Self-adaptive switching system and method for power supply and distribution

Abstract

The invention discloses a self-adaptive switching system and method for power supply and distribution, and relates to the technical field of smart power grids, and the method comprises the steps: loading a quantum algorithm library and power grid operation data to a quantum calculation platform, carrying out the parallel processing of the power grid operation data through a quantum algorithm, analyzing the abnormal condition in a current power grid, and obtaining a switching result according to the analysis result. According to the method, the optimal power supply path and the switching strategy from the power generation equipment to the load point are optimized, the optimal solution can be quickly found when a large-scale complex power grid is processed through the quantum algorithm, and the optimal power supply path and the switching strategy are obtained. The transmission loss of a power transmission line and power generation equipment is effectively reduced, and the power supply efficiency of a power grid is improved. In addition, by introducing a time attenuation factor and a priority weight, a switching strategy is further optimized, and efficient operation of a power grid and reasonable distribution of loads are ensured.

Description

Technical Field

[0001] The present invention relates to the technical field of smart grids, and particularly to an adaptive switching system and method for power supply and distribution. Background Art

[0002] With the continuous expansion of the power grid scale and the continuous progress of technology, ordinary power supply and distribution networks are gradually developing towards the intelligent direction. Early power supply and distribution networks mainly relied on manual operations and simple automation devices. Although these devices could achieve basic power supply functions, they were unable to cope with complex and changing load demands and abnormal detections.

[0003] There are still many deficiencies in the existing power supply and distribution networks in terms of intelligence. In terms of load demand prediction, existing methods lack the ability to efficiently process real-time data, making it difficult to accurately predict changes in load demands in the future for a period of time, and restricting the adaptive scheduling ability of the power grid. In addition, in terms of abnormal detection, existing methods cannot effectively handle complex fault modes in the power grid, and are prone to false alarms or missed detections, affecting the safe operation of the power grid. Summary of the Invention

[0004] In view of the above existing problems, the present invention is proposed.

[0005] Therefore, the present invention provides an adaptive switching method for power supply and distribution, which solves the problems of inaccurate load demand prediction and poor reliability of abnormal detection in the power grid.

[0006] To solve the above technical problems, the present invention provides the following technical solutions:

[0007] In a first aspect, the present invention provides an adaptive switching method for power supply and distribution, which includes collecting power grid operation data and load demand data;

[0008] Designing a corresponding quantum algorithm according to the power grid scale and power grid complexity and training it to form a quantum algorithm library;

[0009] Loading the quantum algorithm library and power grid operation data onto a quantum computing platform, using the quantum algorithm to perform parallel processing on the power grid operation data, analyzing abnormal situations in the current power grid, and generating a power grid status report according to the analysis results;

[0010] Inputting the power grid status report and load demand data into the quantum computing platform, using the quantum algorithm to calculate the optimal solution from the power generation equipment to the load point, and obtaining the optimal power supply path and switching strategy;

[0011] Sending the optimal power supply path and switching strategy to the central control center, and using an abnormal detection model to monitor changes in the power grid status and generate protection action suggestions.

[0012] As a preferred solution of the adaptive switching method for power supply and distribution of the present invention, wherein: the optimal power supply path and switching strategy are sent to the central control center, and an anomaly detection model is used to monitor the state change of the power grid and generate protection action suggestions, including the following steps,

[0013] Convert the optimal power supply path and switching strategy into JSON format, and use the HTTPS communication protocol to send the converted optimal power supply path and switching strategy to the central control center;

[0014] The central control center analyzes the converted optimal power supply path and switching strategy to generate control instructions for the switching time point, switching action, and load distribution ratio;

[0015] Select Autoencoder as the anomaly detection model;

[0016] Input the control instructions into the anomaly detection model to obtain the reconstruction error value;

[0017] Set the power grid state error threshold, and judge whether the current power grid state is normal according to the interval where the reconstruction error value is within the power grid state error threshold;

[0018] Generate protection action suggestions according to the judgment result.

[0019] In a second aspect, the present invention provides an adaptive switching system for power supply and distribution, including a data acquisition module that acquires power grid operation data and load demand data;

[0020] An algorithm design and training module that designs corresponding quantum algorithms according to the power grid scale and power grid complexity and conducts training to form a quantum algorithm library;

[0021] A power grid state report generation module that loads the quantum algorithm library and power grid operation data onto the quantum computing platform, uses the quantum algorithm to perform parallel processing on the power grid operation data, analyzes the abnormal conditions in the current power grid, and generates a power grid state report according to the analysis results;

[0022] A scheme planning module that inputs the power grid state report and load demand data into the quantum computing platform, uses the quantum algorithm to find the optimal solution from the power generation equipment to the load point, and obtains the optimal power supply path and switching strategy;

[0023] A protection action suggestion module that sends the optimal power supply path and switching strategy to the central control center, and uses the anomaly detection model to monitor the state change of the power grid and generate protection action suggestions.

