Isolation switch mechanical state detection method and system based on vibration signals

Abstract

The invention discloses a vibration signal-based isolation switch mechanical state detection method and system, and relates to the technical field of high-voltage isolation switch diagnosis, and the method comprises the steps: arranging a vibration sensor on an isolation switch to collect a vibration signal; performing short-time Fourier transform on the vibration signal; applying a Wiener filter to the Fourier transform signal; extracting feature data of the filtered vibration signal; performing normalization processing on the feature data; inputting the normalized feature data and system state information into a neural network for training; performing fault diagnosis on the vibration signals by using the training neural network; and receiving an analysis result through the data server, generating a fault diagnosis report, and displaying the mechanical state data. According to the isolation switch mechanical state detection method based on the vibration signals, the signal quality is improved through two times of filtering, and the defect diagnosis is more accurate and more comprehensive through multi-point arrangement of the sensors, so that the abruptness is avoided.

Description

Technical Field

[0001] The present invention relates to the technical field of high-voltage isolating switch diagnosis, and in particular to a method and system for detecting the mechanical state of an isolating switch based on a vibration signal. Background Art

[0002] High-voltage disconnectors are the most widely used high-voltage switchgear in substations, and have a great impact on the safe operation of substations. Over the years, accidents of high-voltage disconnectors caused by mechanical failures have occurred frequently and with serious consequences. At present, there are a large number of methods for diagnosing serious faults of insulator cracks at home and abroad, but there are few diagnostic methods for defects such as transmission jamming, inadequate opening and closing, and three-phase phase differences in disconnectors. The existing methods are mainly based on preventive measures and empirical diagnosis. For the detection of mechanical characteristics of disconnectors, although there are currently detection technologies based on experience, mechanical signals, motor current, torque and position information, these technologies are still imperfect. For example, the detection results of experience-based detection technologies depend on personal experience; stress strain detection method and spindle torque sensor detection method are feasible under laboratory conditions, but not suitable for substation environments; manual torque detection method is convenient to operate, but the data fluctuates greatly and cannot effectively distinguish between abnormal and normal data; the detection technology based on motor current signals does not consider the different effects of the working voltage of the disconnector motor in different time periods in actual substations. Since these existing detection technologies cannot intuitively and effectively reflect the mechanical state of the disconnector, it is necessary to study the detection technology of the mechanical characteristics of the disconnector, and propose a detection method for the mechanical state of the disconnector based on vibration signals.

[0003] At present, the main defects of disconnectors are divided into the following three types: transmission jamming, inadequate opening and closing, and three-phase out of phase.

[0004] High-voltage disconnectors work outdoors, and the surfaces of their metal parts are in direct contact with the air, which is prone to chemical reactions such as oxidation, which can cause the material to rust. At the same time, harmful solid particles and gas components in the atmosphere react chemically with water to generate electrolyte solutions, which adhere to the surface of metal parts and accelerate electrochemical corrosion. In addition to external environmental factors, the disconnector itself has mechanical parts such as rotating bearings and shaft pins. When the sealing design is unreasonable, external factors such as dust and rainwater enter the gap of the rotating shaft, and long-term accumulation can also cause internal corrosion of mechanical parts. Rust is a common problem of high-voltage disconnectors, and the rust problem usually causes different degrees of transmission mechanism jamming, which reduces the mechanical strength of the components and makes it difficult to operate the disconnector. In severe cases, it will cause damage and deformation of parts, and a sharp increase in operating torque.

[0005] Two travel limit switches are installed in the electric operating mechanism of the high-voltage disconnector, which cut off the power supply of the drive motor at the end of the opening and closing operations, respectively, to control the start and stop actions of the opening and closing operations. When the contact touches the limit position block, the closed contact is disconnected or the open contact is closed, thereby changing the state of the control circuit. If the position of the limit block is offset, the limit block does not contact the contact when the closing operation is about to end, and the limit switch cannot be cut off, causing the drive motor to continue to rotate, resulting in an over-closing fault; or because the limit nut is loose, the screw length changes, causing the position of the limit switch to change, and the motor stops running before the upper rotating arm of the operating insulator passes the "dead point", which causes the disconnector to fail to open and close.

