High-voltage cable dielectric loss monitoring method, device, terminal equipment and storage medium
By constructing a dielectric loss prediction model and using the convolutional attention mechanism to process multi-dimensional dielectric loss data, the problem of failing to fully consider the influencing factors in existing technologies is solved, and the accuracy and real-time performance of high-voltage cable dielectric loss monitoring are achieved.
Patent Information
- Application Number
- CN202510143217.6
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Filing Date
- 2025-02-10
- Publication Date
- 2025-09-30
- Estimated Expiration
- 2045-02-10
AI Technical Summary
Existing high-voltage cable dielectric loss monitoring methods fail to fully consider the influence of factors such as temperature, electric field interference and electromagnetic interference, resulting in inaccurate monitoring results.
By constructing a dielectric loss prediction model and using the convolutional attention mechanism to process multi-dimensional dielectric loss data, temperature, electric field interference and electromagnetic interference data, iterative training is performed to generate a dielectric loss prediction model and output the dielectric loss value at the next moment.
It improves the accuracy and comprehensiveness of high-voltage cable dielectric loss monitoring, can predict dielectric loss changes in real time, and reduces the risk of power accidents caused by insulation defects.
Smart Images
Figure CN119667301B_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to the technical field of cable insulation performance monitoring, and in particular to a method, device, terminal equipment and storage medium for monitoring dielectric loss of a high-voltage cable. Background Art
[0002] With the continuous advancement of power grids, the length of high-voltage cables is increasing, making their safe and stable operation increasingly important. Monitoring high-voltage cables can reduce the risk of power accidents caused by insulation defects and ensure the safety and stability of the power grid. Existing monitoring methods for high-voltage cables primarily assess the overall aging of high-voltage cables using dielectric loss. This involves the internal heating of the dielectric due to dielectric conductivity and polarization hysteresis under the influence of an applied electric field. This dielectric loss not only dissipates electrical energy but also causes heating and aging of the insulation material, potentially leading to insulation failure.
[0003] Current methods for monitoring dielectric loss in high-voltage cables can be categorized into two main types: absolute and relative measurement. Absolute measurement methods include traditional methods such as the bridge method and the resonance method, while relative measurement methods include the zero-crossing comparison method, the sine fitting method, and the harmonic analysis method. These methods are all digital and fail to consider the impact of related factors on dielectric loss during measurement. Summary of the Invention
[0004] Embodiments of the present invention provide a high-voltage cable dielectric loss monitoring method, apparatus, terminal equipment, and storage medium, which can comprehensively consider the influence of factors related to high-voltage cable dielectric loss, improve the accuracy and comprehensiveness of high-voltage cable dielectric loss value prediction, and thereby improve the accuracy of high-voltage cable dielectric loss monitoring.
[0005] An embodiment of the present invention provides a method for monitoring dielectric loss of a high-voltage cable, comprising:
[0006] Acquire multidimensional dielectric loss data to be processed for a preset period of time of a target high-voltage cable; wherein the multidimensional dielectric loss data to be processed includes: dielectric loss data to be processed and associated data of the dielectric loss data to be processed; the associated data includes temperature, electric field interference, and electromagnetic interference;
[0007] Inputting the multidimensional dielectric loss data to be processed for the preset period into a dielectric loss prediction model, so that the dielectric loss prediction model outputs a dielectric loss value of the multidimensional dielectric loss data to be processed for the preset period at the next moment;
[0008] The dielectric loss of the target high-voltage cable is monitored according to the dielectric loss value at the next moment.
[0009] Furthermore, the construction of the dielectric loss prediction model includes:
[0010] Obtain a multidimensional dielectric loss time series training set; wherein the multidimensional dielectric loss time series training set is composed of multidimensional dielectric loss data samples at a plurality of consecutive moments, each multidimensional dielectric loss data sample includes a dielectric loss data sample, a temperature sample, an electric field interference sample, and an electromagnetic interference sample, and each multidimensional dielectric loss data sample is marked with a corresponding dielectric loss true value;
[0011] An initial dielectric loss prediction model is constructed, and the initial dielectric loss prediction model is iteratively trained using the multi-dimensional dielectric loss time series training set until the initial dielectric loss prediction model converges, thereby generating a dielectric loss prediction model.
[0012] Furthermore, the initial dielectric loss prediction model is iteratively trained using the multi-dimensional dielectric loss time series training set, including:
[0013] Iteratively training the initial dielectric loss prediction model using a multidimensional dielectric loss data sample sequence of a preset time period as input to the initial dielectric loss prediction model and using the actual dielectric loss value at the next moment of the preset time period as output of the initial dielectric loss prediction model;
[0014] In each iterative training process, the initial dielectric loss prediction model uses a convolutional attention mechanism to process dielectric loss data samples, temperature samples, electric field interference samples, and electromagnetic interference samples, and outputs a dielectric loss prediction value at the next moment in a preset time period;
[0015] Determine the dielectric loss value of the current iteration according to the dielectric loss prediction value at the next moment in the preset period and the dielectric loss actual value at the next moment in the preset period;
[0016] The weight of the initial dielectric loss prediction model is adjusted according to the dielectric loss value of the current iteration.
