Magnetic suspension type safe electric device and working state remote monitoring method thereof

Abstract

The invention relates to the technical field of safe electricity, and particularly discloses a magnetic suspension type safe electricity device and a working state remote monitoring method thereof. Real-time temperature data of the magnetic suspension type partition mechanism and the external environment are synchronously collected through a first temperature sensor arranged on the magnetic suspension type partition mechanism and a second temperature sensor arranged in the external environment, and a working state remote monitoring server is arranged at the rear end. The real-time temperature data of the magnetic suspension type partition mechanism and the external environment are subjected to time sequence analysis by using a deep learning algorithm, time sequence change modes of the mechanism temperature and the environment temperature are mined respectively, and the two time sequence change modes are subjected to characteristic comparison analysis to filter out the interference influence of the external environment temperature on the state temperature of the mechanism; therefore, more accurate abnormal state detection of the magnetic suspension type partition mechanism is realized. Therefore, the state monitoring requirements of the magnetic suspension type partition mechanism under different environmental conditions can be fully met, the accuracy of a monitoring result is improved, and the safety and reliability of power equipment are further improved.

Description

Technical Field

[0001] The present application relates to the field of magnetic levitation safety electrical technology, and more specifically, to a magnetic levitation safety electrical device and a remote monitoring method for its working status. Background Art

[0002] With the development of science and technology and the improvement of people's quality of life, the safety and reliability of power equipment are receiving more and more attention. Traditional power equipment, especially those involving high voltage and high current, may cause safety hazards due to overheating and other problems during operation, which not only threatens the service life of the equipment itself, but also may cause unpredictable loss of life and property to users. Existing power equipment safety monitoring technology mainly relies on regular manual inspections. However, due to the long manual inspection cycle, high cost, and the great influence of human factors, it is usually difficult to achieve real-time monitoring, and it is easy to have false alarms or missed alarms.

[0003] In this regard, the invention patent with publication number CN118300280A discloses a magnetic levitation safety electrical device and a state detection method, which includes a phase line interface terminal, a ground line interface terminal, and a magnetic levitation partition mechanism arranged between the phase line interface terminal and the ground line interface terminal. The temperature data of the magnetic levitation partition mechanism is monitored and collected in real time through a temperature sensor, and an artificial intelligence-based data processing and analysis algorithm is introduced to perform time series analysis of the temperature time series data, so as to capture the time series pattern and change trend of the temperature value in the time dimension, thereby realizing real-time monitoring of the working status of the magnetic levitation partition mechanism and abnormal early warning prompts, reducing the subjective errors and abnormal state response lag problems caused by traditional manual regular inspections or regular maintenance.

[0004] However, although the above scheme has improved the effect of power equipment safety monitoring to a certain extent, it still faces some challenges in practical application. For example, the temperature change of the external environment will affect the temperature inside the equipment, thereby interfering with the accuracy of the monitoring results. In addition, the above scheme deploys the temperature data processing task of the magnetic levitation partition mechanism on the edge side, which makes the real-time data processing and analysis limited by the computing power of the edge side equipment.

[0005] Therefore, in order to further improve the accuracy and real-time performance of the working status detection of the magnetic levitation safety electrical device, an optimized magnetic levitation safety electrical device and a remote monitoring method for its working status are expected. Summary of the invention

[0006] In order to solve the above technical problems, the present application is proposed. The embodiment of the present application provides a magnetic levitation safety electrical device and a remote monitoring method for its working status, which synchronously collects the real-time temperature data of the magnetic levitation partition mechanism and the external environment through a first temperature sensor deployed in the magnetic levitation partition mechanism and a second temperature sensor deployed in the external environment, and deploys a working status remote monitoring server at the back end, and uses a deep learning algorithm to perform time series analysis on the real-time temperature data of the magnetic levitation partition mechanism and the external environment, respectively mining the time series change patterns of the mechanism temperature and the ambient temperature, and performing feature comparison analysis on the two to filter out the interference effect of the external environment temperature on the mechanism state temperature, thereby achieving more accurate abnormal state detection of the magnetic levitation partition mechanism. In this way, it can fully adapt to the state monitoring needs of the magnetic levitation partition mechanism under different environmental conditions, improve the accuracy of the monitoring results, and further improve the safety and reliability of power equipment.

[0007] According to one aspect of the present application, a magnetic levitation safety electrical device is provided, which includes: a phase line interface terminal for connecting a phase line; a ground line interface terminal for connecting a ground line; a magnetic levitation partition mechanism arranged between the phase line interface terminal and the ground line interface terminal, wherein the magnetic levitation partition mechanism includes a primary coil and a secondary coil opposite to the primary coil, the primary coil is electrically connected to the phase line interface terminal, and the secondary coil is electrically connected to the ground line interface terminal; a first temperature sensor deployed in the magnetic levitation partition mechanism, which is used to collect a time queue of the real-time temperature of the magnetic levitation partition mechanism, wherein the first temperature sensor is connected to an external working status remote monitoring server through a wireless communication module.

[0008] In the above-mentioned magnetic levitation safety electrical device, the working status remote monitoring server includes: a temperature data receiving module, which is used to receive the time queue of the real-time temperature of the magnetic levitation partition mechanism collected by the first temperature sensor, and the time queue of the external environment temperature collected by the second temperature sensor; a mechanism temperature fluctuation feature extraction module, which is used to perform empirical mode decomposition and temperature fluctuation feature extraction on the time queue of the real-time temperature of the magnetic levitation partition mechanism to obtain a series of mechanism state temperature intrinsic mode function component waveform feature vectors; a mechanism temperature timing information aggregation module, which is used to perform significant aggregation based on characteristic energy distribution on the series of mechanism state temperature intrinsic mode function component waveform feature vectors to obtain a mechanism state temperature timing significant aggregation representation vector; a feature filtering module, which is used to filter out environmental temperature interference on the mechanism state temperature timing significant aggregation representation vector based on the temperature timing characteristics in the time queue of the external environment temperature to obtain a filtered mechanism state temperature timing significant aggregation representation vector; a working status detection module, which is used to determine whether the state of the magnetic levitation partition mechanism is abnormal based on the filtered mechanism state temperature timing significant aggregation representation vector.

[0009] In the above-mentioned magnetic levitation safety electrical device, the mechanism temperature fluctuation feature extraction module includes: an empirical mode decomposition unit, which is used to perform empirical mode decomposition on the time queue of the real-time temperature of the magnetic levitation isolation mechanism to obtain a series of mechanism state temperature inherent mode function components; a temperature timing feature extraction unit, which is used to pass each mechanism state temperature inherent mode function component in the series of mechanism state temperature inherent mode function components through a temperature timing feature extractor based on a 1D-CNN model to obtain the waveform feature vector of the series of mechanism state temperature inherent mode function components.

