Electric arc risk situation awareness prediction method and system based on multi-characteristic dynamic mapping

Abstract

The invention discloses an arc risk situation awareness prediction method and system based on multi-characteristic dynamic mapping, and belongs to the technical field of electrical fault detection.The method comprises the steps that arc characteristics are extracted according to a current waveform of an electric power system, and the arc generation position is determined by combining the propagation rule of the arc characteristics; predicting the development trend of the arc form image and the energy signal according to the arc form image and the energy signal of the current waveform; according to the development trend of the arc characteristics, taking the arc as a heat source, reflecting the contact area of the arc and air by the arc form image, and combining with the heat transfer rule of the arc to obtain the temperature field distribution of the arc; and according to the arc generation position and the temperature field distribution, in combination with the distance perception between the ignition point of the evaluation object and the arc boundary, predicting the arc risk on the installation line of the evaluation object, and making corresponding emergency prevention measures according to the risk evaluation result. And accurate prediction, reliable identification and rapid cutting-off of the fault arc are realized.

Description

Technical Field

[0001] The present invention relates to the technical field of electrical fault detection, and in particular to an arc risk situation awareness prediction method and system based on multi-characteristic dynamic mapping. Background Art

[0002] The global energy crisis and climate warming are becoming increasingly serious, prompting the widespread application of new green renewable energy such as photovoltaics, wind power, and fuel cells. Modern power systems based on new energy sources rely on the development of DC power distribution technology, and DC power distribution systems have become an important development direction. DC fault arcs often occur in electrical systems containing high-voltage DC power supplies. On the DC side, any device damage may cause an arc, and factors such as animal bites, aging of electronic devices, cable weathering, and loose installation joints will increase the possibility of DC line damage and thus fault arcs. Since there is no zero crossing point of AC fault arcs, if protective measures are not taken in time, it will cause huge damage to photovoltaic modules and transmission lines, and even cause fires, resulting in serious economic losses and casualties, highlighting the urgency and importance of arc fault diagnosis research.

[0003] At present, the relevant research on the field of fault arc diagnosis at home and abroad is based on two aspects. One is to extract the fault arc image features (color, shape and spatial relationship features) through the inspection technology of the Internet of Things robot or drone for visual analysis. Among them, the substation intelligent inspection robot is an intelligent detection device with intelligent detection devices such as high-definition visible light camera and infrared thermal imager. The intelligent inspection robot can complete the control closed loop of 24-hour rapid data collection, real-time transmission of equipment information, big data analysis and early warning and rapid feedback, so as to realize the automatic detection and intelligent analysis of the power equipment status, but it cannot meet the real-time requirements of fault detection. If the fault detection camera is installed on the entire line, the investment is large and the cost is high. It is easy to have monitoring blind spots and cannot meet the requirements of full coverage of fault points. The second is to extract the electrical signal characteristics of the fault arc for fault diagnosis. The researchers proposed the time-frequency domain detection method. The time domain detection method mainly realizes arc fault detection by analyzing and processing the collected voltage, current and other signals to obtain typical characteristics under fault conditions such as current standard deviation, current change rate and other information. The time domain detection method usually performs Fourier transform on the current data to obtain current spectrum information under different states and analyze its differences. For the time-frequency domain detection method, wavelet transform analysis method is used to detect arc faults. In addition, some researchers have proposed DC arc detection methods based on arc electromagnetic radiation characteristics and physical signals, and fault detection is achieved by measuring arc electromagnetic radiation signals or physical signals.

[0004] In the field of image technology diagnosis, all shape features currently lack a complete mathematical model and no theoretical support, so image technology is not particularly widely used. Spatial relationship features are very sensitive to local image features, but they are too dependent on the translation, rotation or size transformation of image objects, which often leads to the failure to achieve the desired effect in detection. In addition, spatial relationship features are used in conjunction with other image features and are generally not used alone. In the field of electrical signal diagnosis, Fourier transform is a global transform that can only analyze stationary random signals and cannot reflect the local changes of the analyzed signal in the time domain, so it is impossible to find the moment of the fault arc. Wavelet feature extraction will have the disadvantage of a significant decrease in the diagnosis rate in certain scenarios, poor adaptability to multiple working conditions, and feature weakening in some scenarios, which brings challenges to the extraction of fault arc features and the improvement of detection accuracy. In a DC system in an open environment, the increase in interference factors will make the physical signal features unclear, and the fault arc will be missed, so this method is not suitable for large-scale low-voltage DC systems. However, the hardware implementation cost of fault arc detection based on electromagnetic radiation is high, and it is difficult to be widely used in DC power grids. Summary of the invention

[0005] In response to the problems existing in the prior art, the present invention provides an arc risk situation awareness and prediction method and system based on multi-characteristic dynamic mapping, which predicts arc energy and arc area through a current-image correlation model, predicts the final arc risk development trend, and realizes accurate prediction, reliable identification, and rapid disconnection of fault arcs.