[0017] Furthermore, the multi-dimensional dielectric loss data sample also includes a true value of frequency;
[0018] The multi-dimensional dielectric loss time series training set is used to iteratively train the initial dielectric loss prediction model, including:
[0019] Iteratively training the initial dielectric loss prediction model using a multidimensional dielectric loss data sample sequence of a preset time period as input to the initial dielectric loss prediction model and using the actual dielectric loss value at the next moment of the preset time period as output of the initial dielectric loss prediction model;
[0020] In each iterative training process, the initial dielectric loss prediction model uses a convolutional attention mechanism to process dielectric loss data samples, temperature samples, electric field interference samples, and electromagnetic interference samples, and outputs a dielectric loss prediction value at the next moment in a preset period and a frequency prediction value at the next moment in the preset period;
[0021] Determine the dielectric loss value of the current iteration according to the dielectric loss prediction value at the next moment in the preset period and the dielectric loss actual value at the next moment in the preset period;
[0022] Adjust the weight of the initial dielectric loss prediction model according to the dielectric loss value of the current iteration;
[0023] Determine the frequency loss value of the current iteration according to the frequency prediction value at the next moment of the preset period and the actual frequency value at the next moment of the preset period;
[0024] If the frequency loss value of the current iteration is less than a preset frequency loss threshold, the multidimensional dielectric loss data sample at the next moment of the preset period and the dielectric loss prediction value at the next moment of the preset period are added to the multidimensional dielectric loss time series training set. Furthermore, the preset frequency loss threshold is determined based on a preset threshold adjustment curve and the current number of iterations.
[0025] Furthermore, after obtaining the multi-dimensional dielectric loss time series training set, the method further includes:
[0026] Wavelet denoising is performed on the multidimensional dielectric loss time series training set to obtain the denoised multidimensional dielectric loss time series training set.
[0027] Based on the above method embodiment, the present invention provides a corresponding device embodiment;
[0028] An embodiment of the present invention provides a high-voltage cable dielectric loss monitoring device, comprising: a data acquisition module, a dielectric loss prediction module, and a monitoring module;
[0029] The data acquisition module is configured to acquire the multi-dimensional dielectric loss data to be processed for a preset period of time of the target high-voltage cable; wherein the multi-dimensional dielectric loss data to be processed includes: the dielectric loss data to be processed and associated data of the dielectric loss data to be processed; the associated data includes temperature, electric field interference, and electromagnetic interference;
[0030] The dielectric loss prediction module is configured to input the multidimensional dielectric loss data to be processed in the preset time period into the dielectric loss prediction model, so that the dielectric loss prediction model outputs the dielectric loss value of the multidimensional dielectric loss data to be processed in the preset time period at the next moment;
[0031] The monitoring module is used to monitor the dielectric loss of the target high-voltage cable according to the dielectric loss value at the next moment.
[0032] Furthermore, it also includes a dielectric loss prediction model construction module;
[0033] The dielectric loss prediction model construction module is used to obtain a multidimensional dielectric loss time series training set; wherein the multidimensional dielectric loss time series training set is composed of multidimensional dielectric loss data samples at a plurality of consecutive moments, each multidimensional dielectric loss data sample includes a dielectric loss data sample, a temperature sample, an electric field interference sample, and an electromagnetic interference sample, and each multidimensional dielectric loss data sample is marked with a corresponding dielectric loss true value;
[0034] An initial dielectric loss prediction model is constructed, and the initial dielectric loss prediction model is iteratively trained using the multi-dimensional dielectric loss time series training set until the initial dielectric loss prediction model converges, thereby generating a dielectric loss prediction model.
[0035] Another embodiment of the present invention provides a terminal device, including a processor, a memory, and a computer program stored in the memory and configured to be executed by the processor. When the processor executes the computer program, it implements the high-voltage cable dielectric loss monitoring method described in the above-mentioned embodiment of the invention.
[0036] Another embodiment of the present invention provides a storage medium, which includes a stored computer program, wherein when the computer program is running, the device where the storage medium is located is controlled to execute the high-voltage cable dielectric loss monitoring method described in the above-mentioned embodiment of the invention.