[0024] In a third aspect, the present invention provides a computer device, including a memory and a processor, where the memory stores a computer program, and: when the computer program is executed by the processor, any step of the adaptive switching method for power supply and distribution as described in the first aspect of the present invention is implemented.

[0025] In a fourth aspect, the present invention provides a computer-readable storage medium, on which a computer program is stored, and: when the computer program is executed by the processor, any step of the adaptive switching method for power supply and distribution as described in the first aspect of the present invention is implemented.

[0026] The beneficial effects of the present invention are as follows: The power grid state report and load demand data are input into the quantum computing platform, and the quantum algorithm is used to calculate the optimal solution from the power generation equipment to the load point, obtaining the optimal power supply path and switching strategy, realizing the optimization of the best path and switching strategy from the power generation equipment to the load point. The quantum algorithm can quickly find the optimal solution when dealing with large-scale complex power grids, thereby effectively reducing the transmission losses of transmission lines and power generation equipment and improving the power supply efficiency of the power grid. In addition, by introducing a time decay factor and priority weights, the switching strategy is further optimized, ensuring the efficient operation of the power grid and the reasonable allocation of loads. BRIEF DESCRIPTION OF THE DRAWINGS

[0027] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the following will briefly introduce the drawings required for the description of the embodiments. Obviously, the following drawings are only some embodiments of the present invention. For those of ordinary skill in the art, other drawings can be obtained based on these drawings without creative efforts.

[0028] Figure 1 It is a flowchart of the adaptive switching method for power supply and distribution in Embodiment 1.

[0029] Figure 2 It is a determination diagram for determining whether the current power grid state is normal in Embodiment 1. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0030] In order to make the above objects, features, and advantages of the present invention more obvious and understandable, the following will make a detailed description of the specific embodiments of the present invention with reference to the accompanying drawings of the specification.

[0031] Embodiment 1, referring to Figure 1 and Figure 2 , is the first embodiment of the present invention. This embodiment provides an adaptive switching method for power supply and distribution, including the following steps:

[0032] S1. Grid operation data includes the voltage and current of power generation equipment, grid frequency (i.e., the number of times alternating current changes per second, usually measured in Hertz (Hz)), temperature, and transmission line parameters (i.e., resistance, reactance, capacitance, used for subsequent calculation of transmission losses of transmission lines).

[0033] Load demand data includes load growth rate, peak load, total load demand, time-of-use load demand, and user-classified load demand (i.e., dividing the respective load demands according to different types of users, such as industrial, commercial, and residential).

[0034] S2. Design a corresponding quantum algorithm according to the grid scale and grid complexity.

[0035] It includes the following steps.

[0036] Use the SCADA grid management program to obtain the grid scale of the power grid.

[0037] Open the client of SCADA, query a certain area (such as Area A) and its corresponding grid facilities (such as 5 power stations and 8 substations), use the export function to organize the queried area and grid facilities into Excel format, name it Area A_grid scale.xlsx, and obtain the grid scale of the power grid.

[0038] The grid scale of the power grid includes the number of power stations, substations, transformers, circuit breakers, relays in the power grid, and the length of transmission lines.

[0039] Select QAOA as the quantum algorithm based on the grid scale, define a unique identifier for each node in the power grid, and identify the connection relationship between each node.

[0040] Assign a unique prefix to each type of node (such as G for power station, S for substation, and L for load point), and assign numbers to each node in ascending order (i.e., G1, G2, S1, S2, L1, L2). After the assignment, generate a unique identifier according to the current geographical coordinates of the node, such as G_40.7128_74.0060 representing a power station located in a certain city.

[0041] Use an adjacency list (a data structure for storing nodes in a graph and their list of adjacent nodes) to create an empty list for each node in the power grid, and add the adjacent nodes of each node to the corresponding list. For example, if node S1 is connected to nodes L1 and L2, then add L1 and L2 to the list of S1, and also add S1 to the lists of L1 and L2.