[0006] The present invention is achieved through the following technical solutions: In a first aspect, the present application provides an arc risk situation awareness prediction method based on multi-characteristic dynamic mapping, characterized in that it includes the following steps: Step 1: extract arc characteristics according to the current waveform of the power system, and determine the arc occurrence location in combination with the propagation law of the arc characteristics; Step 2: The arc morphology image and energy signal are obtained based on the arc characteristics, and the development trend of the arc characteristics is determined in combination with the prediction model; Step 3: According to the development trend of arc characteristics, the arc itself is used as the heat source. The arc morphology image reflects the contact area between the arc and the air. Combined with the heat transfer law of the arc, the temperature field distribution of the arc is obtained. Step 4: Based on the arc location and temperature field distribution, combined with the distance perception between the ignition point of the assessed object and the arc boundary, predict the arc risk on the installation line of the assessed object, and formulate corresponding emergency prevention measures based on the risk assessment results.

[0007] Furthermore, the current waveform of the power system described in step 1 is calculated by frequency fsPerform real-time sampling of the power system to obtain current signals.

[0008] Furthermore, extracting the arc characteristics in step 1 includes transforming the current waveform into a two-dimensional function of time and frequency based on a time-frequency transformation method, and determining the arc characteristics according to the two-dimensional function.

[0009] Furthermore, the method for determining the arc occurrence position in step 1 is as follows: The arc location is determined based on the propagation law that the arc characteristics decay as the line grows and the arc location model is combined.

[0010] Furthermore, the dynamic mapping model in step 2 is constructed based on the variational autoencoder correlation analysis method, and the dynamic mapping model outputs the mapping relationship between the arc features and the arc morphology image and energy signal.

[0011] Furthermore, in step 2, the arc morphology image and energy signal at the current moment are combined with the prediction model to obtain its development trend within a period of time, forming a development trend T1 of the arc morphology image and energy signal. At the next moment, the same current feature is predicted to obtain a development trend T2, which is compared with the development trend T1. According to the comparison result, the development trend T1 is corrected and updated to obtain an updated development trend.

[0012] Furthermore, the prediction model is a random forest prediction model, a linear regression model or a logistic regression model.

[0013] Furthermore, in step 3, according to the updated development trend, the arc itself is used as the heat source, the arc morphology image reflects the contact area between the arc and the air, and the arc temperature field distribution is obtained in combination with the heat transfer law of the arc; The heat transfer law of the arc is that the temperature attenuation medium is air, the attenuation direction is horizontal, and the temperature attenuation is inversely proportional to the square of the distance, thus forming a temperature field.

[0014] In a second aspect, the present application provides an arc risk situation awareness prediction system based on multi-characteristic dynamic mapping, characterized in that it includes: The arc location module is used to extract arc characteristics based on the current waveform of the power system and determine the arc location based on the propagation law of the arc characteristics; A trend prediction module is used to determine the development trend of the arc characteristics based on the arc morphology image and energy signal of the arc characteristics and in combination with the prediction model; The temperature distribution module is used to obtain the temperature field distribution of the arc based on the development trend of the arc characteristics and the arc itself as the heat source. The arc morphology image reflects the contact area between the arc and the air. Combined with the heat transfer law of the arc, the arc is used to obtain the arc temperature field distribution. The prediction module predicts the arc risk on the installation line of the assessed object based on the arc location and temperature field distribution, combined with the distance perception between the ignition point of the assessed object and the arc boundary, and formulates corresponding emergency prevention measures based on the risk assessment results.

[0015] In a third aspect, the present application provides an electronic device, characterized in that it includes: Memory for storing computer programs; A processor is used to implement the steps of the arc risk situation awareness and prediction method based on multi-characteristic dynamic mapping as described in any one of claims 1 to 8 when executing the computer program.