[0037] The following beneficial effects are achieved by implementing the present invention:
[0038] The present invention provides a method, apparatus, terminal device, and storage medium for monitoring dielectric loss of a high-voltage cable. The monitoring method obtains unprocessed dielectric loss data and associated data of the unprocessed dielectric loss data for a preset time period of a target high-voltage cable, wherein the associated data includes temperature, electric field interference, and electromagnetic interference; combines the data into unprocessed multidimensional dielectric loss data, which is then input into a dielectric loss prediction model so that the dielectric loss prediction model outputs a dielectric loss value at the next moment of the unprocessed multidimensional dielectric loss data for the preset time period, and performs dielectric loss monitoring on the target high-voltage cable based on the dielectric loss value at the next moment. By obtaining the unprocessed multidimensional dielectric loss data including the unprocessed dielectric loss data and associated data of the unprocessed dielectric loss data, various associated factors affecting dielectric loss can be fully considered when predicting the dielectric loss value, making the prediction result more comprehensive and accurate, thereby improving the accuracy of dielectric loss monitoring of the target high-voltage cable. BRIEF DESCRIPTION OF THE DRAWINGS
[0039] Figure 1The present invention is a flowchart of a method for monitoring dielectric loss of a high-voltage cable provided by an embodiment of the present invention.
[0040] Figure 2 This is a training flow chart of an initial dielectric loss prediction model provided by an embodiment of the present invention.
[0041] Figure 3 The figure is a structural diagram of a high-voltage cable dielectric loss monitoring device provided by one embodiment of the present invention. DETAILED DESCRIPTION
[0042] The following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. All other embodiments obtained by ordinary technicians in this field based on the embodiments of the present invention without making any creative efforts shall fall within the scope of protection of the present invention.
[0043] To make the objectives, technical solutions, and advantages of this application more clear, the technical solutions in this application will be clearly and completely described below in conjunction with the accompanying drawings in the embodiments of this application. Obviously, the described embodiments are only part of the embodiments of this application, not all of them. Based on the embodiments in this application, all other embodiments obtained by ordinary technicians in this field without making any creative efforts are within the scope of protection of this application.
[0044] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by those skilled in the art to which this application belongs; the terms used herein are only for the purpose of describing specific embodiments and are not intended to limit this application; the terms "including" and "having" and any variations thereof in the specification and claims of this application and the above-mentioned figure descriptions are intended to cover non-exclusive inclusions.
[0045] In the description of the embodiments of this application, the technical terms "first" and "second" are used only to distinguish different objects and should not be understood to indicate or imply relative importance or implicitly specify the quantity, specific order, or primary and secondary relationship of the indicated technical features. In the description of the embodiments of this application, the meaning of "plurality" is more than two, unless otherwise clearly and specifically defined.
[0046] References herein to "embodiments" mean that a particular feature, structure, or characteristic described in connection with the embodiments may be included in at least one embodiment of the present application. The appearance of this phrase in various places in the specification does not necessarily refer to the same embodiment, nor does it constitute an independent or alternative embodiment that is mutually exclusive of other embodiments. It is understood, both explicitly and implicitly, by those skilled in the art that the embodiments described herein may be combined with other embodiments.
[0047] In the description of the embodiments of this application, the term "and / or" is simply a description of the association relationship between associated objects, indicating that three relationships can exist. For example, A and / or B can represent the following three situations: A exists alone, A and B exist simultaneously, and B exists alone. In addition, the character " / " in this document generally indicates that the associated objects are in an "or" relationship.
[0048] In the description of the embodiments of the present application, the term "multiple" refers to more than two (including two). Similarly, "multiple groups" refers to more than two groups (including two groups), and "multiple pieces" refers to more than two pieces (including two pieces).
[0049] In the description of the embodiments of the present application, unless otherwise expressly specified or limited, technical terms such as "installed," "connected," "connected," and "fixed" should be understood in a broad sense. For example, they can refer to fixed connections, detachable connections, or integration; mechanical connections or electrical connections; direct connections or indirect connections through an intermediate medium; internal connections between two components or interactions between two components. Those skilled in the art can understand the specific meanings of the above terms in the embodiments of the present application based on specific circumstances.
[0050] like Figure 1 FIG. 1 is a high-voltage cable dielectric loss monitoring method provided by an embodiment of the present invention, comprising:
[0051] Step S1: obtaining multi-dimensional dielectric loss data to be processed for a preset period of time of a target high-voltage cable; wherein the multi-dimensional dielectric loss data to be processed includes: dielectric loss data to be processed and associated data of the dielectric loss data to be processed; the associated data includes temperature, electric field interference, and electromagnetic interference;
[0052] Step S2: inputting the multidimensional dielectric loss data to be processed for the preset period into a dielectric loss prediction model, so that the dielectric loss prediction model outputs a dielectric loss value at the next moment of the multidimensional dielectric loss data to be processed for the preset period;
[0053] Step S3: monitoring the dielectric loss of the target high-voltage cable according to the dielectric loss value at the next moment.