[0042] Using the Pauli-Z matrix as a binary decision variable (a mathematical tool for representing whether a certain condition holds. Binary decision variables usually take values of 0 or 1, representing no and yes respectively), represent the node states of the connection relationships between each pair of nodes (divided into connected state and unconnected state, using the +1 eigenvalue and -1 eigenvalue of the Pauli-Z matrix to represent the node states respectively, that is, +1 for the connected state and -1 for the unconnected state);

[0043] Extract the current parameters and resistance parameters of the transmission line from the power grid database, and calculate the transmission loss of the transmission line. The expression is:

[0044] Z = I 2 ×R;

[0045] Among them, Z represents the transmission loss of the transmission line, I represents the value of the current parameter of the transmission line, and R represents the value of the resistance parameter of the transmission line;

[0046] Assume that the current and resistance of a transmission line are 100 A and 0.05 Ω respectively. Using this formula, the calculated transmission loss of this transmission line is 500 W (watt);

[0047] Define the interaction of node states according to the transmission loss of the transmission line (that is, the transmission loss generated when nodes are connected by a transmission line) and perform weighted summation to form a cost function. At the same time, convert the cost function into the form of a quantum Hamiltonian to obtain the cost Hamiltonian;

[0048] The formation process of the cost function is as follows: First, select a pair of nodes, such as (x, y), substitute (x, y) into the formula of the transmission loss of the transmission line, calculate the transmission loss (that is, Z′) generated when the nodes are connected by the transmission line and perform weighted summation. Finally, use the Pauli-Z matrix to define the node connection state (O) between x and y. Then the cost function C is expressed as:

[0049] C = ∑ x<y Z′ × O;

[0050] Converting the cost function into the form of a quantum Hamiltonian means replacing the node connection state O between x and y xy with a form suitable for quantum computing;

[0051] Based on the cost Hamiltonian, select aluminum as the superconducting material to design a superconducting circuit, and introduce Josephson junctions (structures composed of two layers of superconductors sandwiching an insulating layer) into the superconducting circuit to form superconducting qubits;

[0052] The reason for choosing aluminum is that aluminum has a low resistivity and high ductility, which is convenient for processing into complex circuit structures;

[0053] Coat a photoresist on a support material using lithography technology, and form a pattern of the required circuit shape through mask exposure. Based on the circuit shape pattern, deposit aluminum on the support material using electron beam evaporation, and at the same time etch and separate materials other than aluminum to form the final superconducting circuit;

[0054] Integrate Josephson junctions into the superconducting loop. After the superconducting circuit is cooled below its critical temperature (about 1200 °C), apply a drive signal to each qubit of the superconducting circuit using microwave pulses, and use a capacitive coupler to connect each group of qubits (a group of 2) after the drive signal is applied and combine the electric field effect (the interaction force generated by charged particles through the electric field) to change the plate area of the capacitive coupler, thereby adjusting the coupling capacitance value of the capacitor to form superconducting qubits;

[0055] Initialize the superconducting qubits to a uniform superposition state (that is, the superconducting qubits are in an equal-probability superposition state of all possible states), and use the Pauli-X matrix to define the drive term of each initialized superconducting qubit (that is, the external physical quantity that controls the state evolution of the superconducting qubit) to obtain the drive Hamiltonian;

[0056] Cool the superconducting qubits to a temperature of 0.01 K to form the ground state |0> with the lowest energy, and use Hadamard to emit a radio frequency signal to the superconducting qubits to convert the ground state |0> to a uniform superposition state

[0057] Perform a Pauli-X operation (that is, a control pulse signal) on the superconducting qubits in the uniform superposition state, flip the probability amplitude of the superconducting qubits to |0> or |1>, and add the drive term of the superconducting qubits and the probability amplitude of the superconducting qubits to obtain the drive Hamiltonian.

[0058] S3. Train the quantum algorithm to form a quantum algorithm library.

[0059] Including the following steps,

[0060] Use graph theory to identify the specific topological structure of the power grid (divided into tree structure, ring structure, and mesh structure) to obtain the complexity of the power grid;

[0061] Use graph theory to list the connection relationships between power grid devices and power grid nodes as the nodes and edges of the power grid graph respectively. Use the breadth-first search algorithm to create an access status array and a parent node array. Select any power grid node and add it to the queue (marked as visited). Take a node from the queue as the current node and traverse all the neighbor nodes of the current node until the queue is empty;

[0062] If it is found during the entire traversal process that the neighbor nodes of any node have been visited and are not its parent node, it indicates that there is a loop in the graph (if it is a single loop, the topological structure of the power grid is a ring structure; if there are multiple loops, the topological structure of the power grid is a mesh structure; if there is no loop, the topological structure of the power grid is a tree structure).

