A Dual-Power Dynamic Reconfiguration Method and System

Abstract

The present invention discloses a dual - power dynamic reconstruction method and system, which relates to the technical field of power management. The method includes: deploying acquisition terminals to collect waveform data of each node in the power supply topology network; using a CNN - LSTM network to identify faults in the time - series waveform sequence; constructing power supply source tags for the power supply topology network; performing first - order reconstruction on nodes other than the fault node set, and outputting a first power supply topology network and a second power supply topology network; performing load - balancing analysis on the node paths corresponding to the hybrid power source tags, and outputting hybrid node reconstruction paths; generating a reconstructed power supply topology network. It solves the technical problems existing in the prior art, such as only being able to respond to abnormalities on the power supply side, being unable to accurately identify and isolate node - level faults, lacking adaptive adjustment to load changes and topological structures, resulting in unbalanced dual - power loads and poor power supply stability. It realizes dual - power load balancing and achieves the technical effect of improving the power supply stability of dual - power supplies.

Description

Technical Field

[0001] This application relates to the technical field of power management, and specifically relates to a dual-power dynamic reconstruction method and system. Background Art

[0002] The automatic switching of dual power supplies mainly switches the main power supply and the standby power supply automatically to avoid power outages caused by single power supply failures. Existing automatic transfer switches (ATS) and static transfer switches (STS) both perform switching control based on real-time monitoring of the power supply status. When the main power supply voltage is abnormal (such as power outage, voltage drop, frequency fluctuation, etc.), the load is switched to the standby power supply. However, the existing methods only respond to voltage / frequency abnormalities on a single power supply side and cannot accurately identify and isolate node-level faults (such as branch line short circuits, local overloads) in the power supply topology network; and lack the ability to adaptively adjust to dynamic load changes and topological structures, resulting in the standby power supply being in an overloaded, no-load or low-load state, making the stability of dual-power supply insufficient and the overall resource utilization rate low.

[0003] Therefore, in the current related technologies, there are technical problems such as only being able to respond to power supply side abnormalities, unable to accurately identify and isolate node-level faults, lacking adaptive adjustment to load changes and topological structures, resulting in unbalanced dual-power loads and poor power supply stability. Summary of the Invention

[0004] By providing a dual-power dynamic reconstruction method and system, this application solves the technical problems in the prior art, such as only being able to respond to power supply side abnormalities, unable to accurately identify and isolate node-level faults, lacking adaptive adjustment to load changes and topological structures, resulting in unbalanced dual-power loads and poor power supply stability, realizes balanced dual-power loads, and achieves the technical effect of improving the stability of dual-power supply.

[0005] The present application provides a dual-power dynamic reconstruction method, which includes: deploying acquisition terminals to collect waveform data of each node in the power supply topology network to generate a timing waveform sequence, where the power supply topology network includes a first power supply and a second power supply; using a CNN-LSTM network to perform fault identification on the timing waveform sequence to obtain a set of fault nodes; constructing power supply labels for each node in the power supply topology network, where the power supply labels include a first power supply label, a second power supply label, and a hybrid power supply label; performing first-order reconstruction on the nodes other than the set of fault nodes according to the first power supply label and the second power supply label to output a first power supply topology network based on the first power supply and a second power supply topology network based on the second power supply; taking load balancing of the first power supply and the second power supply as the second-order reconstruction target, performing load balancing analysis on the node paths corresponding to the hybrid power supply label to output a hybrid node reconstruction path; updating the first power supply topology network and the second power supply topology network according to the hybrid node reconstruction path to generate a reconstructed power supply topology network.

[0006] In a possible implementation manner, the dual-power dynamic reconstruction method further performs the following processing: using a CNN-LSTM network to perform feature recognition on the timing waveform sequence to output CNN short-period features and LSTM long-period features, where the CNN short-period features include local voltage changes and current peak changes, and the LSTM long-period features include timing oscillations, drift data, and periodic changes; analyzing the states of each node in the power supply topology network according to the CNN short-period features and the LSTM long-period features, and marking the nodes that do not meet the preset feature threshold as fault nodes to obtain a set of fault nodes.

[0007] In a possible implementation manner, the dual-power dynamic reconstruction method further performs the following processing: the hybrid power supply label includes a first type of hybrid power supply label and a second type of hybrid power supply label, where the first type of hybrid power supply label is a label that requires simultaneous power supply from the first power supply and the second power supply, and the second type of hybrid power supply label is a label for switching power supply between the first power supply and the second power supply; performing load balancing analysis on the node paths corresponding to the hybrid power supply label, and respectively performing load balancing analysis on the node paths corresponding to the first type of hybrid power supply label and the second type of hybrid power supply label to output a hybrid node reconstruction path.

[0008] In a possible implementation manner, the dual-power dynamic reconstruction method further performs the following processing: for the nodes corresponding to the first type of hybrid power supply tags, construct a combined power supply path diagram of the first power supply and the second power supply; identify the real-time power demands of each node in the combined power supply path diagram; perform power supply ratio analysis using a ratio model based on the real-time power demands, calculate the load difference between the first power supply and the second power supply under different ratios, and take minimizing the load difference as the load balancing target to output the reconstruction path of the first type of hybrid nodes, thereby obtaining the reconstruction path of the hybrid nodes.

[0009] In a possible implementation manner, the dual-power dynamic reconstruction method further performs the following processing: for the nodes corresponding to the second type of hybrid power supply tags, respectively evaluate the line impedance, network structure complexity, and source load status of their connections to the first power supply and the second power supply to obtain a first load index and a second load index; select the smaller one of the first load index and the second load index as the node belonging path to output the reconstruction path of the second type of hybrid nodes; integrate the reconstruction path of the first type of hybrid nodes and the reconstruction path of the second type of hybrid nodes to obtain the reconstruction path of the hybrid nodes.