[0054] With respect to step S1, multidimensional dielectric loss data to be processed for a target high-voltage cable to be monitored during a preset time period is obtained. This multidimensional dielectric loss data must include the dielectric loss data to be processed for the target high-voltage cable at each moment in the preset time period, as well as associated data such as the temperature, electric field interference, and electromagnetic interference of the target high-voltage cable. The preset time period is a historical period of fixed length. The end time of the preset time period is the current monitoring time of the target high-voltage cable, and the start time of the preset time period is determined based on the current monitoring time of the target high-voltage cable and the fixed time length.
[0055] For step S2, the multidimensional dielectric loss data to be processed for a preset time period is input into the dielectric loss prediction model, and the dielectric loss prediction model outputs the dielectric loss value of the multidimensional dielectric loss data to be processed for the preset time period at the next moment, that is, the dielectric loss value at the next moment after the current monitoring moment.
[0056] In a preferred embodiment, the construction of the dielectric loss prediction model includes: obtaining a multidimensional dielectric loss time series training set; wherein the multidimensional dielectric loss time series training set is composed of multidimensional dielectric loss data samples at several consecutive moments, each multidimensional dielectric loss data sample includes a dielectric loss data sample, a temperature sample, an electric field interference sample, and an electromagnetic interference sample, and each multidimensional dielectric loss data sample is marked with a corresponding dielectric loss true value; constructing an initial dielectric loss prediction model, and iteratively training the initial dielectric loss prediction model with the multidimensional dielectric loss time series training set until the initial dielectric loss prediction model converges, thereby generating a dielectric loss prediction model.
[0057] Specifically, the construction of the dielectric loss prediction model mainly includes a training data acquisition step and a model training step. When acquiring the training data, in order to minimize the interference of irrelevant external electromagnetic wave signals, the multi-dimensional dielectric loss data samples of each continuous moment of the present invention are all obtained through simulation experiments. First, a laboratory cable dielectric loss data acquisition platform is constructed in an anechoic chamber, and a cable layout and model transmission path similar to the actual high-voltage cable field are set in the laboratory, and a signal transmitter and a spectrum analyzer are used to simulate industrial interference signals. Secondly, a voltage sensor and a current sensor are used to collect high-voltage cable analog data respectively, and the collected high-voltage cable analog signal is converted into a digital signal with a 50H industrial frequency as the sampling frequency. At the same time, temperature monitoring equipment, electric field probes and magnetic field probes are used to detect parameters such as real-time temperature, electric field interference, and magnetic field interference respectively, as temperature samples, electric field interference samples and electromagnetic interference samples corresponding to the multi-dimensional dielectric loss data samples at each moment. Thirdly, based on high-voltage cable data, a comprehensive approach to dielectric loss detection, including harmonic analysis, bipolar zero-crossing comparison, and high-order sinusoidal fitting, is employed to obtain dielectric loss data at multiple consecutive moments, which are then used as dielectric loss data samples corresponding to the multidimensional dielectric loss data samples at multiple consecutive moments. It should be noted that the harmonic analysis method uses fast Fourier transform harmonic analysis of current and voltage signals, extracts the signal phase angle, and obtains dielectric loss data; the bipolar zero-crossing comparison method simultaneously compares the zero-crossing time difference between the positive and negative phases of voltage and current to obtain the corresponding dielectric loss data; and the high-order sinusoidal fitting method uses the fundamental frequency, the amplitude, and the phase angle of each harmonic as variables to approximate the original signal, combining it with Hamming window interpolation to obtain the corresponding dielectric loss data. For each dielectric loss data sample, a multidimensional dielectric loss data sample is constructed by combining the temperature sample, the electric field interference sample, the electromagnetic interference sample, and the actual dielectric loss value. The harmonic analysis method is then used to obtain the actual dielectric loss value corresponding to the multidimensional dielectric loss data sample. Finally, the sampling rate is adjusted in real time by using the true value of the adopted frequency and a dynamic algorithm to obtain multidimensional dielectric loss data samples at multiple consecutive moments, forming a multidimensional dielectric loss time series training set.
[0058] In a preferred embodiment, after obtaining the multi-dimensional dielectric loss time series training set, the method further includes: performing wavelet denoising on the multi-dimensional dielectric loss time series training set to obtain a denoised multi-dimensional dielectric loss time series training set.
[0059] Specifically, after obtaining the multidimensional dielectric loss time series training set, the multidimensional dielectric loss data samples at each moment containing the noisy signal are placed in a two-dimensional space using wavelet transform. By utilizing the completely different characteristics of the signal and noise, the denoising effect of each sample is achieved through wavelet decomposition, denoising and signal reconstruction of the noisy signal, and the denoised multidimensional dielectric loss time series training set is obtained.
[0060] Preferably, the denoised multi-dimensional dielectric loss time series training set is divided into a validation set and a test set according to a certain ratio.