[0016] Compared with the prior art, the present invention has the following beneficial technical effects: The arc risk situation awareness prediction method based on multi-characteristic dynamic mapping provided by the present application determines the arc occurrence location according to the propagation law of the arc characteristics, obtains the arc morphology image and arc energy development trend based on the prediction model according to the mapping relationship between the arc characteristics and the arc morphology image and arc energy, takes the arc itself as the heat source, and the arc morphology image reflects the contact area between the arc and the air, and combines the relevant knowledge of heat transfer to obtain the arc temperature field distribution. According to the arc occurrence location and temperature field distribution, combined with the distance perception between the ignition point of the assessed object and the arc boundary, the arc risk on the installation line of the assessed object is predicted, and corresponding emergency prevention measures are formulated according to the risk assessment results. The method is to collect current waveforms and combine correlation models to predict arc development trends and risk level assessments, and make corresponding emergency measures, thereby achieving accurate prediction, reliable identification, and rapid disconnection of fault arcs.

[0017] The present application also proposes an arc risk situation awareness prediction system based on multi-characteristic dynamic mapping and an electronic device, which have all the advantages of the above-mentioned arc risk situation awareness prediction method based on multi-characteristic dynamic mapping. BRIEF DESCRIPTION OF THE DRAWINGS

[0018] In order to more clearly illustrate the technical solutions of the embodiments of the present application, the drawings required for use in the embodiments will be briefly introduced below. It should be understood that the following drawings only show certain embodiments of the present application and therefore should not be regarded as limiting the scope. For ordinary technicians in this field, other related drawings can be obtained based on these drawings without creative work.

[0019] Figure 1 This is a flow chart of the arc risk situation awareness and prediction method of the present invention; Figure 2 This is a schematic diagram of the arc risk situation perception prediction project of the present invention; Figure 3 This is the arc current waveform diagram of the arc drawing method of the present invention; Figure 4 The current signal waveform and its wavelet packet transform characteristics of the present invention; Figure 5a The arc image at the initial stage of the arc drawing method predicted by the present invention; Figure 5b The arc image of the stable combustion stage of the arc drawing method predicted by the present invention; Figure 6a This is a current waveform diagram of the fuse arcing method of the present invention; Figure 6b The arc image of the initial stage of the fuse arcing method predicted by the present invention; Figure 6c The arc image of the stable combustion stage of the fuse arcing method predicted by the present invention; Figure 7 This is a graph of the arcing shape area evolution trend predicted by the present invention based on the measured arc current waveform. DETAILED DESCRIPTION

[0020] In order to make the purpose, technical solution and advantages of the embodiments of the present application clearer, the technical solution in the embodiments of the present application will be clearly and completely described below in conjunction with the drawings in the embodiments of the present application. Obviously, the described embodiments are part of the embodiments of the present application, rather than all the embodiments. The components of the embodiments of the present application described and shown in the drawings here can be arranged and designed in various different configurations.

[0021] Therefore, the following detailed description of the embodiments of the present application provided in the accompanying drawings is not intended to limit the scope of the present application for which protection is sought, but merely represents selected embodiments of the present application. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in the field without creative work are within the scope of protection of the present application.

[0022] An arc risk situation awareness prediction method based on multi-characteristic dynamic mapping includes the following steps: Step 1: extract arc characteristics according to the current waveform of the power system, and determine the arc occurrence location in combination with the propagation law of the arc characteristics; Step 2: The arc morphology image and energy signal are obtained based on the arc characteristics, and the development trend of the arc characteristics is determined in combination with the prediction model; Step 3: According to the development trend of arc characteristics, the arc itself is used as the heat source. The arc morphology image reflects the contact area between the arc and the air. Combined with the heat transfer law of the arc, the temperature field distribution of the arc is obtained. Step 4: Based on the arc location and temperature field distribution, combined with the distance perception between the ignition point of the assessed object and the arc boundary, predict the arc risk on the installation line of the assessed object, and formulate corresponding emergency prevention measures based on the risk assessment results.

[0023] This application obtains arc characteristics through time-frequency transformation, and thus determines the location of arc occurrence in combination with propagation laws. Based on the dynamic mapping method, the mapping set reflects the correspondence between arc characteristics and arc morphology images and arc energy, thereby predicting the development trend of arc morphology images and arc energy. The development trend is corrected by continuously updating the mapping set to ensure high accuracy of the development trend prediction. According to the predicted development trend of the arc characteristics, the arc itself is used as the heat source, and the arc morphology image reflects the contact area between the arc and the air. The arc temperature field distribution is obtained in combination with relevant knowledge of heat transfer. Based on the perception of the distance between the ignition point of the object to be evaluated on the line and the arc boundary, the arc risk on the installation line of the object to be evaluated is evaluated in combination with artificial intelligence, and corresponding emergency prevention measures are formulated according to the risk assessment results.