[0010] In a possible implementation manner, the dual-power dynamic reconstruction method further performs the following processing: update the first power supply topology network according to the reconstruction path of the hybrid nodes to obtain a first updated power supply topology network, and update the second power supply topology network according to the reconstruction path of the hybrid nodes to obtain a second updated power supply topology network; couple the first updated power supply topology network and the second updated power supply topology network to generate a reconstructed power supply topology network.

[0011] In a possible implementation manner, the dual-power dynamic reconstruction method further performs the following processing: perform load stability simulation, overload risk assessment, and network voltage quality detection on the reconstructed power supply topology network; if any one of the load stability simulation, overload risk assessment, and network voltage quality does not meet the corresponding threshold, roll back to the previous stage of the reconstruction state and update the reconstruction path of the hybrid nodes.

[0012] The present application also provides a dual - power dynamic reconstruction system, which includes: a timing waveform sequence generation module for deploying a collection terminal to collect waveform data of each node in a power supply topology network, generating a timing waveform sequence, where the power supply topology network includes a first power supply and a second power supply; a fault identification module for using a CNN - LSTM network to identify faults in the timing waveform sequence to obtain a set of fault nodes; a power supply label construction module for constructing power supply labels for each node in the power supply topology network, where the power supply labels include a first power supply label, a second power supply label, and a hybrid power supply label; a power supply topology network output module for performing a first - order reconstruction on the nodes except the set of fault nodes according to the first power supply label and the second power supply label, and outputting a first power supply topology network based on the first power supply and a second power supply topology network based on the second power supply; a load - balancing analysis module for taking the load balancing of the first power supply and the second power supply as the second - order reconstruction target, performing a load - balancing analysis on the node paths corresponding to the hybrid power supply label, and outputting a reconstructed path for hybrid nodes; a reconstructed power supply topology network generation module for updating the first power supply topology network and the second power supply topology network according to the reconstructed path for hybrid nodes to generate a reconstructed power supply topology network.

[0013] It is intended to deploy a collection terminal to collect waveform data of each node in a power supply topology network through a dual - power dynamic reconstruction method and system proposed in the present application; use a CNN - LSTM network to identify faults in a timing waveform sequence; construct power supply labels for a power supply topology network; perform a first - order reconstruction on the nodes except the set of fault nodes to output a first power supply topology network and a second power supply topology network; perform a load - balancing analysis on the node paths corresponding to the hybrid power supply label to output a reconstructed path for hybrid nodes; and generate a reconstructed power supply topology network. This solves the technical problems existing in the prior art, such as being able to only respond to abnormalities on the power supply side, being unable to accurately identify and isolate node - level faults, lacking adaptive adjustment to load changes and topological structures, resulting in unbalanced loads of dual - power supplies and poor power supply stability. It realizes the load balancing of dual - power supplies and achieves the technical effect of improving the power supply stability of dual - power supplies. BRIEF DESCRIPTION OF THE DRAWINGS

[0014] In order to more clearly illustrate the technical solutions of the embodiments of the present disclosure, the drawings of the embodiments of the present disclosure will be briefly introduced below. Flowcharts are used in this application to illustrate the operations performed by the system according to the embodiments of the application. It should be understood that the operations in the front or below do not necessarily need to be executed precisely in sequence. On the contrary, according to needs, they can be executed in reverse order or simultaneously. At the same time, other operations can also be added to these processes, or one or several operations can be removed from these processes.

[0015] Figure 1Schematic diagram of a dual - power dynamic reconfiguration method provided by an embodiment of the present application.

[0016] Figure 2 Schematic diagram of a dual - power dynamic reconfiguration system structure provided by an embodiment of the present application.

[0017] Explanation of reference numerals: Timing waveform sequence generation module 10, fault identification module 20, power supply label construction module 30, power supply topology network output module 40, load - balancing analysis module 50, reconfigured power supply topology network generation module 60. Detailed implementation manners

[0018] The above description is only an overview of the technical solution of the present application. In order to understand the technical means of the present application more clearly, it can be implemented according to the content of the specification. And in order to make the above - mentioned and other purposes, features and advantages of the present application more obvious and understandable, the following specifically gives the detailed implementation manners of the present application.

[0019] In order to make the purpose, technical solution and advantages of the present application clearer, the present application will be further described in detail below with reference to the accompanying drawings. The described embodiments should not be regarded as limitations of the present application. All other embodiments obtained by those of ordinary skill in the art without creative efforts fall within the protection scope of the present application.

[0020] In the following description, reference is made to "some embodiments", which describe a subset of all possible embodiments. However, it can be understood that "some embodiments" can be the same subset or different subsets of all possible embodiments and can be combined with each other without conflict. The terms "first" and "second" involved are only used to distinguish similar objects and do not represent a specific order for the objects. The terms "including" and "having" and any variations thereof are intended to cover non - exclusive inclusion. For example, a process, method, system, product or server that includes a series of steps or units does not necessarily have to be limited to those steps or units clearly listed, but may include other steps or modules that are not clearly listed or are inherent to these processes, methods, products or devices. Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by those skilled in the technical field to which the present application belongs. The terms used herein are only for the purpose of describing the embodiments of the present application.

[0021] An embodiment of the present application provides a dual - power dynamic reconfiguration method, as Figure 1 shown. The method includes:

[0022] Step S100, deploying a collection terminal to collect waveform data of each node of the power supply topology network, generating a timing waveform sequence, where the power supply topology network includes a first power supply and a second power supply.

