

4. The method according to claim 3, characterized in that The abnormal warning signal model includes an encoder and a decoder, the encoder includes three LSTM layers, and the decoder includes three anti-LSTM layers; The step of obtaining the abnormal warning signal according to the time series feature vector through the abnormal warning signal model includes: Compressing the temporal feature vector into a low-dimensional latent vector by the encoder; Reconstructing the low-dimensional latent vector through the decoder to obtain a reconstructed time series feature vector; If the mean square error between the reconstructed time series feature vector and the time series feature vector is greater than or equal to a preset dynamic threshold, the abnormal warning signal is obtained based on the reconstructed time series feature vector.

5. The method according to claim 3, characterized in that The thermal state classification model includes a lightweight ViT module and a classification head; The step of obtaining the thermal state label according to the thermal imaging image by using the thermal state classification model includes: Segmenting the thermal imaging image into at least one image block through the lightweight ViT module, and composing an image block sequence based on the image block; Obtaining, by means of the classification head, the classification probability corresponding to the workshop equipment according to the image block sequence; Based on the classification probability and the preset classification probability interval, a thermal state label corresponding to the workshop equipment is determined.

6. The method according to claim 3, characterized in that The preset multimodal feature fusion model includes a CNN branch module, a BiGRU branch module and an attention fusion layer; The preset multimodal feature fusion model is used to obtain the workshop equipment health index according to the joint characterization information and the abnormality recognition result, including: Obtaining a first feature vector according to the abnormal warning signal through the CNN branch module; Obtaining a second feature vector according to the time series feature vector through the BiGRU branch module; The workshop equipment health index is obtained through the attention fusion layer according to the first feature vector and the second feature vector.

7. The method according to claim 1, characterized in that The preset workshop equipment management optimization model includes a first fully connected layer, a second fully connected layer, a third fully connected layer and a Gaussian strategy layer, the ratio of the number of neurons between the first fully connected layer, the second fully connected layer and the third fully connected layer is 1:2:4, the activation function of the Gaussian strategy layer is a Tanh function, and the target optimization management strategy includes the optimization control parameters of the workshop equipment; The target optimization management strategy is obtained by presetting the workshop equipment management optimization model according to the workshop equipment health index and the environmental information, including: Obtaining a first intermediate vector according to the workshop equipment health index and the environmental information through the first fully connected layer; Obtaining a second intermediate vector according to the first intermediate vector through the second fully connected layer; Obtaining, through the third fully connected layer, a preliminary distribution of control parameters according to the second intermediate vector; The optimized control parameters are obtained according to the preliminary distribution of the control parameters through the Gaussian strategy layer.

8. A workshop equipment management device based on a fusion gateway, characterized in that: The device comprises: The first unit is used to obtain workshop equipment operation information, workshop equipment log information and environmental information, and pre-process the workshop equipment operation information and workshop equipment log information to obtain joint representation information, wherein the workshop equipment operation information is used to represent the real-time operation status of the workshop equipment, the workshop equipment log information is used to represent the historical operation status of the workshop equipment, and the joint representation information is used to represent the status of the workshop equipment at the time sequence, event and image levels; The second unit is used to obtain an anomaly recognition result according to the joint representation information by using a preset anomaly recognition model; The third unit is used to obtain a workshop equipment health index according to the joint characterization information and the abnormality recognition result by using a preset multimodal feature fusion model; The fourth unit is used to obtain a target optimization management strategy according to the workshop equipment health index and the environmental information by using a preset workshop equipment management optimization model.

9. A terminal, characterized in that: It comprises a processor and a memory, wherein the memory stores a plurality of instructions; the processor loads instructions from the memory to execute the steps in the method as claimed in any one of claims 1 to 7.

10. A medium, characterized in that The medium stores a plurality of instructions, and the instructions are suitable for being loaded by a processor to execute the steps in the method according to any one of claims 1 to 7.

Images