[0006] Three-phase linkage is an important mechanical motion feature of high-voltage disconnectors. For high-voltage disconnectors, three-phase out-of-phase refers to the larger value of the minimum contact distances of the other two phases when one of the three phases reaches the just-closed position first during the opening and closing operation. After the high-voltage disconnector is disassembled and packed and shipped out of the factory, it needs to be installed and debugged on site. If the debugging is not appropriate, the mechanical dimensions of the cross-link, inter-phase link, and master-slave crank arm will be different from the standard state, which can cause leading or lagging three-phase out-of-phase faults. Summary of the invention

[0007] In view of the above-mentioned problems, the present invention is proposed.

[0008] Therefore, the technical problem solved by the present invention is that the direct detection method is not suitable for the detection of this object because the GIS disconnector has a closed structure, a small internal space, a compact transmission mechanism, and operates in a high-voltage and high-current environment. The sensor is introduced at a high potential, the signal line is difficult to lead out, the signal is interfered by the strong electromagnetic environment, the sealing of the original structure is destroyed, and it is easy to bring insulation risks.

[0009] The state sensing method based on motor current and the state sensing method based on vibration signal theoretically have complete information on the mechanical state of GIS disconnectors. In theory, they can monitor the contact state of GIS disconnectors. However, the existing methods have the following disadvantages: they cannot reflect the phase change relationship between voltage and current, and the motor currents of different motor models have a certain degree of dispersion, resulting in unclear current characteristics. Even under the premise that the current characteristics are relatively obvious, the existing diagnostic algorithm belongs to the physical analysis method, which analyzes defects by analyzing the current envelope, and belongs to threshold judgment. This judgment will cause two problems. First, it is difficult to directly quantify the threshold setting for different disconnectors. Experts need to intervene to complete physical analysis of the movement process of the disconnector, and then perform threshold correspondence. At the same time, it is difficult to adapt to the differences in thresholds caused by the discreteness of the mechanical state of the switch in the industrial field.

[0010] In order to solve the above technical problems, the present invention provides the following technical solutions: a method for detecting the mechanical state of an isolating switch based on a vibration signal, comprising:

[0011] Arrange a vibration sensor on the disconnector to collect vibration signals;

[0012] Perform short-time Fourier transform on the vibration signal to analyze the time-frequency domain characteristics;

[0013] Apply a Wiener filter to the Fourier transformed signal to remove noise;

[0014] Extracting characteristic data of the filtered vibration signal;

[0015] Normalize the feature data;

[0016] Input the normalized feature data and system status information into the neural network for training;

[0017] Fault diagnosis of vibration signals using trained neural networks;

[0018] The analysis results are received through the data server, a fault diagnosis report is generated, and the mechanical status data is displayed.

[0019] As a preferred solution of the method for detecting the mechanical state of the disconnector based on the vibration signal of the present invention, the short-time Fourier transform is expressed as:

[0020]

[0021] Where ω is the angular frequency, w(t) is the window function, * is the conjugate, e -jωτ is the estimated center frequency.

[0022] As a preferred solution of the method for detecting the mechanical state of the disconnector based on the vibration signal of the present invention, the short-time Fourier transform (STFT) of the signal x (t,ω) is discretized and expressed as,

[0023]

[0024] Among them, x(n+m)w(m) is a short-time series.

[0025] As a preferred solution of the method for detecting the mechanical state of the disconnector based on the vibration signal of the present invention, the Wiener filter is expressed as:

[0026]

[0027] Among them, S s (f) is the power spectrum density of the preprocessed vibration signal s(t), S n(f) is the power spectral density of the noise n(t).