[0023] Preferably, acquisition terminals, including high-precision voltage and current sensors, etc., are deployed at various key positions of the power supply topology network, such as the power output end, the intersection of transmission lines, the incoming and outgoing line ends of the substation, important power distribution equipment, and the access points of large power-consuming equipment, etc., to monitor the power data of each node of the power supply topology network in real time. Specifically, the sensors of the acquisition terminal are used to monitor and record the voltage and current data of the nodes of the power supply topology network. Since the voltage and current of the power supply system are periodic signals that change with time, their waveforms contain rich information, such as amplitude, frequency, phase, and harmonic components, etc. The acquisition terminal samples the signal at a certain sampling frequency, converts the continuous analog signal into a discrete digital signal, and arranges them in chronological order to form a time-series waveform sequence, which contains the electrical signal characteristic information of the corresponding node at each time point, such as the amplitude, frequency, phase, etc. of the voltage, reflecting the changes of voltage, current, etc. with time at the nodes of the power supply topology network. When a fault (such as a short circuit, overload, etc.) occurs at a certain node, the corresponding time-series waveform sequence will change accordingly. By analyzing the time-series waveform sequence, the power operation status of each node in the power supply topology network can be understood, and the fault node can be identified. Among them, the power supply topology network includes a first power supply and a second power supply, and the first power supply and the second power supply are independent of each other and jointly supply power to each node in the network to improve the reliability and stability of power supply.

[0024] Step S200, use a CNN-LSTM network to perform fault identification on the time-series waveform sequence to obtain a set of fault nodes.

[0025] Preferably, the CNN-LSTM network includes a CNN (Convolutional Neural Network) and an LSTM (Long Short-Term Memory Network). Among them, the convolutional neural network can automatically extract local features in the data. When processing a time series waveform sequence, it can slide a convolutional kernel over the data to capture local patterns and features in the waveform, such as the peak, valley, and slope changes of the waveform. The long short-term memory network is used to process long-term dependencies in the sequence data and can effectively capture the data associations at different time points in the time series waveform sequence, such as the change trend and periodicity of voltage or current over a period of time. Before inputting the time series waveform sequence into the CNN-LSTM network, the data is preprocessed, including smoothing the data to remove noise interference and make the waveform smoother, and normalizing the data to map the value range of the data to a specific interval to accelerate the convergence speed of the model. Then, the preprocessed time series waveform sequence is input into the CNN-LSTM network, and the CNN-LSTM network is used to identify faults in the time series waveform sequence. Specifically, first, the CNN layer performs a convolutional operation on the waveform sequence to extract local features of the waveform, including key information in different local regions of the waveform. Then, the LSTM layer receives the features extracted by the CNN and further processes them, considering the time order of the sequence to capture the long-term dependencies between the features. The feature vectors obtained by feature extraction are then input into the fully connected layer for classification, that is, the input waveform sequence is classified into different categories, such as normal state, different types of fault states, etc., so as to establish the corresponding relationship between the waveform features and the fault nodes, thereby determining the specific nodes where the faults occur, and finally quickly and accurately obtaining the fault node set.

[0026] Further, step S200 further includes step S210 of using a CNN-LSTM network to perform feature recognition on the time series waveform sequence and outputting CNN short-period features and LSTM long-period features. The CNN short-period features include local voltage changes and current peak changes, and the LSTM long-period features include time series oscillations, drift data, and periodic changes. Step S220 is to analyze the states of each node in the power supply topology network according to the CNN short-period features and the LSTM long-period features, mark the nodes that do not meet the preset feature threshold as fault nodes, and obtain the fault node set.

[0027] Preferably, the time - series waveform sequence collected from the power supply topology network is used as the input of the CNN - LSTM network. The CNN processes the time - series waveform sequence to capture features within a short period, including local voltage changes and current peak changes. Among them, the local voltage change reflects the voltage fluctuation of a node at a certain moment, and the current peak change can reflect the extreme value change of the current within a short time, which helps to judge whether there is an instantaneous fault or abnormality at the node; the LSTM extracts long - period features from the time - series waveform sequence, including time - series oscillation, drift data, and periodic changes. Among them, the time - series oscillation reflects the periodic fluctuation of the signal over a long time, the drift data reflects the trend that the signal gradually deviates from the normal range over time, and the periodic change can help to discover the periodic regular changes of the signal on a long - time scale, which helps to detect faults that develop relatively slowly or have periodicity.

[0028] Preferably, according to the normal operating state and historical data of the power supply topology network, preset feature thresholds are set for the short - period features of the CNN and the long - period features of the LSTM respectively, which are used to judge whether a node is normal. There are normal ranges for the voltage fluctuation range, current peak range, time - series oscillation amplitude, drift speed, and periodic change of the node. Then, the states of each node in the power supply topology network are analyzed, that is, the short - period features of the CNN and the long - period features of the LSTM extracted are respectively compared with the preset feature thresholds to analyze the states of each node in the power supply topology network. For each node, check whether features such as its local voltage change, current peak change, time - series oscillation, drift data, and periodic change are within the normal range. If the short - period features of the CNN or the long - period features of the LSTM of a certain node do not meet the preset feature thresholds, mark this node as a faulty node. Finally, all the nodes marked as faulty nodes are integrated to form a faulty node set.

[0029] Step S300, construct power supply labels for each node of the power supply topology network, where the power supply labels include a first power supply label, a second power supply label, and a hybrid power supply label.