[0063] Based on the complexity of the power grid, select P layers as the number of layers of QAOA and initialize the angle parameters of QAOA.

[0064] The P layers here are selected according to the complexity of the power grid. Suppose the power grid has 100 nodes and 200 edges, and there are multiple loops, then initially select P as 4.

[0065] P being 4 means that the angle parameters of 4 layers need to be initialized. For each layer, use a random number generator to generate two random numbers in the interval [0, 2π] (θ represents the angle parameter for driving the evolution time of the Hamiltonian, and φ represents the angle parameter for driving the evolution time of the cost Hamiltonian). For example, in the first layer (P = 1), randomly draw a value from the interval [0, 2π]. For example, θ1 is 1.57 and φ1 is 3.04.

[0066] Based on the cost Hamiltonian and the driving Hamiltonian, use the COBYLA optimizer to iteratively adjust the initialized angle parameters to obtain the optimal angle parameters.

[0067] Input the initialized angle parameters into the COBYLA optimizer and determine a target function (whose core is to reflect the specific objective of the optimization problem). In the first round of iteration, COBYLA will choose a random direction as the initial trial direction to adjust the angle parameters (try to increase or decrease the value of the angle parameters and observe which change is more beneficial to the target function), and at the same time record the adjusted target function value. In the second round of iteration, if the target function value becomes better, record this direction as the effective direction (continue to adjust along this direction in the next round of iteration), otherwise mark it as the invalid direction (perform a small - range random perturbation on the angle parameters in the direction opposite to the invalid direction). As the number of iterations increases, COBYLA will gradually narrow the adjustment range of the angle parameters until the target function value becomes very small (i.e., to the power of negative six or smaller), and then obtain the optimal angle parameters.

[0068] Save the optimal angle parameters as a data file in a structured form (i.e., key - value pair form, where the key is the name of the angle parameter and the value is the corresponding angle parameter value) to form a quantum algorithm library.

[0069] The quantum algorithm library in this step can be used in different scenarios, such as testing new hardware and verifying algorithm performance. And the structured data file is easy to query and update, which helps to maintain the integrity and consistency of the quantum algorithm library in the long term.

[0070] S4. Load the quantum algorithm library and power grid operation data onto the quantum computing platform, and use the quantum algorithm to perform parallel processing on the power grid operation data to analyze the abnormal conditions in the current power grid. According to the analysis results, generate a power grid status report.

[0071] It includes the following steps:

[0072] Select IBM Quantum as the quantum computing platform and load the quantum algorithm library and power grid operation data onto IBM Quantum;

[0073] The reason for choosing IBM Quantum is that quantum processors with various different scales and performances can meet the needs from basic research to complex applications;

[0074] Log in to IBM Quantum, use the command pip install qiskit to install the Qiskit library and obtain the API key, and load the quantum algorithm library and power grid operation data onto IBM Quantum;

[0075] Select the voltage level of the power generation equipment, the current intensity, and the frequency stability of the power grid as the operation indicators of the power grid nodes;

[0076] The reason for choosing these three as the operation indicators of the power grid nodes is that the voltage level determines the output force of the power grid nodes and the synchronization ability with the power grid. By monitoring the current, the power consumption demand of the nodes and whether there is an overload phenomenon can be understood. And many electrical equipment is very sensitive to the change of the power grid frequency. Unstable frequency will lead to a decrease in the efficiency or failure of the electrical equipment;

[0077] According to the operation indicators of the power grid nodes, define the state categories of the power grid nodes and the specific number of qubits required, and at the same time formulate corresponding mapping rules for the state categories of each power grid node to map the state of the power grid node into qubit representation;

[0078] For each operation indicator (taking the voltage level as an example), divide it into several intervals according to the actual needs (multiply each interval to get the number of state categories): normal (voltage lower limit ≤ voltage value ≤ voltage upper limit), too low (voltage value < voltage lower limit), too high (voltage value > voltage upper limit);

[0079] Combine the intervals of different operation indicators to form the state categories of the power grid nodes (such as low voltage, high current load, slight frequency fluctuation);

[0080] The number of qubits depends on the total number of state categories. If there are N state categories in total, the number of qubits q required should satisfy that 2 to the power of q is greater than or equal to N (for example, if there are 3 operating indicators, and each indicator is divided into 3 intervals, then the total number of state categories is 3×3×3 = 27, and the number of qubits required is 2 to the fifth power = 32);