[0010] In the above-mentioned magnetic levitation safety electrical device, the mechanism temperature timing information aggregation module includes: a cluster initial center calculation unit, which is used to determine the mechanism state temperature waveform feature cluster initial center vector based on the static energy distribution of the series of mechanism state temperature inherent modal function component waveform feature vectors; a dynamic aggregation weight generation unit, which is used to calculate the dynamic aggregation weight factor of each mechanism state temperature inherent modal function component waveform feature vector in the series of mechanism state temperature inherent modal function component waveform feature vectors based on the spatial span of each mechanism state temperature inherent modal function component waveform feature vector relative to the mechanism state temperature waveform feature cluster initial center vector to obtain a series of dynamic aggregation weight factors; an inherent modal function component explicit aggregation unit, which is used to calculate the weighted sum of the series of mechanism state temperature inherent modal function component waveform feature vectors based on the series of dynamic aggregation weight factors to obtain the mechanism state temperature timing significant aggregation representation vector.

[0011] In the above-mentioned magnetic levitation safety electrical device, the cluster initial center calculation unit includes: a static energy factor calculation subunit, which is used to calculate the static energy factor of each mechanism state temperature inherent modal function component waveform feature vector in the series of mechanism state temperature inherent modal function component waveform feature vectors to obtain a series of static energy factors; a cluster initial center vector selection subunit, which is used to select the mechanism state temperature inherent modal function component waveform feature vector corresponding to the largest static energy factor in the series of static energy factors as the mechanism state temperature waveform feature cluster initial center vector.

[0012] In the above-mentioned magnetic levitation safety electrical device, the static energy factor calculation subunit is used to: calculate the kurtosis of the waveform characteristic vector of the temperature intrinsic modal function component of the mechanism state, and input the kurtosis into the sigmoid activation function to obtain the static energy factor.

[0013] In the above-mentioned magnetic levitation safety electrical device, the dynamic aggregation weight generation unit includes: a dynamic aggregation energy factor calculation subunit, which is used to calculate the dynamic aggregation energy factor of each mechanism state temperature inherent modal function component waveform feature vector in the series of mechanism state temperature inherent modal function component waveform feature vectors based on the spatial span between each mechanism state temperature inherent modal function component waveform feature vector in the series of mechanism state temperature inherent modal function component waveform feature vectors and the initial center vector of the mechanism state temperature waveform feature cluster, and the static energy factor of each mechanism state temperature inherent modal function component waveform feature vector to obtain a series of dynamic aggregation energy factors; a masking subunit, which is used to input the series of dynamic aggregation energy factors into the gated mask unit to obtain the series of dynamic aggregation weight factors.

[0014] In the above-mentioned magnetic levitation safety electrical device, the dynamic aggregation energy factor calculation subunit is used to: use the square value of the number of characteristic vectors between the waveform characteristic vector of the temperature inherent modal function component of the mechanism state temperature and the initial center vector of the mechanism state temperature waveform characteristic cluster as the spatial span coefficient, and calculate the weighted ratio between the product of the static energy factor of the waveform characteristic vector of the temperature inherent modal function component of the mechanism state temperature and the static energy factor of the initial center vector of the mechanism state temperature waveform characteristic cluster and the spatial span coefficient to obtain the dynamic aggregation energy factor.

[0015] In the above-mentioned magnetic levitation safety electrical device, the feature filtering module includes: an ambient temperature time series feature extraction unit, which is used to input the time queue of the external ambient temperature into an ambient temperature time series feature capturer based on a forward LSTM model to obtain an external ambient temperature time series feature representation vector; an environmental interference filtering unit, which is used to input the mechanism state temperature time series significant aggregation representation vector and the external ambient temperature time series feature representation vector into an ambient temperature interference filtering module to obtain the filtered mechanism state temperature time series significant aggregation representation vector.

[0016] In the above-mentioned magnetic levitation safety electrical device, the working status detection module is used to: input the filtered mechanism state temperature time series significant aggregation representation vector into the classifier-based state detection result generator to obtain the state detection result, and the state detection result is used to indicate whether there is an abnormality in the state of the magnetic levitation partition mechanism.

[0017] According to another aspect of the present application, a method for remotely monitoring the working state of a magnetic levitation safety electrical device is provided, comprising:

[0018] A time queue for receiving the real-time temperature of the magnetic suspension partition mechanism collected by the first temperature sensor and a time queue for receiving the external environment temperature collected by the second temperature sensor;

[0019] Performing empirical mode decomposition and temperature fluctuation feature extraction on the time queue of the real-time temperature of the magnetic levitation partition mechanism to obtain a series of waveform feature vectors of the intrinsic mode function components of the mechanism state temperature;

[0020] Performing significant aggregation based on characteristic energy distribution on the series of waveform characteristic vectors of the intrinsic modal function components of the mechanism state temperature to obtain a significant aggregation representation vector of the mechanism state temperature time series;

[0021] Based on the temperature time series characteristics in the time queue of the external environment temperature, filtering the significant aggregation representation vector of the mechanism state temperature time series to remove the environmental temperature interference to obtain a filtered significant aggregation representation vector of the mechanism state temperature time series;

[0022] Based on the filtered mechanism state temperature time series significant aggregation representation vector, determine whether the state of the magnetic levitation partition mechanism is abnormal.

[0023] Compared with the prior art, the magnetic levitation safety electrical device and its working state remote monitoring method provided by the present application synchronously collect the real-time temperature data of the magnetic levitation partition mechanism and the external environment through the first temperature sensor deployed in the magnetic levitation partition mechanism and the second temperature sensor deployed in the external environment, and deploy the working state remote monitoring server at the back end, and use the deep learning algorithm to perform time series analysis on the real-time temperature data of the magnetic levitation partition mechanism and the external environment, respectively dig out the time series change patterns of the mechanism temperature and the ambient temperature, and perform feature comparison analysis on the two to filter out the interference effect of the external environment temperature on the mechanism state temperature, thereby achieving more accurate abnormal state detection of the magnetic levitation partition mechanism. In this way, it can fully adapt to the state monitoring needs of the magnetic levitation partition mechanism under different environmental conditions, improve the accuracy of the monitoring results, and further improve the safety and reliability of power equipment. BRIEF DESCRIPTION OF THE DRAWINGS

[0024] By describing the embodiments of the present application in more detail in conjunction with the accompanying drawings, the above and other purposes, features and advantages of the present application will become more apparent. The accompanying drawings are used to provide a further understanding of the embodiments of the present application and constitute a part of the specification. Together with the embodiments of the present application, they are used to explain the present application and do not constitute a limitation of the present application. In the accompanying drawings, the same reference numerals generally represent the same components or steps.

[0025] Figure 1 It is a block diagram of a remote monitoring server for the working status of a magnetic levitation safety electrical device according to an embodiment of the present application.