[0028] As a preferred solution of the vibration signal-based disconnector mechanical state detection method of the present invention, the input neural network training includes detecting the vibration signal with selected N vibration sensors;

[0029] Collect measurement signals detected by N vibration sensors at corresponding positions and perform preprocessing;

[0030] Perform feature selection on the preprocessed N vibration sensor signals;

[0031] Normalize the characteristic signal to provide a standard form for the input of the neural network;

[0032] The normalized feature information and known system status information are used as training samples and input into the neural network for training until the preset requirements of the model are met.

[0033] As a preferred solution of the method for detecting the mechanical state of the isolating switch based on vibration signals described in the present invention, the weights of the feature quantities are analyzed after the features are fused, the data collected by the vibration sensor with fluctuations greater than a preset threshold are removed, and the data with overlap higher than the preset threshold are selected for data training and fault diagnosis.

[0034] As a preferred solution of the vibration signal-based disconnector mechanical state detection method of the present invention, wherein: the display of mechanical state data includes transmitting the analyzed and optimized data to the background data server software through the data receiving part;

[0035] The data receiving part receives vibration signal data through the RS232 serial port, communicates with the data acquisition terminal according to the communication protocol, and stores the received data in the form of files;

[0036] The data analysis and diagnosis part performs algorithm analysis on the received vibration signal data, forms a diagnosis report, and gives the fault diagnosis results;

[0037] The data display section can display the vibration signal data curve of each disconnector, support the selection of multiple data of the same disconnector, display multiple data curves in different time periods or states, and be able to view the diagnosis results.

[0038] Another object of the present invention is to provide a mechanical state detection system for an isolating switch based on a vibration signal, which can solve the problems of the existing high-voltage isolating switch fault detection method, such as the inability to effectively avoid environmental interference, high failure rate of single-point sensors, and low diagnostic accuracy, by constructing a mechanical state detection system for an isolating switch based on a vibration signal.

[0039] To solve the above technical problems, the present invention provides the following technical solutions: a vibration signal-based disconnector mechanical state detection system, comprising: a data acquisition module, a time domain analysis module, a noise filtering module, a feature extraction module, a normalization module, a model training module, a fault diagnosis module and a visualization module; the data acquisition module is used to arrange a vibration sensor on the disconnector to collect vibration signals; the time domain analysis module is used to perform short-time Fourier transform on the vibration signal and analyze the time-frequency domain characteristics; the noise filtering module is used to apply a Wiener filter to the Fourier transform signal to remove noise; the feature extraction module is used to extract feature data of the filtered vibration signal; the normalization module is used to perform normalization processing on the feature data; the model training module is used to input the normalized feature data and system state information into a neural network for training; the fault diagnosis module is used to use the trained neural network to perform fault diagnosis of the vibration signal; the visualization module is used to receive the analysis results through a data server, generate a fault diagnosis report, and display the mechanical state data.

[0040] A computer device comprises a memory and a processor, wherein the memory stores a computer program, and when the processor executes the computer program, the steps of the above-mentioned method for detecting the mechanical state of an isolating switch based on a vibration signal are implemented.

[0041] A computer-readable storage medium stores a computer program, which, when executed by a processor, implements the steps of the above-mentioned method for detecting the mechanical state of an isolating switch based on a vibration signal.

[0042] Beneficial effects of the present invention: The method for detecting the mechanical state of an isolating switch based on a vibration signal provided by the present invention firstly uses a short-time Fourier transform to analyze the time-frequency domain characteristics of the vibration signal, thereby ensuring that the time domain information of the signal is effectively captured. Subsequently, the signal is secondary filtered using a Wiener filter to further remove noise and retain valid signals. This two-time filtering method can greatly improve the quality of the signal and reduce the interference of external noise on the detection results. Secondly, by arranging multiple vibration sensors on the isolating switch, vibration signals can be collected from different positions, avoiding local failures or misjudgments that may occur in a single sensor. Multi-point arrangement of sensors can provide more comprehensive monitoring data, thereby more accurately diagnosing the mechanical defects of the isolating switch. Through two-time filtering and multi-point sensor data fusion, the present invention improves the accuracy and robustness of fault diagnosis. BRIEF DESCRIPTION OF THE DRAWINGS

[0043] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the accompanying drawings required for use in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other accompanying drawings can be obtained based on these accompanying drawings without paying creative work.