[0030] Preferably, obtain the locations of the first power source and the second power source, the connection mode of the transmission lines, the distribution of transformers and switchgear, etc. through the power supply network topology. Use graph theory to establish a mathematical model of the power supply topology network, abstract each node (such as power sources, transformers, loads, etc.) as vertices in the graph, and abstract the transmission lines as edges in the graph; analyze the connections between each node and the first power source and the second power source, including checking the direction of the transmission lines, the status of the switches, etc., to determine the electrical connections existing between the nodes and the power sources; then conduct simulation tests, energize the first power source and the second power source respectively, install sensors to monitor the electrical parameters such as voltage and current of the nodes in real time, and judge the power supply status of the nodes according to the monitoring data. For example, if the node voltage is consistent with the output voltage of the first power source and has nothing to do with the output voltage of the second power source, it is judged that the node is powered only by the first power source; furthermore, establish power supply labels for each node, including the first power source label, the second power source label, and the hybrid power source label. Among them, when a certain node is powered only by the first power source, label this node with the first power source label, indicating that the power supply of this node completely depends on the first power source. If the first power source fails or the power supply is interrupted, this node will lose power supply; similarly, if a certain node is powered only by the second power source, label this node with the second power source label. If the second power source fails or the power supply is interrupted, this node will lose power supply; if a certain node can be powered by both the first power source and the second power source, or switch the power supply between the two power sources, label this node with the hybrid power source label. When one power source fails, the other power source can continue to supply power to the node, or both power sources can supply power simultaneously during the peak power consumption period, enhancing the reliability and stability of the node power supply.

[0031] Step S400, perform first-order reconstruction on the nodes other than the set of faulty nodes according to the first power source label and the second power source label, and output a first power supply topology network based on the first power source and a second power supply topology network based on the second power source.

[0032] Preferably, the nodes in the faulty node set are excluded from the power supply topology network, and the remaining normal nodes are divided according to the first power supply label and the second power supply label. Specifically, the nodes with the first power supply label are grouped into one group, which are normally powered only by the first power supply; the nodes with the second power supply label are grouped into another group, which are powered only by the second power supply; then a first-order reconstruction is performed, that is, with the first power supply as the core, the nodes with the first power supply label are reconnected to form a first power supply topology network based on the first power supply, and it is ensured that power can be smoothly transmitted from the first power supply to each node. For example, some lines may be in a complex connection state due to the existence of faulty nodes, and now they are adjusted to a more direct and effective connection method to reduce power transmission losses; similarly, with the second power supply as the core, the nodes with the second power supply label are reconnected to construct a second power supply topology network based on the second power supply, including all normal nodes with the second power supply label.

[0033] Step S500 , taking balancing the loads of the first power source and the second power source as a second-order reconstruction target, performing load balancing analysis on the node paths corresponding to the hybrid power source labels, and outputting hybrid node reconstruction paths.

[0034] Preferably, after completing the first-order reconstruction based on the first power supply and the second power supply, two relatively independent power supply topology networks (the first power supply topology network and the second power supply topology network) are obtained. At this time, the node corresponding to the hybrid power supply label may be unbalancedly powered by the first power supply and the second power supply. The load balancing of the first power supply and the second power supply is taken as the second-order reconstruction goal, and the node path corresponding to the hybrid power supply label is subjected to load balancing analysis to make the load distribution of the two power supplies more reasonable, so as to avoid the situation where one power supply is overloaded and the other power supply is underloaded. Specifically, for the node path corresponding to the hybrid power supply label, data related to the load is collected, such as the power demand of each node, the current or power currently passing through each path, and other information, and these data are analyzed to evaluate the current Based on the current load distribution, determine which paths have heavier loads and which paths have lighter loads, then adjust the state of some switches to change the direction of current flow, so that the load is transferred from the heavier path to the lighter path to achieve load balancing; finally determine the adjustment plan for the node path corresponding to the hybrid power supply label, that is, the hybrid node reconstruction path, which can adjust some nodes that were originally borne by the first power supply to bear a part of the load by the second power supply. By replanning the power supply path, the loads of the two power supplies are within a reasonable range, thereby making the power supply of the first power supply and the second power supply to the hybrid power supply node more balanced, reducing the risk of equipment damage and unstable power supply due to load imbalance, and improving the performance and reliability of the entire power supply network.

[0035] Further, step S500 further includes step S510. The hybrid power supply tags include first - type hybrid power supply tags and second - type hybrid power supply tags. Among them, the first - type hybrid power supply tags are tags that require both the first power supply and the second power supply to supply power simultaneously, and the second - type hybrid power supply tags are tags that are powered by switching between the first power supply and the second power supply; step S520, perform load - balancing analysis on the node paths corresponding to the hybrid power supply tags, perform load - balancing analysis on the node paths corresponding to the first - type hybrid power supply tags and the second - type hybrid power supply tags respectively, and output the reconstructed hybrid node paths.

[0036] Preferably, the hybrid power supply tags include first - type hybrid power supply tags and second - type hybrid power supply tags. Among them, the first - type hybrid power supply tags are tags that require both the first power supply and the second power supply to supply power simultaneously. During the power - supply process, the two power supplies work together to provide power for the nodes; the second - type hybrid power supply tags are tags that are powered by switching between the first power supply and the second power supply. That is, in different situations, the nodes select one of the power supplies to supply power according to different conditions or strategies. For example, when the first power supply fails, has too high a load, or other abnormal situations occur, the nodes will automatically switch to the second power supply for power supply to ensure their normal operation. Then, perform load - balancing analysis on the node paths corresponding to the hybrid power supply tags. Specifically, clarify the power - demand characteristics of the nodes with the first - type hybrid power supply tags, including the power - demand ratio of the first power supply and the second power supply and voltage requirements under different working states, and then analyze the load conditions when the current first power supply and the second power supply supply power to these nodes. For example, measure the actual current and power obtained by each node from the first power supply and the second power supply, understand the load magnitudes borne by each power supply when supplying power to these nodes, and then, according to the power demands of the nodes and the current load conditions, find an optimized power - supply path, which may include adjusting the connection mode of the transmission lines, changing the states of certain switches, etc., to make the loads of the first power supply and the second power supply when supplying power to these nodes more balanced.