[0026] Figure 2 It is a block diagram of a mechanism temperature fluctuation feature extraction module in a magnetic levitation safety electrical device according to an embodiment of the present application.

[0027] Figure 3 4 is a block diagram of a mechanism temperature timing information aggregation module in a magnetic levitation safety electrical device according to an embodiment of the present application.

[0028] Figure 4 The block diagram is a dynamic aggregation weight generation unit in a magnetic levitation safety electrical device according to an embodiment of the present application.

[0029] Figure 5 4 is a block diagram of a feature filtering module in a magnetic levitation safety electrical device according to an embodiment of the present application.

[0030] Figure 6 Flow chart of a method for remotely monitoring the working status of a magnetic levitation safety electrical device according to an embodiment of the present application. DETAILED DESCRIPTION

[0031] As shown in this application and claims, unless the context clearly indicates an exception, the words "a", "an", "an" and / or "the" do not refer to the singular and may also include the plural. Generally speaking, the terms "include" and "comprise" only indicate the inclusion of the steps and elements that have been clearly identified, and these steps and elements do not constitute an exclusive list. The method or device may also include other steps or elements.

[0032] Below, the exemplary embodiments according to the present application will be described in detail with reference to the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present application, rather than all the embodiments of the present application, and it should be understood that the present application is not limited to the exemplary embodiments described here.

[0033] As described in the above background technology, invention patent CN118300280A proposes a magnetic levitation safety electrical device and a state detection method, which includes: a phase line interface terminal for connecting the phase line; a ground line interface terminal for connecting the ground line; a magnetic levitation isolation mechanism arranged between the phase line interface terminal and the ground line interface terminal, wherein the magnetic levitation isolation mechanism includes a primary coil and a secondary coil opposite to the primary coil, the primary coil is electrically connected to the phase line interface terminal, and the secondary coil is electrically connected to the ground line interface terminal.

[0034] During the operation of the above-mentioned magnetic levitation safety electrical device, the primary coil converts the electrical energy from the phase line into magnetic energy, and the secondary coil is used to convert the magnetic energy into electrical energy through electromagnetic induction, so as to achieve electrical connectivity between the phase line and the ground line without contacting the phase line and the ground line. In addition, in order to reduce the subjective errors and abnormal state response lag caused by traditional manual regular inspections or regular maintenance, the above-mentioned magnetic levitation safety electrical device also includes a temperature sensor and an early warning processor, which are used to monitor the temperature of the magnetic levitation partition mechanism in real time, and analyze and process the mechanism temperature data to achieve real-time monitoring of its working state and abnormal early warning prompts.

[0035] However, since the temperature change of the external environment will affect the temperature inside the device, and the above scheme does not take into account the interference of the external temperature change on the temperature monitoring of the mechanism, which may affect the accuracy of the monitoring results.

[0036] In response to the above technical problems, this application proposes a new type of magnetic levitation safety electrical device and its working state remote monitoring method, which synchronously collects the real-time temperature data of the magnetic levitation partition mechanism and the external environment through a first temperature sensor deployed in the magnetic levitation partition mechanism and a second temperature sensor deployed in the external environment, and deploys a working state remote monitoring server at the back end, and uses a deep learning algorithm to perform time series analysis on the real-time temperature data of the magnetic levitation partition mechanism and the external environment, respectively mining the time series change patterns of the mechanism temperature and the ambient temperature, and filtering out the interference of the external environment temperature on the mechanism state temperature by performing feature comparison analysis on the two, thereby achieving more accurate abnormal state detection of the magnetic levitation partition mechanism. In this way, it can fully adapt to the state monitoring needs of the magnetic levitation partition mechanism under different environmental conditions, improve the accuracy of the monitoring results, and further improve the safety and reliability of power equipment.

[0037] Figure 1 FIG. 1 is a block diagram of a remote monitoring server for the working state of the magnetic levitation safety electrical device according to an embodiment of the present application. Figure 1As shown, the working status remote monitoring server 100 includes: a temperature data receiving module 110, which is used to receive the time queue of the real-time temperature of the magnetic levitation partition mechanism collected by the first temperature sensor, and the time queue of the external environment temperature collected by the second temperature sensor; a mechanism temperature fluctuation feature extraction module 120, which is used to perform empirical mode decomposition and temperature fluctuation feature extraction on the time queue of the real-time temperature of the magnetic levitation partition mechanism to obtain a series of mechanism state temperature intrinsic mode function component waveform feature vectors; a mechanism temperature time series information aggregation module 130, which is used to perform significant aggregation based on characteristic energy distribution on the series of mechanism state temperature intrinsic mode function component waveform feature vectors to obtain a mechanism state temperature time series significant aggregation representation vector; a feature filtering module 140, which is used to filter out environmental temperature interference on the mechanism state temperature time series significant aggregation representation vector based on the temperature time series characteristics in the time queue of the external environment temperature to obtain a filtered mechanism state temperature time series significant aggregation representation vector; a working status detection module 150, which is used to determine whether the state of the magnetic levitation partition mechanism is abnormal based on the filtered mechanism state temperature time series significant aggregation representation vector.

[0038] In the above-mentioned working status remote monitoring server 100, the temperature data receiving module 110 is used to receive the time queue of the real-time temperature of the magnetic levitation partition mechanism collected by the first temperature sensor. Specifically, in the technical solution of the present application, first, the time queue of the real-time temperature of the magnetic levitation partition mechanism is collected by the first temperature sensor, and the time queue of the external environment temperature is collected by the second temperature sensor. Next, considering the limitations of edge computing resources, in order to improve the efficiency and accuracy of data processing, the present application further transmits the time queue of the real-time temperature of the magnetic levitation partition mechanism and the time queue of the external environment temperature to the working status remote monitoring server through a wireless communication module, so as to utilize the powerful computing power of the server for subsequent temperature data processing and analysis, thereby realizing remote real-time monitoring of the working status of the magnetic levitation partition mechanism. In an embodiment of the present application, the wireless communication module can be based on 5G, Wi-Fi or other wireless communication technologies.

[0039] In the above-mentioned remote monitoring server 100 for working status, the mechanism temperature fluctuation feature extraction module 120 is used to perform empirical mode decomposition and temperature fluctuation feature extraction on the time queue of the real-time temperature of the magnetic levitation partition mechanism to obtain a series of mechanism state temperature intrinsic mode function component waveform feature vectors. Figure 2 FIG. 1 is a block diagram of a temperature fluctuation feature extraction module of a magnetic levitation safety electrical device according to an embodiment of the present application. Figure 2As shown, the mechanism temperature fluctuation feature extraction module 120 includes: an empirical mode decomposition unit 121, which is used to perform empirical mode decomposition on the time queue of the real-time temperature of the magnetic levitation partition mechanism to obtain a series of mechanism state temperature inherent mode function components; a temperature time series feature extraction unit 122, which is used to pass each mechanism state temperature inherent mode function component in the series of mechanism state temperature inherent mode function components through a temperature time series feature extractor based on a 1D-CNN model to obtain a series of mechanism state temperature inherent mode function component waveform feature vectors.