[0044] Figure 1 An overall flow chart of a method for detecting the mechanical state of an isolating switch based on a vibration signal provided by an embodiment of the present invention.

[0045] Figure 2 A schematic diagram of data fusion of a method for detecting the mechanical state of an isolating switch based on vibration signals provided in one embodiment of the present invention.

[0046] Figure 3 A closing curve diagram of a disconnector mechanical state detection method based on vibration signals provided by an embodiment of the present invention.

[0047] Figure 4 A trip curve diagram of a disconnector mechanical state detection method based on vibration signals provided by an embodiment of the present invention.

[0048] Figure 5 An overall structural diagram of a vibration signal-based isolating switch mechanical state detection system provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0049] In order to make the above-mentioned purposes, features and advantages of the present invention more obvious and easy to understand, the specific implementation methods of the present invention are described in detail below in conjunction with the drawings of the specification. Obviously, the described embodiments are part of the embodiments of the present invention, but not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary persons in the art without creative work should fall within the scope of protection of the present invention.

[0050] In the following description, many specific details are set forth to facilitate a full understanding of the present invention, but the present invention may also be implemented in other ways different from those described herein, and those skilled in the art may make similar generalizations without violating the connotation of the present invention. Therefore, the present invention is not limited to the specific embodiments disclosed below.

[0051] Example 1

[0052] Reference Figure 1-Figure 4 , is an embodiment of the present invention, and provides a method for detecting a mechanical state of an isolating switch based on a vibration signal, comprising:

[0053] Arrange a vibration sensor on the disconnector to collect vibration signals;

[0054] Perform short-time Fourier transform on the vibration signal to analyze the time-frequency domain characteristics;

[0055] Apply a Wiener filter to the Fourier transformed signal to remove noise;

[0056] Extracting characteristic data of the filtered vibration signal;

[0057] Normalize the feature data;

[0058] Input the normalized feature data and system status information into the neural network for training;

[0059] Fault diagnosis of vibration signals using trained neural networks;

[0060] The analysis results are received through the data server, a fault diagnosis report is generated, and the mechanical status data is displayed.

[0061] Vibration sensors are also called acceleration sensors. There are many types of vibration sensors. Generally, they use the changes in the piezoelectric effect of piezoelectric ceramics to sense the external inertial force and convert the force signal into an electrical signal. The vibration signal is a physical function that carries the information of the device or structure, while the vibration information is a characteristic quantity that reflects the operating status and mechanism characteristics of the device or structure. Vibration measurement is to detect, analyze, and extract characteristic quantities of the vibration signals exhibited by mechanical equipment to determine the degree of deterioration of the machine itself and predict its life.

[0062] The core component of the high-voltage disconnector is the mechanical structure, which is used to transmit force and motion. If any moving part or related parts fail, the stability of the mechanical motion will inevitably be destroyed. With the participation of the transmitted force, the non-stationary phenomenon of this force and motion can be characterized by vibration. Therefore, installing a vibration sensor on the high-voltage disconnector can effectively monitor the vibration signal generated by the mechanical motion. By analyzing the vibration signal and selecting appropriate characteristic parameters, the non-stationary signal under the fault state can be effectively identified, and then the fault state of the disconnector can be determined.