[0037] Preferably, for the nodes with the second type of hybrid power tags, clarify the power switching conditions. For example, determine the switching timing based on the load size or voltage stability, and then analyze whether the current power switching strategy is reasonable, as well as the time ratio and load conditions of the first power source and the second power source supplying power to these nodes under different working states, including checking whether the power switching is timely during peak load and low load, and whether there is a situation where a certain power source undertakes too much load for a long time; then, according to the evaluation results, optimize the power switching strategy and power supply path. For example, if a certain power source undertakes too heavy a load in some cases, adjust the switching conditions to enable another power source to participate in power supply earlier, or optimize the connection path between the nodes and the power sources to reduce losses and imbalance during power supply. For example, when the first power source has an overload risk at high load, switch some loads to the second power source in advance, or adjust the line to enable the second power source to supply power to these nodes more efficiently. Finally, integrate the results of the load balancing analysis of the node path corresponding to the first type of hybrid power tag and the second type of hybrid power tag to generate a reconstructed path for hybrid nodes, including how the nodes corresponding to each hybrid power tag are connected to the first power source and the second power source, as well as the power supply methods and switching strategies of the power sources under different working states, so as to achieve load balancing between the first power source and the second power source in the entire power supply system and improve the reliability and efficiency of power supply.

[0038] Further, step S520 further includes step S521, for the nodes corresponding to the first type of hybrid power tags, construct a combined power supply path diagram of the first power source and the second power source; step S522, identify the real-time power demands of each node in the combined power supply path diagram; step S523, based on the real-time power demands, use a ratio model to conduct power supply ratio analysis, calculate the load difference between the first power source and the second power source under different ratios, and take minimizing the load difference as the load balancing target to output a reconstructed path for the first type of hybrid nodes, thereby obtaining the reconstructed path for hybrid nodes.

[0039] Preferably, for the nodes corresponding to the first type of hybrid power tags, construct a combined power supply path diagram to show how the first power source and the second power source are connected to these nodes through transmission lines, switchgear, etc., as well as the connection relationships between the nodes; then, through the power sensors installed at the nodes, monitor and identify the power demand conditions of each node in the diagram in real time, obtain the power data obtained by each node from the first power source and the second power source at different times, and further determine the real-time power demands, that is, a load is simultaneously supplied with power by two power sources in proportion.

[0040] Preferably, using the obtained real-time power demand data, a ratio model is adopted for power supply ratio analysis, that is, according to the capacity of the power source, the loss of the transmission line, the power demand characteristics of the node, etc., the load conditions of the first power source and the second power source under different power supply ratios are calculated. For example, by changing the power supply ratios of the first power source and the second power source to a certain node, the load changes of the two power sources are observed, so as to find different power supply ratio combinations and their corresponding load states; then the load difference between the two power sources under each combination is calculated to evaluate the load balance degree of the two power sources under different power supply ratios. The smaller the load difference, the more balanced the load distribution of the two power sources; the larger the load difference, the more uneven the load distribution. Then, minimizing the load difference between the first power source and the second power source is taken as the goal of load balance. Among different power supply ratio combinations and their corresponding load differences, the power supply ratio combination with the smallest load difference is found, and then the original combined power supply path is adjusted and optimized to output the first type of hybrid node reconstruction path, improving the load balance degree and operation efficiency of the power supply system.

[0041] Further, step S520 further includes step S524. For the nodes corresponding to the second type of hybrid power source label, the line impedance, network structure complexity, and source load state connected to the first power source and the second power source are respectively evaluated to obtain a first load index and a second load index; step S525, selecting the smaller one of the first load index and the second load index as the node attribution path, and outputting the second type of hybrid node reconstruction path; step S526, combining the first type of hybrid node reconstruction path and the second type of hybrid node reconstruction path to obtain a hybrid node reconstruction path.

[0042] Preferably, for the nodes corresponding to the second type of hybrid power supply tags (i.e., the nodes powered by switching between the first power supply and the second power supply), the line impedance, network structure complexity, and source load status of the lines connecting them to the first power supply and the second power supply are respectively evaluated. Specifically, the line impedance affects the loss during power transmission. If the line impedance is large, more electrical energy loss will occur when transmitting the same power, increasing the power supply cost. Evaluating the line impedance of the lines connected to the first power supply and the second power supply is to evaluate the loss situation during the process of power transmission from the power supply to the nodes; the network structure complexity includes the connection mode of the lines, the number and layout of intermediate devices (such as switches, transformers, etc.). A complex network structure may increase the difficulty of maintenance, the time and cost of troubleshooting, and may also affect the reliability of power supply. Evaluating the network structure complexity of the lines connected to the two power supplies to determine which connection mode is more convenient for management and maintenance; the source load status refers to the load conditions of the first power supply and the second power supply themselves. If a power supply is already in a high-load state, connecting more loads may cause it to be overloaded, affecting the power supply quality and even causing faults. Evaluating the source load status of the lines connected to the first power supply and the second power supply helps to reasonably allocate the nodes; furthermore, the first load index corresponding to the connection to the first power supply and the second load index corresponding to the connection to the second power supply are obtained, and the smaller one of them is selected as the attribution path of the node, indicating that the comprehensive situation of connecting to this power supply is better, such as lower loss, simpler network, and less impact on the load of the power supply, etc. Then, the reconstruction path of the second type of hybrid nodes is output, that is, it is determined whether these nodes should be connected to the first power supply or the second power supply. Finally, the reconstruction path of the first type of hybrid nodes (for the nodes that require simultaneous power supply from the first power supply and the second power supply) and the reconstruction path of the second type of hybrid nodes (for the nodes that can be powered by switching between the first power supply and the second power supply) are integrated to obtain a complete reconstruction path of the hybrid nodes, which helps to achieve more reasonable power distribution and load balancing, and improve the reliability and efficiency of the power supply system.