[0040] Specifically, the empirical mode decomposition unit 121 is used to perform empirical mode decomposition on the time queue of the real-time temperature of the magnetic levitation partition mechanism to obtain a series of mechanism state temperature intrinsic mode function components. It should be understood that, considering that the real-time temperature data of the magnetic levitation partition mechanism usually has nonlinear and non-stationary characteristics, in order to more carefully and accurately extract the temperature change characteristics of the magnetic levitation partition mechanism, the present application adopts the empirical mode decomposition method (EMD) to process the time queue of the real-time temperature of the magnetic levitation partition mechanism to decompose it into a series of mechanism state temperature intrinsic mode function components. Among them, each intrinsic mode function component contains the temperature change information of the magnetic levitation partition mechanism on a specific time scale, reflects the local time domain change characteristics of the temperature, and helps to capture the subtle fluctuations and change trends of the temperature more finely. It is worth mentioning that the empirical mode decomposition method is an adaptive signal processing method, which is suitable for analyzing nonlinear and non-stationary signals, and can decompose complex signals into a series of simple intrinsic mode functions with different frequencies.

[0041] Specifically, the temperature time series feature extraction unit 122 is used to pass each mechanism state temperature intrinsic modal function component in the series of mechanism state temperature intrinsic modal function components through a temperature time series feature extractor based on a 1D-CNN model to obtain the series of mechanism state temperature intrinsic modal function component waveform feature vectors. It should be understood that in order to further extract high-level temperature change features from each mechanism state temperature intrinsic modal function component, the present application adopts a 1D-CNN model to construct a temperature time series feature extractor, so as to utilize the powerful time series data processing capability of the 1D-CNN model to extract features of each mechanism state temperature intrinsic modal function component, and perform sliding convolution operations along the time dimension on each mechanism state temperature intrinsic modal function component to extract key features therein, so as to effectively identify and learn the temperature time series dynamic change mode in each mechanism state temperature intrinsic modal function component, and generate corresponding temperature waveform feature descriptions to obtain a series of mechanism state temperature intrinsic modal function component waveform feature vectors. It is worth mentioning that the 1D CNN network (one-dimensional convolutional neural network) is a deep learning model, which is mainly used to process time series data or any other form of one-dimensional sequence data, and is particularly suitable for processing the temperature intrinsic mode function components of each mechanism state.

[0042] In the above-mentioned working status remote monitoring server 100, the mechanism temperature time series information aggregation module 130 is used to perform significant aggregation on the series of mechanism state temperature intrinsic modal function component waveform feature vectors based on characteristic energy distribution to obtain the mechanism state temperature time series significant aggregation representation vector. That is, in order to further conduct a comprehensive analysis of the temperature state of the mechanism from a global perspective to achieve a comprehensive evaluation of the working state of the magnetic levitation partition mechanism, it is necessary to perform feature aggregation on the series of mechanism state temperature intrinsic modal function component waveform feature vectors to capture the overall temperature change trend of the magnetic levitation partition mechanism. It is worth mentioning that, considering that the series of mechanism state temperature intrinsic modal function component waveform feature vectors may contain some irrelevant information such as noise and redundancy, errors may be introduced when directly performing feature aggregation. Therefore, in order to better simulate and understand the dynamic changes of mechanism state temperature, this application proposes a feature significance aggregation method based on characteristic energy distribution. By considering the spatial relationship and characteristic energy distribution between the series of mechanism state temperature intrinsic modal function component waveform feature vectors, the feature importance of each mechanism state temperature intrinsic modal function component waveform feature vector is dynamically adjusted, and the information is integrated into a more compact and representative representation according to a specific aggregation strategy.

[0043] in, Figure 3FIG. 1 is a block diagram of a temperature timing information aggregation module of a magnetic levitation safety electrical device according to an embodiment of the present application. Figure 3 As shown, the mechanism temperature timing information aggregation module 130 includes: a cluster initial center calculation unit 131, which is used to determine the mechanism state temperature waveform feature cluster initial center vector based on the static energy distribution of the series of mechanism state temperature intrinsic modal function component waveform feature vectors; a dynamic aggregation weight generation unit 132, which is used to calculate the dynamic aggregation weight factor of each mechanism state temperature intrinsic modal function component waveform feature vector in the series of mechanism state temperature intrinsic modal function component waveform feature vectors based on the spatial span of each mechanism state temperature intrinsic modal function component waveform feature vector relative to the mechanism state temperature waveform feature cluster initial center vector to obtain a series of dynamic aggregation weight factors; an intrinsic modal function component explicit aggregation unit 133, which is used to calculate the weighted sum of the series of mechanism state temperature intrinsic modal function component waveform feature vectors based on the series of dynamic aggregation weight factors to obtain the mechanism state temperature timing significant aggregation representation vector.

[0044] Specifically, the cluster initial center calculation unit 131 includes: a static energy factor calculation subunit, which is used to calculate the static energy factor of each mechanism state temperature inherent modal function component waveform feature vector in the series of mechanism state temperature inherent modal function component waveform feature vectors to obtain a series of static energy factors; a cluster initial center vector selection subunit, which is used to select the mechanism state temperature inherent modal function component waveform feature vector corresponding to the largest static energy factor in the series of static energy factors as the mechanism state temperature waveform feature cluster initial center vector.

[0045] More specifically, the static energy factor calculation subunit is further used to: calculate the kurtosis of the waveform feature vector of the temperature intrinsic modal function component of the mechanism state, and input the kurtosis into a sigmoid activation function to obtain the static energy factor.

[0046] Figure 4 FIG. 1 is a block diagram of a dynamic aggregation weight generation unit in a magnetic levitation safety electrical device according to an embodiment of the present application. Figure 4As shown, the dynamic aggregation weight generating unit 132 includes: a dynamic aggregation energy factor calculating subunit 1321, which is used to calculate the dynamic aggregation energy factor of each mechanism state temperature intrinsic modal function component waveform feature vector in the series of mechanism state temperature intrinsic modal function component waveform feature vectors based on the spatial span between each mechanism state temperature intrinsic modal function component waveform feature vector in the series of mechanism state temperature intrinsic modal function component waveform feature vectors and the initial center vector of the mechanism state temperature waveform feature clustering, and the static energy factor of each mechanism state temperature intrinsic modal function component waveform feature vector to obtain a series of dynamic aggregation energy factors; a masking subunit 1322, which is used to input the series of dynamic aggregation energy factors into a gated mask unit to obtain the series of dynamic aggregation weight factors.