[0063] The mechanical vibration signals of the GIS disconnector under different mechanical states are studied, and the mechanical vibration characteristics that can characterize the mechanical state of the GIS disconnector are extracted. The mechanical vibration model of the GIS disconnector is established, and the mechanical vibration signals of the disconnector under different mechanical states (stuck, misaligned, etc.) are analyzed and feature extracted, and the mapping relationship between the mechanical vibration characteristics and the mechanical state is constructed. Under different functional states (different contact spring elastic forces), the vibration characteristics of the GIS disconnector during steady-state flow are studied, and the GIS disconnector electromagnetic vibration model is established. The vibration signals under different functional states (different fatigue levels of the contact spring) are analyzed and feature extracted, and the mapping relationship between the electromagnetic vibration characteristics and the functional state is constructed to extract the vibration characteristics that can characterize the functional state of the GIS disconnector.

[0064] The present invention arranges vibration sensors at multiple points on the isolating switch, performs noise reduction and feature extraction on the vibration signals collected by the vibration sensors, and then performs feature fusion of the vibration signals; further performs feature quantity weight analysis and optimization, data training and fault diagnosis.

[0065] The specific process of noise reduction processing on the vibration signal collected by the vibration sensor is as follows:

[0066] The present invention adopts short-time Fourier transform to pre-process the collected non-stationary vibration signal, divides the collected vibration signal into several small segments, performs Fourier transform on each segment, and then displays the time-frequency characteristics of the vibration signal by depicting a spectrum of frequency changes over time.

[0067] The pre-processed vibration signal is then filtered twice using the Wiener filter, and the processed vibration signal is feature extracted. Vibration signals are generally non-stationary signals. By first performing a short-time Fourier transform on the non-stationary signal, the vibration signal can be initially processed into a stationary signal, and then the stationary signal is subjected to Wiener filtering (Wiener filtering can only reduce noise on stationary signals), which is equivalent to performing two noise reductions to improve the quality of the signal.

[0068] The short-time Fourier transform is as follows:

[0069] Short-time Fourier transform is generally used to reduce noise of non-steady-state signals. Assuming that the signal collected by the first vibration sensor is x(t), which is a non-steady-state signal, and a window function w(t) with a very narrow time width is given, the signal will slide along the time axis. At this time, the Fourier transform formula of the signal x(t) collected by the first vibration sensor can be expressed as:

[0070]

[0071] Where ω is the angular frequency, w(t) is the window function, * is the conjugate, e-jωτ is the estimated center frequency;

[0072] Because of the existence of the window function w(t), the window function divides the frequency into several segments. Therefore, the short-time Fourier transform has local characteristics, so it is both a time function and a frequency function. By giving a time t, the short-time Fourier transform (STFT) STFT of the signal x (t,ω) can be regarded as the spectrum at that moment, that is, the local spectrum.

[0073] For example, if you need to obtain the local spectrum of the vibration signal collected by the first vibration sensor in 0-0.1s, you can assign t a value of 0.1, perform Fourier transform on the signal during this period, and obtain the local spectrum of the vibration signal collected by this sensor.

[0074] Since continuous signals cannot be accurately represented in MATLAB, the short-time Fourier transform (STFT) of the signal is required in the present invention. x (t,ω) is discretized, and the discrete signal of the continuous time domain signal x(t) is assumed to be x(n), and the window function is w(m);

[0075] The window function moves on the time axis, and the window function length is N, then the short-time Fourier transform STFT x The discrete form of (t,ω) is:

[0076]

[0077] Among them, x(n+m)w(m) is a short-time series.

[0078] The shorter the signal time is, the lower the frequency resolution of the signal is. By continuously increasing the time segments of the signal, the time of each signal segment becomes shorter and the frequency resolution of the time becomes higher.

[0079] The preprocessed vibration signal is then filtered twice using the Wiener filter (because the Wiener filter can only process stationary signals). Assume that the preprocessed vibration signal is s(t), the noise is n(t), the noise n(t) is white noise, and is uncorrelated with the preprocessed vibration signal s(t). Assume that the filter transfer function is H(f), and the frequency spectrum of the output vibration signal is

[0080] D(f) is the frequency spectrum of the preprocessed vibration signal.