[0024] Example 1 See also Figure 1-7 , an arc risk situation awareness prediction method based on multi-characteristic dynamic mapping, comprising the following steps: Step 1: By frequency fs The current signal of the power system is sampled in real time to obtain the real-time current waveform of the power system; the system current signal is sampled in real time to achieve real-time monitoring and form complex and changeable arc data.

[0025] Step 2: extract arc features by transforming the current waveform based on the time-frequency transformation method, that is, representing the current waveform as a two-dimensional function of time and frequency; determine the arc location based on the propagation law that the arc features decay as the line grows, combined with the arc positioning model; The arc morphology image and energy signal are obtained according to the dynamic mapping model and the current waveform. The mapping relationship of the dynamic mapping model is one-to-one corresponding, that is, a unique current signal corresponds to a unique arc morphology image and energy signal.

[0026] In this embodiment, the time-frequency transformation method adopted is wavelet transformation, which can achieve localized analysis of the frequency inside a small time window, and gradually perform multi-scale refinement of the signal through telescoping and translation operations, ultimately achieving time subdivision at high frequencies and frequency subdivision at low frequencies, and can automatically adapt to the requirements of time-frequency signal analysis, thereby focusing on any details of the signal.

[0027] The time-frequency transformation method used is wavelet transformation. Wavelet transformation provides a "time-frequency" window that changes with frequency, and there is also a large family of wavelet bases. Through transformation, the characteristics of certain aspects of the problem can be fully highlighted, and the localized analysis of the frequency inside the small time window can be achieved. Through the telescoping and translation operations, the signal is gradually refined at multiple scales, and finally the time subdivision at high frequencies and the frequency subdivision at low frequencies are achieved. It can automatically adapt to the requirements of time-frequency signal analysis, so as to focus on any details of the signal. The wavelet transformation method is as follows:

[0028] In the formula, f ( t ) is the original signal, Ψ ( τ-t ) is the wavelet basis function, α is the scale parameter, τ is the translation parameter, WT ( α, τ ) are the wavelet coefficients.

[0029] In this embodiment, the arc location model is a pre-trained memory neural network, which is used to determine the arc location. The arc location is determined based on the wavelet characteristics reflecting the arc characteristic propagation attenuation law, the arc characteristic attenuation law as the line grows and the memory neural network.

[0030] The classification function of the memory neural network gives a rough arc occurrence range, or the regression function gives a fine arc occurrence position.

[0031] In this embodiment, the dynamic mapping model should ensure that the data meets relevant conditions, such as appropriate data types, linear relationships, normality, independence, etc. In addition, correlation analysis can only reveal the strength and direction of the relationship between variables, but cannot derive causal relationships. Therefore, special attention should be paid to not confuse correlation and causality during analysis. The dynamic mapping model is preferably a variational autoencoder correlation analysis to form a mapping model library of current waveforms, arc morphology images, and energy signals.

[0032] The current input is based on the dynamic mapping model to obtain the arc morphology image and energy signal. The variational autoencoder is a generative model and an extension of the autoencoder. It combines the ideas of autoencoders and variational inference in deep learning, and aims to generate new data samples by learning the potential distribution of data. The loss function of the variational autoencoder consists of two parts: reconstruction loss and KL divergence. The final loss function is the sum of the two parts, and its calculation formula is as follows:

[0033] in, q ( z | x ) is the conditional distribution of the latent variable output by the encoder; p ( x | z ) is the conditional distribution of the observed data output by the decoder; p ( z ) is the prior distribution of the latent variable (usually a standard normal distribution).

[0034] Step 3, according to the arc morphology image and energy signal at the current moment, combined with the prediction model, obtain its development trend within a period of time, and form the development trend T1 of the arc morphology image and energy signal; at the next moment, predict the same current feature to obtain the development trend T2 and compare it with the development trend T1, and according to the comparison result, correct and update the development trend T1 to obtain the updated development trend T1A, so as to improve the prediction accuracy of the development trend of long-term arc risk.