[0047] Specifically, the dynamic aggregation energy factor calculation subunit 1321 is further used to: use the square value of the number of eigenvectors between the waveform eigenvector of the temperature inherent modal function component of the mechanism state temperature and the initial center vector of the mechanism state temperature waveform feature cluster as the spatial span coefficient, and calculate the weighted ratio between the product of the static energy factor of the waveform eigenvector of the temperature inherent modal function component of the mechanism state temperature and the static energy factor of the initial center vector of the mechanism state temperature waveform feature cluster and the spatial span coefficient to obtain the dynamic aggregation energy factor.

[0048] Specifically, the masking subunit 1322 is used to input the series of dynamic aggregation energy factors into the gated mask unit to obtain the series of dynamic aggregation weight factors, and is further used to: input the series of dynamic aggregation energy factors into the sigmoid function for normalization processing to obtain a series of normalized dynamic aggregation energy factors; based on a preset mask threshold, inactivate the series of normalized dynamic aggregation energy factors to obtain the series of dynamic aggregation weight factors.

[0049] Correspondingly, the mechanism temperature time series information aggregation module 130 is used to: process the series of mechanism state temperature intrinsic mode function component waveform feature vectors using the following feature aggregation formula to obtain the mechanism state temperature time series significant aggregation representation vector, wherein the feature aggregation formula is:

[0050] X=(x 1 , x 2 , ..., x i , ..., x n}

[0051]

[0052] x c=x m

[0053]

[0054] w si =mask(w i )

[0055]

[0056] Wherein, X represents the waveform characteristic vector of the temperature intrinsic modal function component of the series of mechanism states, x 1 、x 2 、x i 、x m and x n respectively represent the first, second, i-th, m-th and n-th mechanism state temperature intrinsic modal function component waveform feature vectors in the series of mechanism state temperature intrinsic modal function component waveform feature vectors, and the value of n is the number of the mechanism state temperature intrinsic modal function component waveform feature vectors, represents the eigenvalue of the jth position in the waveform eigenvector of the temperature intrinsic modal function component of the i-th mechanism state, μ i and σ i 4 denote the characteristic mean and square of the characteristic variance of the waveform characteristic vector of the temperature intrinsic modal function component of the i-th mechanism state, E{·} denotes the expected value of the calculation set, sigmoid denotes the sigmoid activation function, represents the static energy factor of the waveform characteristic vector of the temperature intrinsic modal function component of the i-th mechanism state, argmax represents the index corresponding to the maximum value, m represents the index of the largest static energy factor in the series of static energy factors, x c represents the initial center vector of the clustering of the temperature waveform characteristics of the mechanism state, represents the static energy factor of the initial center vector of the clustering of the temperature waveform characteristics of the mechanism state, a and b are different weight parameters, Count(x i →x m ) represents the number of characteristic vectors between the waveform characteristic vector of the ith mechanism state temperature intrinsic mode function component and the initial center vector of the mechanism state temperature waveform characteristic cluster, represents the dynamic aggregate energy factor of the waveform eigenvector of the temperature intrinsic modal function component of the i-th mechanism state, w i represents the i-th normalized dynamic aggregation energy factor, mask(·) represents mask processing, θ is the preset mask threshold, and w si represents the i-th dynamic aggregation weight factor, x fA vector representing a significant aggregation of the temperature time series of the mechanism state.

[0057] That is, first, the static energy factor of each mechanism state temperature intrinsic modal function component waveform feature vector is calculated based on a predefined metric function to evaluate its stability and importance in the feature space, and the mechanism state temperature intrinsic modal function component waveform feature vector corresponding to the largest static energy factor is selected as the mechanism state temperature waveform feature initial center vector to determine the cluster core of the mechanism temperature feature distribution, thereby providing a stable reference point for subsequent feature aggregation. Then, based on the spatial span of each mechanism state temperature intrinsic modal function component waveform feature vector relative to the mechanism state temperature waveform feature initial center vector, and combined with their respective static energy factors, the dynamic aggregation energy factor of each mechanism state temperature intrinsic modal function component waveform feature vector is further derived. That is, by considering the feature space proximity of each mechanism state temperature intrinsic modal function component waveform feature vector relative to the temperature feature distribution cluster core and the importance difference of the intrinsic attributes, its contribution to the global time domain feature of temperature is more accurately measured. Then, based on the gated mask mechanism, the obtained series of dynamic aggregation energy factors are nonlinearly transformed and screened to generate corresponding weights, so as to weightedly aggregate the waveform feature vectors of the series of mechanism state temperature intrinsic modal function components, thereby enhancing the model's attention to key information and suppressing irrelevant parts to obtain the mechanism state temperature time series significant aggregation representation vector.

[0058] In the above-mentioned remote monitoring server 100 for working status, the feature filtering module 140 is used to filter out the ambient temperature interference of the significant aggregation representation vector of the mechanism state temperature time series based on the temperature time series characteristics in the time queue of the external environment temperature to obtain the filtered significant aggregation representation vector of the mechanism state temperature time series. Figure 5 FIG. 1 is a block diagram of the feature filtering module in the magnetic levitation safety electrical device according to an embodiment of the present application. Figure 5 As shown, the feature filtering module 140 includes: an ambient temperature time series feature extraction unit 141, which is used to input the time queue of the external ambient temperature into an ambient temperature time series feature capturer based on a forward LSTM model to obtain an external ambient temperature time series feature representation vector; an environmental interference filtering unit 142, which is used to input the mechanism state temperature time series significant aggregation representation vector and the external ambient temperature time series feature representation vector into an ambient temperature interference filtering module to obtain the filtered mechanism state temperature time series significant aggregation representation vector.

[0059] Specifically, the ambient temperature time series feature extraction unit 141 is used to input the time queue of the external ambient temperature into the ambient temperature time series feature capturer based on the forward LSTM model to obtain the external ambient temperature time series feature representation vector. It should be understood that in order to filter out the influence of the external ambient temperature on the temperature monitoring of the magnetic levitation partition mechanism, the present application further introduces an ambient temperature compensation mechanism. First, the time queue of the external ambient temperature is subjected to a time series analysis by a forward LSTM model, so as to utilize the long and short-term memory characteristics of the forward LSTM model to capture the dynamic characteristics of the ambient temperature changing over time, thereby extracting a feature representation that can reflect the trend of ambient temperature changes, so as to obtain the external ambient temperature time series feature representation vector. It is worth mentioning that LSTM (Long Short-Term Memory Network) is a special recurrent neural network architecture that performs well in processing sequence data, especially time series analysis and natural language processing tasks.