[0043] Step S600, update the first power supply topology network and the second power supply topology network according to the hybrid node reconstruction path to generate a reconstructed power supply topology network.

[0044] Preferably, the first power supply topology network and the second power supply topology network are updated according to the hybrid node reconstruction path. Specifically, the hybrid nodes are connected to the first power supply topology network and the second power supply topology network according to the planning of the hybrid node reconstruction path. For example, the connection relationship of some transmission lines is changed or the states of some switches are adjusted so that the hybrid nodes can obtain a more balanced power supply from the first power source and the second power source. Then, other power supply nodes and paths in the first power supply topology network and the second power supply topology network are adjusted. For example, the power transmission paths are reallocated and the current distribution on each line is optimized to adapt to the changes after the hybrid nodes are connected, ensuring the power transmission efficiency and stability of the entire network. Through the update of the first power supply topology network and the second power supply topology network, a reconstructed power supply topology network is finally generated, which includes both the parts independently powered by the first power source and the second power source separated during the first-order reconstruction, and realizes the load balance of the hybrid power nodes through the hybrid node reconstruction path, making the entire power supply system more reasonable, efficient and reliable, reducing the situations of power source overload or load imbalance, thereby improving the stability and fault resistance of the power supply system.

[0045] Further, step S600 further includes step S610 of updating the first power supply topology network according to the hybrid node reconstruction path to obtain a first updated power supply topology network, and updating the second power supply topology network according to the hybrid node reconstruction path to obtain a second updated power supply topology network; step S620 of coupling the first updated power supply topology network and the second updated power supply topology network to generate a reconstructed power supply topology network.

[0046] Preferably, according to the node - power connection relationship of the reconstructed path of the hybrid node, the first power supply topology network is updated. For example, if the reconstructed path of the hybrid node indicates that a node originally powered by the first power supply alone now needs to be connected to the second power supply to achieve a better power supply effect, a connection line between the node and the second power supply is added to the first power supply topology network, and relevant line parameters, node attributes and other information are adjusted to obtain the first updated power supply network topology. Similarly, according to the reconstructed path of the hybrid node, the second power supply topology network is updated. For example, if the power supply mode of a node is switched from the second power supply to the first power supply, or its connection mode with the second power supply needs to be changed, corresponding adjustments are made in the second power supply topology network, including deleting unnecessary connections, adding new connections and updating relevant parameters of nodes and lines, to obtain the second updated power supply topology network. Then, the first updated power supply topology network and the second updated power supply topology network are coupled, that is, the first updated power supply topology network and the second updated power supply topology network are integrated into a reconstructed power supply topology network, including integrating nodes, lines and connection relationships in the two networks. For nodes that exist in both networks, their attributes and connection relationships are unified according to the actual situation; for different parts, reasonable fusion is carried out so that the reconstructed power supply topology network can comprehensively reflect the optimization effect brought by the reconstructed path of the hybrid node, forming a more efficient and reliable power supply network structure to meet the operation requirements of the entire power supply system.

[0047] Further, step S600 further includes step S630 of performing load stability simulation, overload risk assessment and network voltage quality detection on the reconstructed power supply topology network; step S640, if one of the load stability simulation, overload risk assessment and network voltage quality does not meet the corresponding threshold, roll back to the previous - stage reconstruction state and update the reconstructed path of the hybrid node.

[0048] Preferably, load stability simulation, overload risk assessment, and network voltage quality detection are performed on the reconstructed power supply topology network. Specifically, for load stability simulation of the reconstructed power supply topology network, that is, through simulation software under different load scenarios (such as peak load, valley load, etc.), the load changes of each node and line in the network are simulated. By observing whether the load distribution in the network is uniform and whether there are situations where the load fluctuations of certain nodes or lines are too large, the load stability of the network is evaluated; analyze the carrying capacity of each power source, line, and device in the reconstructed power supply topology network, combined with the current load situation and the predicted future load change trend, evaluate whether there is an overload risk in the network. By calculating the load rate of each part (the ratio of the actual load to the rated load), if the load rate is too high, it indicates an overload risk; detect the voltage levels of each node in the reconstructed power supply topology network, including indicators such as voltage amplitude, voltage fluctuation, and voltage harmonics. Excessive or too low voltage, large voltage fluctuations, and severe voltage harmonics may all affect the performance and lifespan of equipment.

[0049] Preferably, corresponding thresholds are set for the indicators of load stability simulation, overload risk assessment, and network voltage quality detection according to the equipment operation requirements and power supply network standards. The results obtained from the simulation and detection are compared with the corresponding thresholds to determine whether the requirements are met. If one or more indicators do not meet the corresponding thresholds in the load stability simulation, overload risk assessment, and network voltage quality detection, it indicates that there are still problems with the reconstructed power supply topology network and adjustments are needed. The network is rolled back to the previous stage of the reconstruction state, that is, restored to the state after the previous reconstruction. Then, according to the evaluation results, analyze the reasons for not meeting the thresholds and update the reconstruction path of the hybrid nodes. For example, if the load stability in a certain area does not meet the requirements, it may be that the power supply distribution of the hybrid nodes is unreasonable. Then, re-adjust the connection method and power supply ratio of the hybrid nodes, update the reconstruction path of the hybrid nodes, so as to perform reconstruction and evaluation again until all indicators meet the thresholds, thereby continuously optimizing the reconstructed power supply topology network, improving its stability, reliability, and voltage quality, and ensuring the safe and efficient operation of the power supply system.

[0050] In the above text, with reference to Figure 1 A dual - power dynamic reconstruction method according to an embodiment of the present invention is described in detail. Next, with reference to Figure 2 A dual - power dynamic reconstruction system according to an embodiment of the present invention will be described.