[0061] like Figure 2 As shown, an initial dielectric loss prediction model is constructed. The initial dielectric loss prediction model is a ConvTrans model. It should be noted that the ConvTrans model is a model that is improved and optimized for time series prediction tasks based on the Transformer model. The core structure of the ConvTrans model consists of two parts: Convolutional Self-Attention (convolutional attention mechanism) and LogSparse Transformer (log sparse Transformer). The ConvTrans model uses the convolutional attention mechanism and log sparse Transformer to achieve less memory for modeling of finer granularity long sequences. The convolutional attention mechanism generates query value Q, key K and value V through causal convolution, as shown in the following formula:
[0062]
[0063] Among them, the query value Q and key K are generated through causal convolution as follows:
[0064] Q=Conv(X,W q ,padding)
[0065] K=Conv(X,W k ,padding)
[0066] Where X is the input sequence; W q and W k is the convolution kernel; padding ensures causality.
[0067] The initial dielectric loss prediction model is iteratively trained using the multi-dimensional dielectric loss time series training set.
[0068] In a preferred embodiment, the initial dielectric loss prediction model is iteratively trained using the multidimensional dielectric loss time series training set, including: using a multidimensional dielectric loss data sample sequence of a preset time period as the input of the initial dielectric loss prediction model, and using the true value of the dielectric loss at the next moment of the preset time period as the output of the initial dielectric loss prediction model, and iteratively training the initial dielectric loss prediction model; wherein, in each iterative training process, the initial dielectric loss prediction model uses a convolutional attention mechanism to process dielectric loss data samples, temperature samples, electric field interference samples, and electromagnetic interference samples, and outputs a dielectric loss prediction value at the next moment of the preset time period; determines the dielectric loss loss value of the current iteration based on the dielectric loss prediction value at the next moment of the preset time period and the true value of the dielectric loss at the next moment of the preset time period; and adjusts the weight of the initial dielectric loss prediction model based on the dielectric loss loss value of the current iteration.
[0069] In a preferred embodiment, the multidimensional dielectric loss data sample also includes a true value of the frequency; the multidimensional dielectric loss time series training set iteratively trains the initial dielectric loss prediction model, including: using the multidimensional dielectric loss data sample sequence of a preset time period as the input of the initial dielectric loss prediction model, and using the true value of the dielectric loss at the next moment of the preset time period as the output of the initial dielectric loss prediction model, and iteratively training the initial dielectric loss prediction model; wherein, in each iterative training process, the initial dielectric loss prediction model uses a convolutional attention mechanism to process the dielectric loss data sample, temperature sample, electric field interference sample, and electromagnetic interference sample, and outputs the preset time series training set. The dielectric loss prediction value of the next moment of the preset period and the frequency prediction value of the next moment of the preset period are determined; the dielectric loss loss value of the current iteration is determined according to the dielectric loss prediction value of the next moment of the preset period and the actual value of the dielectric loss of the next moment of the preset period; the weight of the initial dielectric loss prediction model is adjusted according to the dielectric loss loss value of the current iteration; the frequency loss value of the current iteration is determined according to the frequency prediction value of the next moment of the preset period and the actual value of the frequency of the next moment of the preset period; if the frequency loss value of the current iteration is less than the preset frequency loss threshold, the multidimensional dielectric loss data sample of the next moment of the preset period and the dielectric loss prediction value of the next moment of the preset period are added to the multidimensional dielectric loss time series training set.
[0070] In a preferred embodiment, the preset frequency loss threshold is determined according to a preset threshold adjustment curve and the current number of iterations.