[0060] Specifically, the environmental interference filtering unit 142 is used to input the significant aggregation representation vector of the mechanism state temperature time series and the external environment temperature time series feature representation vector into the environmental temperature interference filtering module to obtain the filtered mechanism state temperature time series significant aggregation representation vector. That is, the mechanism state temperature time series significant aggregation representation vector and the external environment temperature time series feature representation vector are input into the environmental temperature interference filtering module for feature filtering. The environmental temperature interference filtering module performs position-by-position feature differentiation on the two to eliminate the interference of environmental temperature changes on the temperature monitoring of the magnetic levitation partition mechanism, thereby obtaining the filtered mechanism state temperature time series significant aggregation representation vector, so as to more truly reflect the actual working temperature state of the magnetic levitation partition mechanism and provide reliable data support for subsequent fault diagnosis and predictive maintenance.

[0061] In the above-mentioned remote monitoring server 100 for working status, the working status detection module 150 is used to: input the filtered mechanism state temperature time series significant aggregation representation vector into the state detection result generator based on the classifier to obtain the state detection result, and the state detection result is used to indicate whether the state of the magnetic levitation partition mechanism is abnormal, and further used to: input the filtered mechanism state temperature time series significant aggregation representation vector into the state detection result generator based on the classifier to obtain the state detection result, and the state detection result is used to indicate whether the state of the magnetic levitation partition mechanism is abnormal. That is, the classifier model is used to classify and judge the filtered mechanism state temperature time series significant aggregation representation vector. Through a large amount of training, the classifier can effectively learn the temperature change pattern in the filtered mechanism state temperature time series significant aggregation representation vector, and then accurately distinguish the normal working state and potential abnormal state of the magnetic levitation partition mechanism, and output the corresponding state detection result, so as to provide real-time working state feedback to the operator, so as to realize real-time monitoring and abnormality detection of the magnetic levitation partition mechanism.

[0062] In a preferred example of the present application, the filtered mechanism state temperature time series significant aggregation representation vector is passed through a classifier-based state detection result generator to obtain a state detection result, including the steps of: calculating the absolute value sum of each eigenvalue of the filtered mechanism state temperature time series significant aggregation representation vector to obtain a first filtered mechanism state temperature time series significant aggregation spatial structure value, and calculating the square root of the sum of squares of each eigenvalue of the filtered mechanism state temperature time series significant aggregation representation vector to obtain a second filtered mechanism state temperature time series significant aggregation spatial structure value; multiplying each eigenvalue of the filtered mechanism state temperature time series significant aggregation representation vector with the first filtered mechanism state temperature time series significant aggregation spatial structure value and the second filtered mechanism state temperature time series significant aggregation spatial structure value to obtain a first filtered mechanism state temperature time series significant aggregation structure reference value and a second filtered mechanism state temperature time series significant aggregation structure reference value corresponding to each eigenvalue; multiplying each eigenvalue of the filtered mechanism state temperature time series significant aggregation representation vector with the length of the filtered mechanism state temperature time series significant aggregation representation vector and the square root of the length to obtain a first filtered mechanism state temperature time series significant aggregation structure reference value and a second filtered mechanism state temperature time series significant aggregation structure reference value corresponding to each eigenvalue; multiplying each eigenvalue of the filtered mechanism state temperature time series significant aggregation representation vector with the length of the filtered mechanism state temperature time series significant aggregation representation vector and the square root of the length to obtain a second filtered mechanism state temperature time series significant aggregation structure reference value. The first filtered mechanism state temperature time series significant aggregation scale transformation value and the second filtered mechanism state temperature time series significant aggregation scale transformation value corresponding to the eigenvalue; the first filtered mechanism state temperature time series significant aggregation structure reference value is divided by the difference between the first filtered mechanism state temperature time series significant aggregation space structure value and the first filtered mechanism state temperature time series significant aggregation scale transformation value to obtain the first filtered mechanism state temperature time series significant aggregation transformation adjustment value; the second filtered mechanism state temperature time series significant aggregation structure reference value is divided by the difference between the second filtered mechanism state temperature time series significant aggregation space structure value and the second filtered mechanism state temperature time series significant aggregation scale transformation value to obtain the second filtered mechanism state temperature time series significant aggregation transformation adjustment value; the weighted sum of the first filtered mechanism state temperature time series significant aggregation transformation adjustment value and the second filtered mechanism state temperature time series significant aggregation transformation adjustment value is calculated to obtain each eigenvalue of the optimized filtered mechanism state temperature time series significant aggregation representation vector; the optimized filtered mechanism state temperature time series significant aggregation representation vector is passed through a classifier-based state detection result generator to obtain a state detection result.

[0063] Accordingly, the optimization process of the significant aggregation representation vector of the filtered mechanism state temperature time series is performed using the following optimization formula:

[0064]

[0065] v 1i =(α×v i ) / (α-L×v i )

[0066]

[0067] v i ∈V∈R 1×L

[0068] v 1i ∈V 1 ∈R 1×L

[0069] v 2i ∈V 2 ∈R 1×L

[0070] Wherein, V represents the significant aggregation representation vector of the temperature time series of the mechanism state after filtering, v i represents the ith eigenvalue of the filtered mechanism state temperature time series significant aggregation representation vector, L represents the length of the filtered mechanism state temperature time series significant aggregation representation vector, R represents a real number set, α represents the first filtered mechanism state temperature time series significant aggregation space structure value, β represents the second filtered mechanism state temperature time series significant aggregation space structure value, v 1i represents the first filtered mechanism state temperature time series significant aggregation scale transformation value corresponding to the i-th eigenvalue in the filtered mechanism state temperature time series significant aggregation representation vector, V 1 is the first filtered mechanism state temperature time series significant aggregation transformation adjustment vector composed of the first filtered mechanism state temperature time series significant aggregation scale transformation values, v 2i represents the second filtered mechanism state temperature time series significant aggregation scale transformation value corresponding to the i-th eigenvalue in the filtered mechanism state temperature time series significant aggregation representation vector, V 2 is the second filtered mechanism state temperature time series significant aggregation transformation adjustment vector composed of the second filtered mechanism state temperature time series significant aggregation scale transformation values, ω is a weight hyperparameter, represents point-wise addition, ⊙ represents point-wise multiplication, and V′ represents the optimized filtered mechanism state temperature time series significant aggregation representation vector.

[0071] That is, the present application takes into account that the mechanism state temperature time series significant aggregation representation vector and the external environment temperature time series feature representation vector respectively represent the frequency domain significant aggregation features determined by the feature distribution field dynamic aggregation network based on the imitation force field of the real-time temperature of the magnetic levitation partition mechanism and the local time series correlation features of the external environment temperature. In the working state remote monitoring server, when the time queue of the external environment temperature is input into the ambient temperature time series feature capturer based on the forward LSTM model, the cross-domain feature distribution heterogeneity (frequency domain temperature features and time domain temperature features) of the mechanism state temperature time series significant aggregation representation vector and the external environment temperature time series feature representation vector will make the feature representation of the filtered mechanism state temperature time series significant aggregation representation vector have a complex aggregation spatial structure, thereby expecting to improve its classification regression convergence and generalization effects under the complex aggregation spatial structure.