[0051] A dual-power dynamic reconfiguration system according to an embodiment of the present invention is used to solve the technical problems existing in the prior art, such as only being able to respond to power supply side abnormalities, unable to accurately identify isolation node-level faults, lacking adaptive adjustment to load changes and topological structures, resulting in unbalanced dual-power loads and poor power supply stability. It realizes balanced dual-power loads and achieves the technical effect of improving the power supply stability of dual-power. As Figure 2 shown, a dual-power dynamic reconfiguration system includes: a timing waveform sequence generation module 10, a fault identification module 20, a power supply label construction module 30, a power supply topology network output module 40, a load balance analysis module 50, and a reconstructed power supply topology network generation module 60.

[0052] The timing waveform sequence generation module 10 is used to deploy acquisition terminals to collect waveform data of each node of the power supply topology network and generate a timing waveform sequence. The power supply topology network includes a first power supply and a second power supply. The fault identification module 20 is used to use a CNN-LSTM network to identify faults in the timing waveform sequence and obtain a set of fault nodes. The power supply label construction module 30 is used to construct power supply labels for each node of the power supply topology network. The power supply labels include a first power supply label, a second power supply label, and a hybrid power supply label. The power supply topology network output module 40 is used to perform first-order reconstruction on the nodes other than the set of fault nodes according to the first power supply label and the second power supply label, and output a first power supply topology network based on the first power supply and a second power supply topology network based on the second power supply. The load balance analysis module 50 is used to take balancing the loads of the first power supply and the second power supply as the second-order reconstruction target, perform load balance analysis on the node paths corresponding to the hybrid power supply label, and output a hybrid node reconstruction path. The reconstructed power supply topology network generation module 60 is used to update the first power supply topology network and the second power supply topology network according to the hybrid node reconstruction path to generate a reconstructed power supply topology network.

[0053] Next, the specific configuration of the fault identification module 20 will be described in detail. The fault identification module 20 further includes: using a CNN-LSTM network to perform feature identification on the timing waveform sequence and outputting CNN short-cycle features and LSTM long-cycle features. The CNN short-cycle features include local voltage changes and current peak changes. The LSTM long-cycle features include timing oscillations, drift data, and periodic changes. Analyze the states of each node of the power supply topology network according to the CNN short-cycle features and the LSTM long-cycle features, mark the nodes that do not meet the preset feature thresholds as fault nodes, and obtain a set of fault nodes.

[0054] Next, the specific configuration of the load balancing analysis module 50 will be described in detail. The load balancing analysis module 50 further includes: The hybrid power supply tags include a first type of hybrid power supply tag and a second type of hybrid power supply tag. Among them, the first type of hybrid power supply tag is a tag that requires both the first power supply and the second power supply to supply power simultaneously, and the second type of hybrid power supply tag is a tag that is powered by switching between the first power supply and the second power supply; perform load balancing analysis on the node paths corresponding to the hybrid power supply tags, respectively perform load balancing analysis on the node paths corresponding to the first type of hybrid power supply tag and the second type of hybrid power supply tag, and output a reconstructed hybrid node path.

[0055] Next, the specific configuration of the load balancing analysis module 50 will be further described in detail. The load balancing analysis module 50 further includes: For the nodes corresponding to the first type of hybrid power supply tag, construct a combined power supply path diagram of the first power supply and the second power supply; identify the real-time power requirements of each node in the combined power supply path diagram; based on the real-time power requirements, use a ratio model to perform power supply ratio analysis, calculate the load difference between the first power supply and the second power supply under different ratios, and take minimizing the load difference as the load balancing target, output the reconstructed path of the first type of hybrid node, and obtain the reconstructed hybrid node path.

[0056] Next, the specific configuration of the load balancing analysis module 50 will be further described in detail. The load balancing analysis module 50 further includes: For the nodes corresponding to the second type of hybrid power supply tag, respectively evaluate the line impedance, network structure complexity, and source load status of their connections to the first power supply and the second power supply, and obtain a first load index and a second load index; select the smaller one of the first load index and the second load index as the node belonging path, output the reconstructed path of the second type of hybrid node; combine the reconstructed path of the first type of hybrid node and the reconstructed path of the second type of hybrid node to obtain the reconstructed hybrid node path.

[0057] Next, the specific configuration of the reconstructed power supply topology network generation module 60 will be described in detail. The reconstructed power supply topology network generation module 60 further includes: Update the first power supply topology network according to the reconstructed hybrid node path to obtain a first updated power supply topology network, and update the second power supply topology network according to the reconstructed hybrid node path to obtain a second updated power supply topology network; couple the first updated power supply topology network and the second updated power supply topology network to generate a reconstructed power supply topology network.

[0058] Next, the specific configuration of the reconstructed power supply topology network generation module 60 will be described in detail. The reconstructed power supply topology network generation module 60 further includes: performing load stability simulation, overload risk assessment, and network voltage quality detection on the reconstructed power supply topology network; if any one of the load stability simulation, overload risk assessment, and network voltage quality does not meet the corresponding threshold, reverting to the previous stage of the reconstruction state and updating the hybrid node reconstruction path.

[0059] A dual-power dynamic reconstruction system provided by an embodiment of the present invention can execute a dual-power dynamic reconstruction method provided by any embodiment of the present invention, and has functional modules and beneficial effects corresponding to the execution of the method.

[0060] Although this application makes various references to certain modules in the system according to the embodiments of this application, however, any number of different modules can be used and run on the user terminal and / or server. The included individual units and modules are only divided according to functional logic, but are not limited to the above division, as long as the corresponding functions can be achieved; in addition, the specific names of the functional units are only for the convenience of mutual distinction and do not limit the protection scope of the present invention.