[0071] Specifically, when training the initial dielectric loss prediction model, multidimensional dielectric loss data samples for a preset time period are obtained from the multidimensional dielectric loss time series training set as input to the initial dielectric loss prediction model, and the actual dielectric loss value at the next moment in the preset time period is output by the initial dielectric loss prediction model, and the initial dielectric loss prediction model is iteratively trained. During each iterative training process, after the multidimensional dielectric loss data samples for a preset time period are obtained from the multidimensional dielectric loss time series training set and input into the initial dielectric loss prediction model, the initial dielectric loss prediction model uses a convolutional attention mechanism to process the dielectric loss data samples, temperature samples, electric field interference samples, and electromagnetic interference samples, and outputs dielectric loss prediction values corresponding to the multidimensional dielectric loss data samples for the preset time period. For example, if the preset time period is [n, n+k], and the fixed time length of the preset time period is k, multidimensional dielectric loss data samples for [n, n+k] are obtained from the multidimensional dielectric loss time series training set as input to the initial dielectric loss prediction model, so that the initial dielectric loss prediction model can single-step predict the dielectric loss prediction value at time n+k+1 and the frequency prediction value at time n+k+1. The actual dielectric loss value and frequency actual value at time n+k+1 are obtained from the multi-dimensional dielectric loss time series training set. The dielectric loss value of the current iteration is determined based on the dielectric loss prediction value at time n+k+1 and the actual dielectric loss value at time n+k+1. The frequency loss value of the current iteration is determined based on the frequency prediction value at time n+k+1 and the actual frequency value at time n+k+1. In the present invention, the dielectric loss value serves as the loss function of the initial dielectric loss prediction model. The weight of the initial dielectric loss prediction model is adjusted by the dielectric loss value of the current iteration to assist in optimizing the weight of the initial dielectric loss prediction model. When the dielectric loss value is minimized, the initial dielectric loss prediction model converges to generate a dielectric loss prediction model. The frequency loss value is used to determine the reliability of the predicted data. If the frequency loss value at the next moment in the preset period is less than the preset frequency loss threshold for the corresponding number of iterations, the dielectric loss prediction value at the next moment in the preset period is considered reliable data. The multidimensional dielectric loss data sample and the dielectric loss prediction value at the next moment in the preset period are added to the multidimensional dielectric loss time series training set to assist in the prediction of subsequent time steps. That is, in the next iterative training, the multidimensional dielectric loss data sample [n+1, n+k+1] is used as the input of the initial dielectric loss prediction model, so that the initial dielectric loss prediction model predicts the dielectric loss prediction value at time n+k+2 and the frequency prediction value at time n+k+2 in a single step. This method allows the multidimensional dielectric loss time series training set to be continuously iteratively updated as the model training process progresses. If the frequency loss value at the next moment in the preset period is greater than or equal to the preset frequency loss threshold for the corresponding number of iterations, the dielectric loss prediction value at the next moment in the preset period is considered unreliable data and is discarded.Among them, the preset frequency loss threshold is determined according to the threshold adjustment curve, and decreases as the number of iterations increases. The threshold adjustment curve is a cosine decline curve, whose horizontal axis is the number of training rounds, that is, the number of iterations, and whose vertical axis is the preset frequency loss threshold.
[0072] Preferably, the performance of the generated dielectric loss prediction model is evaluated, and the prediction effect of the dielectric loss prediction model is evaluated by multiple evaluation indicators such as mean absolute error (MAE), relative absolute error (RAE), root mean square error (RMSE), and determination coefficient (R2), and the model weight of the dielectric loss prediction model is fine-tuned according to the prediction effect.
[0073] In step S3, dielectric loss of the target high-voltage cable is monitored based on the dielectric loss value at the next moment. Preferably, the dielectric loss prediction model is set in the monitoring equipment of the target high-voltage cable to predict the dielectric loss value of the next moment of the multi-dimensional dielectric loss data to be processed for the preset period in real time based on the multi-dimensional dielectric loss data to be processed for the preset period. The preset period is updated with the dielectric loss value at the next moment, and the prediction is continuously iterated to achieve real-time and long-term monitoring of the target high-voltage cable. At the same time, to avoid error accumulation, the predicted dielectric loss value is regularly revised.
[0074] Based on the above method embodiments, the present invention provides corresponding device embodiments.
[0075] like Figure 3 As shown, an embodiment of the present invention provides a high-voltage cable dielectric loss monitoring device, comprising: a data acquisition module, a dielectric loss prediction module and a monitoring module;
[0076] The data acquisition module is configured to acquire the multi-dimensional dielectric loss data to be processed for a preset period of time of the target high-voltage cable; wherein the multi-dimensional dielectric loss data to be processed includes: the dielectric loss data to be processed and associated data of the dielectric loss data to be processed; the associated data includes temperature, electric field interference, and electromagnetic interference;
[0077] The dielectric loss prediction module is configured to input the multidimensional dielectric loss data to be processed in the preset time period into the dielectric loss prediction model, so that the dielectric loss prediction model outputs the dielectric loss value of the multidimensional dielectric loss data to be processed in the preset time period at the next moment;
[0078] The monitoring module is used to monitor the dielectric loss of the target high-voltage cable according to the dielectric loss value at the next moment.
[0079] In a preferred embodiment, it further includes a dielectric loss prediction model building module;
[0080] The dielectric loss prediction model construction module is used to obtain a multidimensional dielectric loss time series training set; wherein the multidimensional dielectric loss time series training set is composed of multidimensional dielectric loss data samples at a plurality of consecutive moments, each multidimensional dielectric loss data sample includes a dielectric loss data sample, a temperature sample, an electric field interference sample, and an electromagnetic interference sample, and each multidimensional dielectric loss data sample is marked with a corresponding dielectric loss true value;
[0081] An initial dielectric loss prediction model is constructed, and the initial dielectric loss prediction model is iteratively trained using the multi-dimensional dielectric loss time series training set until the initial dielectric loss prediction model converges, thereby generating a dielectric loss prediction model.