[0072] Therefore, in a preferred example, with respect to the spatial structure information of the feature set of the filtered mechanism state temperature time series significant aggregation representation vector in a high-dimensional space, a scale-based frame transformation of each eigenvalue of the filtered mechanism state temperature time series significant aggregation representation vector is performed by taking the class-norm spatial structured representation of the filtered mechanism state temperature time series significant aggregation representation vector as a reference window, and a spatial structure-based frame attention weight adjustment of each eigenvalue of the filtered mechanism state temperature time series significant aggregation representation vector is implemented to ensure the invariance of spatial transformation (translation, scaling and rotation) of the filtered mechanism state temperature time series significant aggregation representation vector under feature space interaction, thereby improving the convergence and generalization effect of classification regression of the feature set of the filtered mechanism state temperature time series significant aggregation representation vector under complex spatial structure representation, and improving the accuracy of the state detection result obtained by passing it through a classifier-based state detection result generator.

[0073] In summary, according to the embodiment of the present application, the magnetic levitation safety electrical device is explained, which synchronously collects the real-time temperature data of the magnetic levitation partition mechanism and the external environment through the first temperature sensor deployed in the magnetic levitation partition mechanism and the second temperature sensor deployed in the external environment, and deploys the working status remote monitoring server at the back end, and uses the deep learning algorithm to perform time series analysis on the real-time temperature data of the magnetic levitation partition mechanism and the external environment, respectively dig out the time series change patterns of the mechanism temperature and the ambient temperature, and perform feature comparison analysis on the two to filter out the interference effect of the external environment temperature on the mechanism state temperature, thereby achieving more accurate abnormal state detection of the magnetic levitation partition mechanism. In this way, it can fully adapt to the state monitoring needs of the magnetic levitation partition mechanism under different environmental conditions, improve the accuracy of the monitoring results, and further improve the safety and reliability of power equipment.

[0074] Furthermore, a method for remotely monitoring the working status of a magnetic levitation safety electrical device is also provided.

[0075] Figure 6 FIG. 1 is a flow chart of a method for remotely monitoring the working state of a magnetic levitation safety electrical device according to an embodiment of the present application. Figure 6 As shown, according to the embodiment of the present application, the remote monitoring method for the working state of the magnetic levitation safety electrical device includes the following steps: S1, receiving the time queue of the real-time temperature of the magnetic levitation partition mechanism collected by the first temperature sensor, and the time queue of the external environment temperature collected by the second temperature sensor; S2, performing empirical mode decomposition and temperature fluctuation feature extraction on the time queue of the real-time temperature of the magnetic levitation partition mechanism to obtain a series of mechanism state temperature intrinsic mode function component waveform feature vectors; S3, performing significant aggregation based on characteristic energy distribution on the series of mechanism state temperature intrinsic mode function component waveform feature vectors to obtain a mechanism state temperature time series significant aggregation representation vector; S4, based on the temperature time series characteristics in the time queue of the external environment temperature, performing environmental temperature interference filtering on the mechanism state temperature time series significant aggregation representation vector to obtain a filtered mechanism state temperature time series significant aggregation representation vector; S5, based on the filtered mechanism state temperature time series significant aggregation representation vector, determining whether the state of the magnetic levitation partition mechanism is abnormal.

[0076] Here, those skilled in the art can understand that the specific operations of each step in the above-mentioned method for remotely monitoring the working state of the magnetic levitation safety electrical device have been referred to above. Figures 1 to 5 The description of the magnetic levitation safety electrical device has been described in detail, and therefore, its repeated description will be omitted.

[0077] The basic principle of the present invention is described above in conjunction with specific embodiments. However, it should be pointed out that the advantages, strengths, effects, etc. mentioned in the present invention are only examples and not limitations, and it cannot be considered that these advantages, strengths, effects, etc. must be possessed by each embodiment of the present invention. In addition, the specific details of the above embodiments are only for the purpose of illustration and facilitation of understanding, rather than limitation, and the above details do not limit the present invention to being implemented by adopting the above specific details.

[0078] In the above embodiments, the description of each embodiment has its own emphasis. For the parts that are not described or recorded in detail in a certain embodiment, please refer to the relevant description of other embodiments. In the several embodiments provided by the present invention, it should be understood that the disclosed system and method can be implemented in other ways. For example, the system embodiment described above is only schematic. For example, the unit division is only a logical function division, and there may be other division methods in actual implementation. The units described as separate components may or may not be physically separated, and the components displayed as units may or may not be physical units, that is, they may be located in one place, or they may be distributed on multiple network units. Some or all of the units may be selected according to actual needs to achieve the purpose of the scheme of this embodiment.

[0079] It will be apparent to those skilled in the art that the invention is not limited to the details of the exemplary embodiments described above and that the invention can be implemented in other specific forms without departing from the spirit or essential features of the invention. Therefore, the embodiments should be considered exemplary and non-limiting in all respects, and the scope of the invention is defined by the appended claims rather than the foregoing description, and it is intended that all variations falling within the meaning and scope of the equivalent elements of the claims be included in the invention. Any reference to a figure in a claim should not be considered as limiting the claim to which it relates.

[0080] In addition, it is obvious that the word "comprising" does not exclude other units or steps, and the singular does not exclude the plural. The multiple units stated in the system claims can also be implemented by one unit through software or hardware.

[0081] Finally, it should be noted that the above description has been given for the purpose of illustration and description. In addition, the above embodiments are only used to illustrate the technical solution of the present invention and are not intended to limit it. Although the present invention has been described in detail with reference to the preferred embodiments, it should be understood by those skilled in the art that the technical solution of the present invention can be modified or replaced by equivalents without departing from the spirit and scope of the technical solution of the present invention.

Claims

1. A magnetic levitation safety electrical device, comprising: A phase line interface terminal, used to connect the phase line; a ground line interface terminal, used to connect the ground line; a magnetic levitation partition mechanism disposed between the phase line interface terminal and the ground line interface terminal, wherein the magnetic levitation partition mechanism comprises a primary coil and a secondary coil opposite to the primary coil, the primary coil is electrically connected to the phase line interface terminal, and the secondary coil is electrically connected to the ground line interface terminal; a first temperature sensor deployed in the magnetic levitation partition mechanism, used to collect a time queue of the real-time temperature of the magnetic levitation partition mechanism, characterized in that the first temperature sensor is connected to an external remote monitoring server for working status via a wireless communication module; Wherein, the working status remote monitoring server includes: A temperature data receiving module, used for receiving a time queue of the real-time temperature of the magnetic suspension partition mechanism collected by the first temperature sensor, and a time queue of the external environment temperature collected by the second temperature sensor; A mechanism temperature fluctuation feature extraction module is used to perform empirical mode decomposition and temperature fluctuation feature extraction on the time queue of the real-time temperature of the magnetic levitation partition mechanism to obtain a series of mechanism state temperature intrinsic mode function component waveform feature vectors; A mechanism temperature time series information aggregation module is used to perform significant aggregation on the series of mechanism state temperature intrinsic mode function component waveform feature vectors based on characteristic energy distribution to obtain a mechanism state temperature time series significant aggregation representation vector; A feature filtering module, configured to filter out the ambient temperature interference of the mechanism state temperature time series significant aggregation representation vector based on the temperature time series features in the time queue of the external ambient temperature to obtain a filtered mechanism state temperature time series significant aggregation representation vector; The working state detection module is used to determine whether there is an abnormality in the state of the magnetic levitation partition mechanism based on the significant aggregation representation vector of the temperature time series of the filtered mechanism state.