[0061] The above specific implementation manners do not constitute a limitation to the protection scope of this application. Those skilled in the art should understand that various modifications, combinations, and substitutions can be made according to design requirements and other factors. Any modifications, equivalent substitutions, and improvements made within the spirit and principle of this application shall be included within the protection scope of this application.

Claims

1. A dual-power dynamic reconfiguration method, characterized in that The method includes: Deploying acquisition terminals to collect waveform data of each node in the power supply topology network, generating a time-series waveform sequence, where the power supply topology network includes a first power source and a second power source; Using a CNN-LSTM network to perform fault identification on the time-series waveform sequence to obtain a set of fault nodes; Constructing power supply tags for each node in the power supply topology network, where the power supply tags include a first power supply tag, a second power supply tag, and a hybrid power supply tag; Performing first-order reconstruction on the nodes other than the set of fault nodes according to the first power supply tag and the second power supply tag, and outputting a first power supply topology network based on the first power source and a second power supply topology network based on the second power source; Taking balancing the loads of the first power source and the second power source as the second-order reconstruction target, performing load balancing analysis on the node paths corresponding to the hybrid power supply tag, and outputting a reconstructed path for hybrid nodes; Updating the first power supply topology network and the second power supply topology network according to the reconstructed path for hybrid nodes to generate a reconstructed power supply topology network; The hybrid power supply tag includes a first type of hybrid power supply tag and a second type of hybrid power supply tag, where the first type of hybrid power supply tag is a tag that requires simultaneous power supply from the first power source and the second power source, and the second type of hybrid power supply tag is a tag for switching power supply between the first power source and the second power source; Performing load balancing analysis on the node paths corresponding to the hybrid power supply tag, respectively performing load balancing analysis on the node paths corresponding to the first type of hybrid power supply tag and the second type of hybrid power supply tag, and outputting a reconstructed path for hybrid nodes; Among them, the method for performing load balancing analysis on the node paths corresponding to the first type of hybrid power supply tag includes: For the nodes corresponding to the first type of hybrid power supply tag, constructing a joint power supply path diagram of the first power source and the second power source; Identifying the real-time power demand of each node in the joint power supply path diagram; Performing power supply ratio analysis using a ratio model based on the real-time power demand, calculating the load difference between the first power source and the second power source under different ratios, taking minimizing the load difference as the load balancing target, outputting a reconstructed path for the first type of hybrid nodes, and obtaining a reconstructed path for hybrid nodes; Among them, the method for performing load balancing analysis on the node paths corresponding to the second type of hybrid power supply tag includes: For the nodes corresponding to the second type of hybrid power supply tag, respectively evaluating the line impedance, network structure complexity, and source load status of their connections to the first power source and the second power source to obtain a first load index and a second load index; Selecting the smaller one of the first load index and the second load index as the node belonging path, and outputting a reconstructed path for the second type of hybrid nodes; Combining the reconstructed path for the first type of hybrid nodes and the reconstructed path for the second type of hybrid nodes to obtain a reconstructed path for hybrid nodes.

2. The dual-power dynamic reconfiguration method according to claim 1, characterized in that Using a CNN-LSTM network to perform fault identification on the time-series waveform sequence to obtain a set of fault nodes, the method includes: Use a CNN-LSTM network to perform feature recognition on the timing waveform sequence, and output CNN short-cycle features and LSTM long-cycle features. The CNN short-cycle features include local voltage changes and current peak changes, and the LSTM long-cycle features include timing oscillations, drift data, and periodic changes; Analyze the states of each node in the power supply topology network according to the CNN short-cycle features and LSTM long-cycle features, mark the nodes that do not meet the preset feature threshold as faulty nodes, and obtain a set of faulty nodes.

3. A dual-power dynamic reconstruction method according to claim 1, characterized in that Update the first power supply topology network and the second power supply topology network according to the hybrid node reconstruction path to generate a reconstructed power supply topology network. The method includes: Update the first power supply topology network according to the hybrid node reconstruction path to obtain a first updated power supply topology network, and update the second power supply topology network according to the hybrid node reconstruction path to obtain a second updated power supply topology network; Couple the first updated power supply topology network and the second updated power supply topology network to generate a reconstructed power supply topology network.

4. A dual-power dynamic reconfiguration method according to claim 1, characterized in that, Verify the feasibility of the generated reconstructed power supply topology network. The method includes: Perform load stability simulation, overload risk assessment, and network voltage quality detection on the reconstructed power supply topology network; If any one of the load stability simulation, overload risk assessment, and network voltage quality does not meet the corresponding threshold, roll back to the previous stage of the reconstruction state and update the hybrid node reconstruction path.

5. A dual-power dynamic reconfiguration system, characterized in that The system is used to implement a dual-power dynamic reconstruction method according to any one of claims 1 to 4. The system includes: A timing waveform sequence generation module for deploying a collection terminal to collect waveform data of each node in the power supply topology network to generate a timing waveform sequence. The power supply topology network includes a first power supply and a second power supply; A fault identification module for using a CNN-LSTM network to perform fault identification on the timing waveform sequence to obtain a set of faulty nodes; A power supply label construction module for constructing power supply labels for each node in the power supply topology network. The power supply labels include a first power supply label, a second power supply label, and a hybrid power supply label; A power supply topology network output module for performing first-order reconstruction on the nodes other than the set of faulty nodes according to the first power supply label and the second power supply label, and outputting a first power supply topology network based on the first power supply and a second power supply topology network based on the second power supply; A load balancing analysis module for taking the load balancing of the first power supply and the second power supply as the second-order reconstruction target, performing load balancing analysis on the node paths corresponding to the hybrid power supply label, and outputting a hybrid node reconstruction path; A reconstructed power supply topology network generation module for updating the first power supply topology network and the second power supply topology network according to the hybrid node reconstruction path to generate a reconstructed power supply topology network.

Images