[0081] The transfer function of the filter can be obtained as

[0082]

[0083] Among them, S s(f) is the power spectrum density of the preprocessed vibration signal s(t), S n (f) is the power spectral density of the noise n(t).

[0084] After the vibration signals collected by different vibration sensors are subjected to noise reduction processing and feature extraction, feature fusion is performed on the vibration signals. The present invention fuses the data collected by different vibration sensors using a multimodal data fusion method based on a neural network. The specific fusion method is shown in Figure 2.

[0085] Since multiple vibration sensors are set up, the neural network not only diagnoses the data, but also fuses the data collected by different vibration sensors and subjected to two noise reduction processes.

[0086] Before network training, the data must first be normalized to ensure that the network layer output is not too small. Here, the initial value pi of the center vector is determined by the training sample, where pi = (pi1, pi2, pi3, pi4, pi5), i = 1, 2, ... 10. If it belongs to the training sample set of the first type of data change, expressed as (S1, S2, ..., Sm), the initial value of each element in the center vector p1 is the average value of each element in these sample input directions. Then the gradient descent method is used to adjust the center vector, width, coefficient and weight of the last layer respectively. During the training process, a larger learning rate is first selected. If the training error of the latter step is greater than the training error of the previous step, the learning rate is reduced to make the network sensitive. Due to the addition of prior information of the center vector of each category, the adjustment of the network weights is limited to a certain range to avoid falling into the local minimum point. The present invention takes into account the lossy fusion and lossless fusion problems of data fusion, and the main description steps of the algorithm are as follows:

[0087] (1) Detecting vibration signals using the selected N vibration sensors;

[0088] (2) collecting measurement signals detected by N vibration sensors at corresponding positions and performing preprocessing;

[0089] (3) performing feature selection on the N vibration sensor signals after preprocessing;

[0090] (4) Normalize the characteristic signal to provide a standard form for the input of the neural network;

[0091] (5) The normalized feature information and the known system status information are used as training samples and sent to the neural network for training until the requirements are met. The trained network is used as the known network. As long as the normalized multi-sensor feature information is sent as input to the network, the network output is the status of the system under test.

[0092] By adopting the multimodal data fusion method of neural network, the independence of data is retained and the information loss in the fusion process is reduced.

[0093] Furthermore, after the features are fused, the weights of the feature quantities are analyzed and optimized, the data with large fluctuations collected by the vibration sensor are removed, and the data with good overlap are selected for data training and fault diagnosis.

[0094] The analyzed and optimized data is transmitted to the background data server software. The background data server software consists of three parts: data reception, data analysis and diagnosis, and data display. The data reception part receives the collected vibration signal data and stores it in the form of a file; the data analysis and diagnosis part analyzes the collected vibration signal data for algorithm analysis, forms a report in the form of a WORD or PDF file, and gives the judgment result; the data display part can display the data curve of each disconnector, and multiple data curves can be displayed by selecting multiple data of the same disconnector; the diagnostic results of the disconnector can be viewed.

[0095] The data server software receives data via RS232 serial port, communicates with the data acquisition terminal according to the communication protocol, and stores the received data in the form of files.

[0096] When new data is received, the configuration file is read to read the disconnector model, parameters and standard curve, and the mechanical defect diagnosis results are obtained through segmented correlation analysis of the received data. Figure 3 and Figure 4 shown.