[0035] In this embodiment, the prediction model is a model used to predict unknown data or events, which can be predicted based on historical data, statistical analysis or machine algorithms. The prediction model should meet the principles of simple establishment process, ability to output continuous change data, high prediction accuracy, etc., preferably a random forest prediction model, which is an integrated model composed of multiple decision trees, and improves the accuracy of prediction through collective decision-making. Other available prediction models include linear regression model, logistic regression model, etc.

[0036] The random forest prediction model is an integrated model composed of multiple decision trees, which improves the accuracy of prediction through collective decision-making. As time accumulates, a new mapping set is formed, and the development trend T2 is formed by mapping and predicting the same current signal; by comparing T1 with T2, the development trend T1 is corrected and updated to improve the prediction accuracy for a longer time; Step 4: Based on the arc morphology image and the energy development trend T1A, the arc itself is used as the heat source. The arc morphology image reflects the contact area between the arc and the air. Combined with the relevant knowledge of heat transfer, the temperature field distribution of the arc is obtained; Among them, the arc can be regarded as a point heat source, the temperature attenuation medium is air, the attenuation direction is horizontal, and the temperature attenuation is roughly inversely proportional to the square of the distance, thus forming a temperature field and clarifying the arc temperature field distribution and arc boundary.

[0037] In the near field area close to the point heat source, the temperature decays relatively quickly, but it is not a simple linear relationship. As the distance from the point heat source increases, after entering the far field area, the temperature decay rate gradually slows down and approaches a stable decay trend. The temperature decay is roughly inversely proportional to the square of the distance, thus forming a temperature field.

[0038] Step 5: According to the arc location and temperature field distribution, combined with the distance perception between the ignition point of the assessed object and the arc boundary, predict the arc risk on the installation line of the assessed object, and take corresponding emergency measures according to the development trend of the arc risk to ensure the safe operation of the system.

[0039] The arc shape and energy signal obtained by mapping are combined with the prediction model to obtain its development trend. Combined with the fault location result, the distance difference ΔS is formed to determine the installation and placement position of the object to be tested. The arc risk level is perceived and evaluated in combination with the ignition point of the object to be tested and the distance from the arc boundary to ensure that the distance and temperature are within the safe operating range. The ignition point of the flammable material is known to be B, and the arc temperature is A, forming a temperature field. As the temperature decreases outward, it should be determined that the flammable materials and important facilities are located outside the circular area with the arc as the center and ΔS as the radius. The temperature of the temperature field is less than B. According to the development trend of the arc risk, appropriate emergency measures are taken to ensure the safe operation of the system.

[0040] Figure 2 The actual operation platform used mainly consists of six parts: DC simulator, arc generator, line load, current sensor, and switch. The DC simulator is mainly composed of an AC power supply and a rectifier, which work together to output DC power to power the experimental circuit; the Arc Fault Generator (AFG) generates an arc; the load represents the electrical equipment; the current sensor samples and records the current signal when the arc occurs, providing a data set for feature extraction; the switch can control the arc occurrence, load, and current sensor switching. The circle, triangle, and five-pointed star represent different objects to be tested in different positions, and the distance between the object to be tested and the arc is R. The dotted line in the figure represents the isotherm of the partial temperature field formed around the fault arc after the arc occurs. Due to the different ignition points of the objects to be evaluated and the distance from the arc, the risk level and probability of fire and other hazards that may be caused are different. The development trend of the arc energy and arc morphology image is obtained through the prediction model, combined with the ignition point of the object to be evaluated and the distance perception from the arc boundary, to ensure that the distance and temperature are within the safe operating range. According to the analysis, the objects under test represented by the triangle and the five-pointed star are in dangerous positions and their placement should be redesigned. At the same time, the arc risk assessment is performed through artificial intelligence algorithms based on data characteristics, and corresponding emergency measures are taken according to the development trend of arc risk to ensure the safe operation of the system.

[0041] Table 1 Sample correlation statistics of current-image features

[0042] As can be seen from Table 1, the Pearson coefficient evaluates the current-image features: arcing area, left and right arc root motion trajectories, upper and lower boundary motion trajectories. Based on the wavelet packet transform, according to the statistical data, it is found that the frequency of the right arc root motion trajectory feature correlation degree below the threshold is 75%, which is 25% more than the lower boundary and 37.5% more than the area, left arc root and upper boundary. Therefore, it is determined that the right arc root motion trajectory feature has the weakest correlation with the wavelet packet transform feature. At the same time, the correlation between other features and wavelet packet features is compared horizontally in Table 1, and it is found that the right arc root motion trajectory has a lower similarity. Therefore, it can be considered that the image feature non-correlation feature can be selected as the right arc root motion trajectory feature. In summary, it is finally believed that the non-correlation feature group that can best characterize the arc instability is: short-time Fourier transform feature-wavelet packet feature-right arc root motion trajectory feature.