[0082] It should be noted that the device embodiments described above are merely illustrative, wherein 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 across multiple network units. Some or all of the modules may be selected according to actual needs to achieve the purpose of the present embodiment. In addition, in the drawings of the device embodiments provided by the present invention, the connection relationship between the modules indicates that there is a communication connection between them, which may be specifically implemented as one or more communication buses or signal lines. A person of ordinary skill in the art can understand and implement the present invention without inventive effort.
[0083] Those skilled in the art will clearly understand that for the sake of convenience and brevity, the specific working process of the device described above can refer to the corresponding process in the aforementioned method embodiment, and will not be repeated here.
[0084] Based on the above method embodiment, the present invention provides a corresponding terminal device embodiment.
[0085] An embodiment of the present invention provides a terminal device, including a processor, a memory, and a computer program stored in the memory and configured to be executed by the processor. When the processor executes the computer program, a high-voltage cable dielectric loss monitoring method as described in any one of the present inventions is implemented.
[0086] The terminal device may be a computing device such as a desktop computer, a notebook computer, a PDA, a cloud server, etc. The terminal device may include, but is not limited to, a processor and a memory.
[0087] The processor may be a central processing unit (CPU), other general-purpose processors, digital signal processors (DSP), application-specific integrated circuits (ASIC), field-programmable gate arrays (FPGA) or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. The general-purpose processor may be a microprocessor or any conventional processor, etc. The processor is the control center of the terminal device, connecting various parts of the entire terminal device using various interfaces and lines.
[0088] The memory can be used to store the computer program, and the processor realizes various functions of the terminal device by running or executing the computer program stored in the memory and calling the data stored in the memory. The memory can mainly include a program storage area and a data storage area, wherein the program storage area can store an operating system, at least one application required for a function, etc.; the data storage area can store data created based on the use of the mobile phone, etc. In addition, the memory can include a high-speed random access memory and can also include a non-volatile memory, such as a hard disk, a memory, a plug-in hard disk, a smart memory card (Smart Media Card, SMC), a secure digital (Secure Digital, SD) card, a flash card (Flash Card), at least one disk storage device, a flash memory device, or other volatile solid-state storage device.
[0089] Based on the above method embodiment, the present invention provides a corresponding storage medium embodiment.
[0090] An embodiment of the present invention provides a storage medium, which includes a stored computer program. When the computer program is running, the device where the storage medium is located is controlled to execute a high-voltage cable dielectric loss monitoring method as described in any one of the present inventions.
[0091] The storage medium is a computer-readable storage medium, and the computer program is stored in the computer-readable storage medium. When the computer program is executed by the processor, the steps of each of the above-mentioned method embodiments can be implemented. The computer program includes computer program code, and the computer program code can be in source code form, object code form, executable file or some intermediate form. The computer-readable medium may include: any entity or device capable of carrying the computer program code, a recording medium, a USB flash drive, a mobile hard disk, a magnetic disk, an optical disk, a computer memory, a read-only memory (ROM), a random access memory (RAM), an electrical carrier signal, a telecommunications signal, and a software distribution medium.
[0092] The above is a preferred embodiment of the present invention. It should be pointed out that for ordinary technicians in this technical field, several improvements and modifications can be made without departing from the principles of the present invention. These improvements and modifications are also considered to be within the scope of protection of the present invention.
Claims
1. A method for monitoring dielectric loss of a high-voltage cable, characterized in that: include: Acquire multidimensional dielectric loss data to be processed for a preset period of time of a target high-voltage cable; wherein the multidimensional dielectric loss data to be processed includes: dielectric loss data to be processed and associated data of the dielectric loss data to be processed; the associated data includes temperature, electric field interference, and electromagnetic interference; Inputting the multidimensional dielectric loss data to be processed for the preset period into a dielectric loss prediction model, so that the dielectric loss prediction model outputs a dielectric loss value of the multidimensional dielectric loss data to be processed for the preset period at the next moment; The dielectric loss of the target high-voltage cable is monitored according to the dielectric loss value at the next moment.
2. A method for monitoring dielectric loss of a high-voltage cable according to claim 1, characterized in that: The construction of the dielectric loss prediction model includes: Obtain a multidimensional dielectric loss time series training set; wherein the multidimensional dielectric loss time series training set is composed of multidimensional dielectric loss data samples at a plurality of consecutive moments, each multidimensional dielectric loss data sample includes a dielectric loss data sample, a temperature sample, an electric field interference sample, and an electromagnetic interference sample, and each multidimensional dielectric loss data sample is marked with a corresponding dielectric loss true value; An initial dielectric loss prediction model is constructed, and the initial dielectric loss prediction model is iteratively trained using the multi-dimensional dielectric loss time series training set until the initial dielectric loss prediction model converges, thereby generating a dielectric loss prediction model.