[0097] Example 2

[0098] Reference Figure 5 , is an embodiment of the present invention, and provides a mechanical state detection system of an isolating switch based on a vibration signal, comprising:

[0099] Data acquisition module 100, time domain analysis module 200, noise filtering module 300, feature extraction module 400, normalization module 500, model training module 600, fault diagnosis module 700 and visualization module 800;

[0100] The data acquisition module 100 is used to arrange a vibration sensor on the disconnector to collect vibration signals;

[0101] The time domain analysis module 200 is used to perform short-time Fourier transform on the vibration signal to analyze the time-frequency domain characteristics;

[0102] The noise filtering module 300 is used to apply a Wiener filter to the Fourier transform signal to remove noise;

[0103] The feature extraction module 400 is used to extract feature data of the filtered vibration signal;

[0104] The normalization module 500 is used to perform normalization processing on the feature data;

[0105] The model training module 600 is used to input the normalized feature data and system status information into the neural network for training;

[0106] The fault diagnosis module 700 is used to perform fault diagnosis of vibration signals using a trained neural network;

[0107] The visualization module 800 is used to receive analysis results through a data server, generate a fault diagnosis report, and display mechanical status data.

[0108] Example 3

[0109] An embodiment of the present invention is different from the first two embodiments in that:

[0110] If the functions are implemented in the form of software functional units and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, or the part that contributes to the prior art or the part of the technical solution, can be embodied in the form of a software product. The computer software product is stored in a storage medium, including several instructions for a computer device (which can be a personal computer, a server, or a network device, etc.) to perform all or part of the steps of the methods described in each embodiment of the present invention. The aforementioned storage medium includes: U disk, mobile hard disk, read-only memory (ROM, Read-Only Memory), random access memory (RAM, Random Access Memory), disk or optical disk, etc., which can store program codes.

[0111] The logic and / or steps represented in the flowchart or otherwise described herein, for example, can be considered as an ordered list of executable instructions for implementing logical functions, and can be embodied in any computer-readable medium for use by an instruction execution system, device or apparatus (such as a computer-based system, a system including a processor, or other system that can fetch instructions from an instruction execution system, device or apparatus and execute instructions), or in conjunction with such instruction execution systems, devices or apparatuses. For the purposes of this specification, "computer-readable medium" can be any device that can contain, store, communicate, propagate or transmit a program for use by an instruction execution system, device or apparatus, or in conjunction with such instruction execution systems, devices or apparatuses.

[0112] More specific examples of computer-readable media (a non-exhaustive list) include the following: an electrical connection with one or more wires (electronic device), a portable computer disk case (magnetic device), a random access memory (RAM), a read-only memory (ROM), an erasable and programmable read-only memory (EPROM or flash memory), an optical fiber device, and a portable compact disk read-only memory (CDROM). In addition, the computer-readable medium may even be a paper or other suitable medium on which the program is printed, since the program may be obtained electronically, for example, by optically scanning the paper or other medium, followed by editing, deciphering or, if necessary, processing in another suitable manner, and then stored in a computer memory.

[0113] It should be understood that the various parts of the present invention can be implemented by hardware, software, firmware or a combination thereof. In the above-mentioned embodiments, a plurality of steps or methods can be implemented by software or firmware stored in a memory and executed by a suitable instruction execution system. For example, if implemented by hardware, as in another embodiment, it can be implemented by any one of the following technologies known in the art or their combination: a discrete logic circuit having a logic gate circuit for implementing a logic function for a data signal, a dedicated integrated circuit having a suitable combination of logic gate circuits, a programmable gate array (PGA), a field programmable gate array (FPGA), etc.

[0114] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention rather than to limit it. Although the present invention has been described in detail with reference to the preferred embodiments, those skilled in the art should understand that the technical solutions of the present invention may be modified or replaced by equivalents without departing from the spirit and scope of the technical solutions of the present invention, which should all be included in the scope of the claims of the present invention.

Claims

1. A method for detecting the mechanical state of an isolating switch based on a vibration signal, characterized in that: include: Arrange a vibration sensor on the disconnector to collect vibration signals; Perform short-time Fourier transform on the vibration signal to analyze the time-frequency domain characteristics; Apply a Wiener filter to the Fourier transformed signal to remove noise; Extracting characteristic data of the filtered vibration signal; Normalize the feature data; Input the normalized feature data and system status information into the neural network for training; Fault diagnosis of vibration signals using trained neural networks; The analysis results are received through the data server, a fault diagnosis report is generated, and the mechanical status data is displayed.