[0043] Figure 3 This is the arc current waveform of the arc pulling method. In the startup phase, the current waveform rises rapidly and reaches the normal working current after 1s. When the arc occurs, the current drops significantly due to the large air resistance. At this time, the arc enters the arc burning stage. After the arc lasts for a period of time, the power is cut off and the arc goes out. In the arc burning state, since the arc carries a lot of heat when it occurs, the insulation layer of the equipment around the arc will be damaged under long-term high temperature, causing a short circuit fault. If there is no human intervention, the arc will continue to burn, and the high temperature will ignite nearby combustibles, causing a large-scale fire.

[0044] Figure 4 The current signal waveform and its wavelet packet transform characteristics. It can be seen that in the normal state of the circuit, the amplitude of the characteristic waveform constructed by the wavelet packet transform is in a stable state, while in the arcing stage, the amplitude of the characteristic quantity is significantly increased and fluctuates strongly. The wavelet packet transform characteristics clearly distinguish the normal state and arcing state of the current signal. Unlike the short-time Fourier transform waveform, the characteristic quantity obtained by the wavelet packet transform waveform has a low degree of overlap between the normal state and the arcing state. The increase in amplitude and separation of the characteristic quantity in the arcing stage are significantly higher than the characteristic quantity in the normal state. This also reflects the above-mentioned fact that the wavelet transform inherits and develops the localization idea of ​​the short-time Fourier transform, while overcoming the shortcomings such as the window size not changing with frequency.

[0045] The arc evolution images recorded by the high-speed camera in the DC arc experimental platform are converted frame by frame, and the image information is converted into discrete visual digital information. According to the imaging principle, the image is divided into 1024*1280 pixel blocks, and then the pixels are converted into corresponding RGB values ​​to form a numerical matrix. Then, a five-step cycle detection and analysis of grayscale conversion, iterative filtering, flood filling, and setting thresholds for binary analysis is performed to effectively filter out interference such as arc radiation, metal vapor, and metal ablation splashing in the arc burning process, and a series of arc images are obtained as shown below.

[0046] Figure 5a This is the arc image at the initial stage of the arc drawing method; Figure 5b To stabilize the arc image in the combustion stage of the drawn arc method; Figure 6b This is the arc image at the initial stage of the fuse arcing method; Figure 6c The arc image of the stable burning stage of the fuse arc striking method; Figure 6a This is the current waveform of the fuse arcing method. Its current waveform is different from that of the arc pulling method. In the arcing state, the arc does not drop gradually, but is in a pulsed state, dropping rapidly to zero and then rising sharply, with the amplitude rising value decreasing gradually, and repeating this process.

[0047] Figure 7 The figure shows the arc current waveform and its arc burning area characteristic waveform. In the initial arcing stage, we can see that the image has been distorted in the initial short period of time through time-frequency transformation, but because the basic conditions for arcing have not yet been met, the arc cannot break through the air to produce arc light, and then the high-speed camera cannot detect the arc light. This is one of the disadvantages of image detection. The image has hysteresis. When the arc voltage and current meet the breakdown conditions, the electrode begins to burn. In the early stage, the arc cannot maintain stable burning, causing the arc to repeatedly "extinguish-reignite" link, which is also reflected in the arc burning area characteristic waveform; when the arc enters the stable burning link in the air medium, the arc repeatedly jumps, and the arc burning area also changes accordingly. This phenomenon fully demonstrates the instability of the arc. When the arc is fully burned, as the arc current decreases, its burning trend will also weaken. The manifestation of the arc burning area is that the area gradually decreases. When the current is reduced to the point where the arc cannot be maintained, the arc is extinguished.

[0048] The method of the present application has at least the following beneficial effects: (1) The DC fault arc feature extraction method based on wavelet transform proposed in the present invention can perform localized analysis of the frequency within a small time window, and gradually perform multi-scale refinement of the signal through telescoping and translation operations, ultimately achieving time subdivision at high frequencies and frequency subdivision at low frequencies. It can automatically adapt to the requirements of time-frequency signal analysis, thereby focusing on any details of the signal, providing accurate fault arc feature extraction technology for complex working conditions existing in actual power grids, and protecting people’s lives and property.