2. The magnetic levitation safety electrical device according to claim 1, characterized in that: The mechanism temperature fluctuation feature extraction module comprises: An empirical mode decomposition unit, used for performing empirical mode decomposition on the time queue of the real-time temperature of the magnetic levitation partition mechanism to obtain a series of mechanism state temperature intrinsic mode function components; The temperature time series feature extraction unit is used to pass each mechanism state temperature intrinsic modal function component in the series of mechanism state temperature intrinsic modal function components through a temperature time series feature extractor based on a 1D-CNN model to obtain a waveform feature vector of the series of mechanism state temperature intrinsic modal function components.

3. The magnetic levitation safety electrical device according to claim 2, characterized in that: The mechanism temperature time series information aggregation module includes: A cluster initial center calculation unit, used to determine the mechanism state temperature waveform feature cluster initial center vector based on the static energy distribution of the series of mechanism state temperature intrinsic modal function component waveform feature vectors; A dynamic aggregation weight generating unit, for calculating the dynamic aggregation weight factor of each mechanism state temperature intrinsic modal function component waveform feature vector in the series of mechanism state temperature intrinsic modal function component waveform feature vectors based on the spatial span of each mechanism state temperature intrinsic modal function component waveform feature vector relative to the initial center vector of the mechanism state temperature waveform feature clustering to obtain a series of dynamic aggregation weight factors; The intrinsic modal function component explicit aggregation unit is used to calculate the weighted sum of the waveform feature vectors of the intrinsic modal function components of the mechanism state temperature based on the series of dynamic aggregation weight factors to obtain the significant aggregation representation vector of the mechanism state temperature time series.

4. The magnetic levitation safety electrical device according to claim 3, characterized in that: The cluster initial center calculation unit includes: A static energy factor calculation subunit, used for calculating the static energy factor of each mechanism state temperature intrinsic modal function component waveform characteristic vector in the series of mechanism state temperature intrinsic modal function component waveform characteristic vectors to obtain a series of static energy factors; The cluster initial center vector selection subunit is used to select the mechanism state temperature intrinsic mode function component waveform feature vector corresponding to the largest static energy factor in the series of static energy factors as the mechanism state temperature waveform feature cluster initial center vector.

5. The magnetic levitation safety electrical device according to claim 4, characterized in that: The static energy factor calculation subunit is used to: The kurtosis of the waveform characteristic vector of the inherent modal function component of the mechanism state temperature is calculated, and the kurtosis is input into a sigmoid activation function to obtain the static energy factor.

6. The magnetic levitation safety electrical device according to claim 5, characterized in that: The dynamic aggregation weight generating unit includes: A dynamic aggregation energy factor calculation subunit is used to calculate the dynamic aggregation energy factor of each mechanism state temperature intrinsic modal function component waveform feature vector in the series of mechanism state temperature intrinsic modal function component waveform feature vectors based on the spatial span between each mechanism state temperature intrinsic modal function component waveform feature vector in the series of mechanism state temperature intrinsic modal function component waveform feature vectors and the initial center vector of the mechanism state temperature waveform feature cluster, and the static energy factor of each mechanism state temperature intrinsic modal function component waveform feature vector to obtain a series of dynamic aggregation energy factors; The masking subunit is used to input the series of dynamic aggregation energy factors into the gated masking unit to obtain the series of dynamic aggregation weight factors.

7. The magnetic levitation safety electrical device according to claim 6, characterized in that: The dynamic aggregation energy factor calculation subunit is used to: Taking the square value of the number of eigenvectors between the waveform eigenvector of the temperature inherent modal function component of the mechanism state and the initial center vector of the temperature waveform feature cluster of the mechanism state as the spatial span coefficient, the weighted ratio between the product of the static energy factor of the waveform eigenvector of the temperature inherent modal function component of the mechanism state and the static energy factor of the initial center vector of the temperature waveform feature cluster of the mechanism state and the spatial span coefficient is calculated to obtain the dynamic aggregation energy factor.

8. The magnetic levitation safety electrical device according to claim 7, characterized in that: The feature filtering module comprises: An ambient temperature time series feature extraction unit, used for inputting the time queue of the external ambient temperature into an ambient temperature time series feature capturer based on a forward LSTM model to obtain an external ambient temperature time series feature representation vector; The environmental interference filtering unit is used to input the mechanism state temperature time series significant aggregation representation vector and the external environment temperature time series feature representation vector into the environmental temperature interference filtering module to obtain the filtered mechanism state temperature time series significant aggregation representation vector.

9. The magnetic levitation safety electrical device according to claim 8, characterized in that: The working status detection module is used to: The filtered mechanism state temperature time series significant aggregation representation vector is input into a classifier-based state detection result generator to obtain a state detection result, and the state detection result is used to indicate whether there is an abnormality in the state of the magnetic levitation partition mechanism.

10. A method for remotely monitoring the working status of a magnetic levitation safety electrical device, characterized in that: include: A time queue for receiving the real-time temperature of the magnetic suspension partition mechanism collected by the first temperature sensor and a time queue for receiving the external environment temperature collected by the second temperature sensor; Performing empirical mode decomposition and temperature fluctuation feature extraction on the time queue of the real-time temperature of the magnetic levitation partition mechanism to obtain a series of waveform feature vectors of the intrinsic mode function components of the mechanism state temperature; Performing significant aggregation based on characteristic energy distribution on the series of waveform characteristic vectors of the intrinsic modal function components of the mechanism state temperature to obtain a significant aggregation representation vector of the mechanism state temperature time series; Based on the temperature time series characteristics in the time queue of the external environment temperature, filtering the significant aggregation representation vector of the mechanism state temperature time series to remove the environmental temperature interference to obtain a filtered significant aggregation representation vector of the mechanism state temperature time series; Based on the filtered mechanism state temperature time series significant aggregation representation vector, determine whether the state of the magnetic levitation partition mechanism is abnormal.

Images