3. A method for monitoring dielectric loss of a high-voltage cable according to claim 2, characterized in that: The initial dielectric loss prediction model is iteratively trained using the multi-dimensional dielectric loss time series training set, including: Iteratively training the initial dielectric loss prediction model using a multidimensional dielectric loss data sample sequence of a preset time period as input to the initial dielectric loss prediction model and using the actual dielectric loss value at the next moment of the preset time period as output of the initial dielectric loss prediction model; In each iterative training process, the initial dielectric loss prediction model uses a convolutional attention mechanism to process dielectric loss data samples, temperature samples, electric field interference samples, and electromagnetic interference samples, and outputs a dielectric loss prediction value at the next moment in a preset time period; Determine the dielectric loss value of the current iteration according to the dielectric loss prediction value at the next moment in the preset period and the dielectric loss actual value at the next moment in the preset period; The weight of the initial dielectric loss prediction model is adjusted according to the dielectric loss value of the current iteration.
4. A method for monitoring dielectric loss of a high-voltage cable according to claim 2, characterized in that: The multi-dimensional dielectric loss data sample also includes a true value of frequency; The multi-dimensional dielectric loss time series training set is used to iteratively train the initial dielectric loss prediction model, including: Iteratively training the initial dielectric loss prediction model using a multidimensional dielectric loss data sample sequence of a preset time period as input to the initial dielectric loss prediction model and using the actual dielectric loss value at the next moment of the preset time period as output of the initial dielectric loss prediction model; In each iterative training process, the initial dielectric loss prediction model uses a convolutional attention mechanism to process dielectric loss data samples, temperature samples, electric field interference samples, and electromagnetic interference samples, and outputs a dielectric loss prediction value at the next moment in a preset period and a frequency prediction value at the next moment in the preset period; Determine the dielectric loss value of the current iteration according to the dielectric loss prediction value at the next moment in the preset period and the dielectric loss actual value at the next moment in the preset period; Adjust the weight of the initial dielectric loss prediction model according to the dielectric loss value of the current iteration; Determine the frequency loss value of the current iteration according to the frequency prediction value at the next moment of the preset period and the actual frequency value at the next moment of the preset period; If the frequency loss value of the current iteration is less than the preset frequency loss threshold, the multidimensional dielectric loss data sample at the next moment of the preset period and the dielectric loss prediction value at the next moment of the preset period are added to the multidimensional dielectric loss time series training set.
5. A method for monitoring dielectric loss of a high-voltage cable according to claim 4, characterized in that: The preset frequency loss threshold is determined according to a preset threshold adjustment curve and the current number of iterations.
6. A method for monitoring dielectric loss of a high-voltage cable according to claim 2, characterized in that: After obtaining the multi-dimensional dielectric loss time series training set, it also includes: Wavelet denoising is performed on the multidimensional dielectric loss time series training set to obtain the denoised multidimensional dielectric loss time series training set.
7. A high-voltage cable dielectric loss monitoring device, characterized in that: include: Data acquisition module, dielectric loss prediction module and monitoring module; The data acquisition module is configured to acquire the multi-dimensional dielectric loss data to be processed for a preset period of time of the target high-voltage cable; wherein the multi-dimensional dielectric loss data to be processed includes: the dielectric loss data to be processed and associated data of the dielectric loss data to be processed; the associated data includes temperature, electric field interference, and electromagnetic interference; The dielectric loss prediction module is configured to input the multidimensional dielectric loss data to be processed in the preset time period into the dielectric loss prediction model, so that the dielectric loss prediction model outputs the dielectric loss value of the multidimensional dielectric loss data to be processed in the preset time period at the next moment; The monitoring module is used to monitor the dielectric loss of the target high-voltage cable according to the dielectric loss value at the next moment.
8. A high-voltage cable dielectric loss monitoring device according to claim 7, characterized in that: It also includes a module for building a dielectric loss prediction model; The dielectric loss prediction model construction module is used to obtain a multidimensional dielectric loss time series training set; wherein the multidimensional dielectric loss time series training set is composed of multidimensional dielectric loss data samples at a plurality of consecutive moments, each multidimensional dielectric loss data sample includes a dielectric loss data sample, a temperature sample, an electric field interference sample, and an electromagnetic interference sample, and each multidimensional dielectric loss data sample is marked with a corresponding dielectric loss true value; An initial dielectric loss prediction model is constructed, and the initial dielectric loss prediction model is iteratively trained using the multi-dimensional dielectric loss time series training set until the initial dielectric loss prediction model converges, thereby generating a dielectric loss prediction model.
9. A terminal device, characterized in that: The method comprises a processor, a memory, and a computer program stored in the memory and configured to be executed by the processor, wherein when the processor executes the computer program, a high-voltage cable dielectric loss monitoring method according to any one of claims 1 to 6 is implemented.
10. A storage medium, characterized in that: The storage medium includes a stored computer program, wherein when the computer program is running, the device where the storage medium is located is controlled to execute the high-voltage cable dielectric loss monitoring method according to any one of claims 1 to 6.