[0049] (2) The variational autoencoder correlation analysis method proposed in the present invention is a generative model that maps input data to a latent space through an encoder and then generates data through a decoder. The distance between different data points in the latent space can be used as a measure of their similarity. Variational autoencoders are widely used in image generation and anomaly detection, and can be used to evaluate the similarity of data points. Its advantages are strong generation capability: by learning the latent distribution of data, new samples similar to the distribution of training data can be generated; unsupervised learning: training can be performed without labeled data; strong latent space interpretability: due to the introduction of the probability distribution of latent variables, its latent space can have good interpretability, and it is easy to interpolate and explore the latent space.

[0050] (3) The electrical signal-image and energy correlation analysis and prediction method proposed in the present invention breaks through the limitations of limited sensor monitoring and achieves full coverage while taking into account real-time performance. It predicts arc morphology images and energy signals through models, clarifies the distribution and boundaries of the fault arc temperature field, evaluates the arc risk level, and takes corresponding emergency measures based on the development trend of the arc risk to ensure the safe operation of the system.

[0051] (4) This invention is based on the combination of electrical professional knowledge and heat transfer to form a cross-disciplinary approach. Knowledge of thermal radiation, heat conduction, etc. provides a basis for arc heat transfer and temperature field distribution, breaking down barriers between disciplines, generating new ideas and methods, promoting the advancement of science and technology, and generating new research directions and solutions.

[0052] Correspondingly, the present application also provides an arc risk situation awareness prediction system based on multi-characteristic dynamic mapping, including: The arc location module is used to extract arc characteristics based on the current waveform of the power system and determine the arc location based on the propagation law of the arc characteristics; A trend prediction module is used to determine the development trend of the arc characteristics based on the arc morphology image and energy signal of the arc characteristics and in combination with the prediction model; The temperature distribution module is used to obtain the temperature field distribution of the arc based on the development trend of the arc characteristics and the arc itself as the heat source. The arc morphology image reflects the contact area between the arc and the air. Combined with the heat transfer law of the arc, the arc is used to obtain the arc temperature field distribution. The prediction module predicts the arc risk on the installation line of the assessed object based on the arc location and temperature field distribution, combined with the distance perception between the ignition point of the assessed object and the arc boundary, and formulates corresponding emergency prevention measures based on the risk assessment results.

[0053] It should be noted that in the several embodiments provided in the present application, it should be understood that the disclosed devices and methods can be implemented in other ways. For example, the device embodiments described above are only schematic. For example, the division of each module is only a logical function division. There may be other division methods in actual implementation. For example, multiple modules can be combined or integrated into another device, or some features can be ignored or not executed. The module described as a separate component may or may not be physically separated. The component displayed as a module may be a physical unit or multiple physical units, that is, it may be located in one place, or it may be distributed in multiple different places. Some or all of the modules may be selected according to actual needs to achieve the purpose of the scheme of this embodiment.

[0054] In addition, each module in each embodiment of the present invention may be integrated into a processing unit, each module may exist physically separately, or two or more modules may be integrated into one unit. The above-mentioned integrated unit may be implemented in the form of hardware or in the form of software functional units.

[0055] An electronic device provided in an embodiment of the present application includes a memory and a processor. The memory stores a computer program. When the processor executes the computer program, the steps of the arc risk situation awareness prediction method based on multi-characteristic dynamic mapping as described in the above embodiment 1 are implemented.

[0056] Another electronic device provided in the embodiment of the present application may also include: an input port connected to the processor, used to transmit multimodal data collected by an external acquisition device to the processor; and a display unit connected to the processor, used to display the processing results of the processor to the outside world; a communication module connected to the processor, used to realize the communication between the electronic device and the outside world. The display unit can be a display panel, a laser scanning display, etc.; the communication mode adopted by the communication module includes but is not limited to mobile high-definition link technology (HML), universal serial bus (USB), high-definition multimedia interface (HDMI), wireless connection (including wireless fidelity technology (WiFi), Bluetooth communication technology, low-power Bluetooth communication technology, and communication technology based on IEEE802.11s).