2. The method for detecting the mechanical state of an isolating switch based on a vibration signal according to claim 1, characterized in that: The short-time Fourier transform is expressed as, Where ω is the angular frequency, w(t) is the window function, * is the conjugate, e -jωτ is the estimated center frequency.

3. The method for detecting the mechanical state of an isolating switch based on a vibration signal according to claim 2, characterized in that: Short-time Fourier transform STFT of the signal x (t,ω) is discretized and expressed as, Among them, x(n+m)w(m) is a short-time series.

4. The method for detecting the mechanical state of an isolating switch based on a vibration signal according to claim 3, characterized in that: The Wiener filter is expressed as, Among them, S s (f) is the power spectrum density of the preprocessed vibration signal s(t), S n (f) is the power spectral density of the noise n(t).

5. The method for detecting the mechanical state of an isolating switch based on a vibration signal according to claim 4, characterized in that: The input neural network is trained by detecting vibration signals using selected N vibration sensors; Collect measurement signals detected by N vibration sensors at corresponding positions and perform preprocessing; Perform feature selection on the preprocessed N vibration sensor signals; Normalize the characteristic signal to provide a standard form for the input of the neural network; The normalized feature information and known system status information are used as training samples and input into the neural network for training until the preset requirements of the model are met.

6. The method for detecting the mechanical state of an isolating switch based on a vibration signal according to claim 5, characterized in that: After the features are fused, the weights of the feature quantities are analyzed, the data collected by the vibration sensor with fluctuations greater than the preset threshold are removed, and the data with overlap higher than the preset threshold are selected for data training and fault diagnosis.

7. The method for detecting the mechanical state of an isolating switch based on a vibration signal according to claim 6, characterized in that: The display of the machine status data includes transmitting the analyzed and optimized data to the backend data server software through the data receiving part; The data receiving part receives vibration signal data through the RS232 serial port, communicates with the data acquisition terminal according to the communication protocol, and stores the received data in the form of files; The data analysis and diagnosis part performs algorithm analysis on the received vibration signal data, forms a diagnosis report, and gives the fault diagnosis results; The data display section can display the vibration signal data curve of each disconnector, support the selection of multiple data of the same disconnector, display multiple data curves in different time periods or states, and be able to view the diagnosis results.

8. A system using the isolation switch mechanical state detection method based on vibration signals as claimed in any one of claims 1 to 7, characterized in that: include: A data acquisition module (100), a time domain analysis module (200), a noise filtering module (300), a feature extraction module (400), a normalization module (500), a model training module (600), a fault diagnosis module (700), and a visualization module (800); The data acquisition module (100) is used to arrange a vibration sensor on the isolating switch to collect vibration signals; The time domain analysis module (200) is used to perform short-time Fourier transform on the vibration signal to analyze the time-frequency domain characteristics; The noise filtering module (300) is used to apply a Wiener filter to the Fourier transform signal to remove noise; The feature extraction module (400) is used to extract feature data of the filtered vibration signal; The normalization module (500) is used to perform normalization processing on the feature data; The model training module (600) is used to input the normalized feature data and system status information into the neural network for training; The fault diagnosis module (700) is used to perform fault diagnosis of vibration signals using a trained neural network; The visualization module (800) is used to receive analysis results through a data server, generate a fault diagnosis report, and display machine status data.

9. A computer device comprising a memory and a processor, wherein the memory stores a computer program, wherein: When the processor executes the computer program, the steps of the method for detecting the mechanical state of an isolating switch based on a vibration signal according to any one of claims 1 to 7 are implemented.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the steps of the method for detecting the mechanical state of an isolating switch based on a vibration signal according to any one of claims 1 to 7 are implemented.

Images