[0057] The above contents are only for explaining the technical idea of ​​the present invention and cannot be used to limit the protection scope of the present invention. Any changes made on the basis of the technical solution in accordance with the technical idea proposed by the present invention shall fall within the protection scope of the claims of the present invention.

Claims

1. A method for predicting arc risk situation awareness based on multi-characteristic dynamic mapping, characterized in that: The following steps are involved: Step 1: extract arc characteristics according to the current waveform of the power system, and determine the arc occurrence location in combination with the propagation law of the arc characteristics; Step 2: The arc morphology image and energy signal are obtained based on the arc characteristics, and the development trend of the arc characteristics is determined in combination with the prediction model; Step 3: According to the development trend of arc characteristics, the arc itself is used as the heat source. The arc morphology image reflects the contact area between the arc and the air. Combined with the heat transfer law of the arc, the temperature field distribution of the arc is obtained. Step 4: Based on the arc location and temperature field distribution, combined with the distance perception between the ignition point of the assessed object and the arc boundary, predict the arc risk on the installation line of the assessed object, and formulate corresponding emergency prevention measures based on the risk assessment results.

2. The arc risk situation awareness prediction method based on multi-characteristic dynamic mapping according to claim 1 is characterized in that: The current waveform of the power system described in step 1, by frequency fs Perform real-time sampling of the power system to obtain current signals.

3. The arc risk situation awareness prediction method based on multi-characteristic dynamic mapping according to claim 1 is characterized in that: Extracting arc characteristics in step 1 includes transforming the current waveform into a two-dimensional function of time and frequency based on a time-frequency transformation method, and determining the arc characteristics according to the two-dimensional function.

4. The arc risk situation awareness prediction method based on multi-characteristic dynamic mapping according to claim 1 is characterized in that: The method for determining the arc occurrence location in step 1 is as follows: The arc location is determined based on the propagation law that the arc characteristics decay as the line grows and the arc location model is combined.

5. The arc risk situation awareness prediction method based on multi-characteristic dynamic mapping according to claim 1 is characterized in that: The dynamic mapping model described in step 2 is constructed based on the variational autoencoder correlation analysis method, and the dynamic mapping model outputs the mapping relationship between the arc features and the arc morphology image and energy signal.

6. The arc risk situation awareness prediction method based on multi-characteristic dynamic mapping according to claim 1 is characterized in that: In step 2, the arc morphology image and energy signal at the current moment are combined with the prediction model to obtain its development trend within a period of time, forming the development trend T1 of the arc morphology image and energy signal. At the next moment, the same current feature is predicted to obtain the development trend T2 and compare it with the development trend T1. According to the comparison result, the development trend T1 is corrected and updated to obtain the updated development trend.

7. The arc risk situation awareness prediction method based on multi-characteristic dynamic mapping according to claim 6 is characterized in that: The prediction model is a random forest prediction model, a linear regression model or a logistic regression model.

8. The arc risk situation awareness prediction method based on multi-characteristic dynamic mapping according to claim 6 is characterized in that: In step 3, according to the updated development trend, the arc itself is used as the heat source, the arc morphology image reflects the contact area between the arc and the air, and the arc temperature field distribution is obtained in combination with the heat transfer law of the arc; The heat transfer law of the arc is that the temperature attenuation medium is air, the attenuation direction is horizontal, and the temperature attenuation is inversely proportional to the square of the distance, thus forming a temperature field.

9. An arc risk situation awareness prediction system based on multi-characteristic dynamic mapping, characterized in that: include: The arc location module is used to extract arc characteristics based on the current waveform of the power system and determine the arc location based on the propagation law of the arc characteristics; A trend prediction module is used to determine the development trend of the arc characteristics based on the arc morphology image and energy signal of the arc characteristics and in combination with the prediction model; The temperature distribution module is used to obtain the temperature field distribution of the arc based on the development trend of the arc characteristics and the arc itself as the heat source. The arc morphology image reflects the contact area between the arc and the air. Combined with the heat transfer law of the arc, the arc is used to obtain the arc temperature field distribution. The prediction module predicts the arc risk on the installation line of the assessed object based on the arc location and temperature field distribution, combined with the distance perception between the ignition point of the assessed object and the arc boundary, and formulates corresponding emergency prevention measures based on the risk assessment results.

10. An electronic device, characterized in that: include: Memory for storing computer programs; A processor is used to implement the steps of the arc risk situation awareness and prediction method based on multi-characteristic dynamic mapping as described in any one of claims 1 to 8 when executing the computer program.

Images