[0081] Each state category is represented by a unique binary string, and each state category is put into one-to-one correspondence with a binary string (such as low voltage, high current load, slight frequency fluctuation → 00010). Finally, the binary string is directly mapped to the state representation of the qubit (the qubit state corresponding to 00010 is 00010 → ∣00010);

[0082] The state categories of power grid nodes include normal state, low load state, and high load state;

[0083] Use the superposition and entanglement of quantum states to perform critical processing and correlation processing on the states of the mapped power grid nodes to obtain the quantum state of the power grid node states;

[0084] Critical processing: First, determine which state categories belong to the critical state (such as low voltage, high voltage, severe frequency fluctuation). Multiply the probability amplitudes by the critical state and the normal state respectively and add them to obtain a quantum composite state (representing the probability distribution between the normal state and the critical state). If the probability of the critical state in the quantum composite state is high, increase the value of the probability amplitude; otherwise, decrease the value of the probability amplitude. Take the adjusted quantum composite state as the result of critical processing;

[0085] Correlation processing: First, identify the strong correlations between power grid nodes (such as two nodes sharing the same transmission line or the frequency fluctuation of one node affecting the voltage level of another node). Use the quantum entanglement state to represent the correlations between power grid nodes. Suppose there are two nodes with states a and b respectively. Multiply a1 by b1 and add a2 multiplied by b2 (a1, a2, b1, b2 respectively represent a certain state category of the power grid node, such as low load and high load of current), and finally multiply by one over the square root of 2 to obtain the final quantum entanglement state. Take the quantum entanglement state as the result of correlation processing;

[0086] Select the initial state of the quantum state of the power grid node state (which is the ∣0 state. If 3 qubits are needed, the initial state is ∣000);

[0087] Use the Ry parameterized rotation gate to adjust the initial state of the power grid node state to a superposition state, and at the same time use the Rz parameterized rotation gate to rotate around the z-axis to adjust the phase of the initial state of the power grid node state;

[0088] Introduce a globe. Use the Ry parametric rotation gate to change the latitude position of the initial state of the power grid node state on the globe (if the initial state is at the North Pole ∣0, after rotating by an angle θ along the y-axis using Ry, the initial state of the power grid node state will move near the equator or even close to the South Pole ∣1, and this ∣1 is the adjusted superposition state), while the Rz parametric rotation gate changes the longitude position of the initial state of the power grid node state on the globe (if the initial state of the power grid node state was originally at the prime meridian 0° longitude, after rotating by an angle θ along the z-axis using Rz, its longitude position will move to 90° longitude, and this 90° longitude is the phase of the initial state of the adjusted power grid node state);

[0089] Use the Qiskit computing framework to combine the superposition state and phase of the adjusted power grid node state to form a QAOA circuit;

[0090] Use the driving Hamiltonian to add a mixed state to the superposition state and phase of the adjusted power grid node state to form the mixed layer structure of the QAOA circuit, and install the circuit interface of the mixed layer structure through the Hadamard gate in the Qiskit computing framework. Through this circuit interface, the superposition state and phase are combined together to form a complete QAOA circuit;

[0091] Use the Aer simulator to run the QAOA circuit on the quantum computing platform, and use the quantum state function to calculate the state probability distribution of the adjusted power grid node. The expression is:

[0092]

[0093] where, Q i represents the quantum state of the i-th power grid node after adjustment, P(Q i ) represents the state probability distribution of the quantum state of the i-th power grid node after adjustment, ψ() represents the quantum state function, ψ(Q i ) represents the quantum state function of the i-th power grid node after adjustment, which is used to describe the complete quantum characteristics of Q i , and d represents the operator for integrating the quantum state of the i-th power grid node after adjustment. The role of integration is to sum up the probability density of the quantum state in the entire space to ensure that the total probability is 1 (normalization condition);

[0094] Use Pandas to analyze the state probability distribution of the adjusted power grid node to find out the abnormal situations existing in the current power grid (including too low or too high voltage level, too large or unstable frequency fluctuation, and too high current load);

[0095] Use the data frame in Pandas to sort the state probability distribution of the adjusted power grid nodes in descending order, and use a pie chart to visually display the probability distribution under different power grid node states. Through this visualization method, quickly discover which states occupy a large proportion and which states are relatively rare;

[0096] Based on the analysis results, generate a power grid status report (including an overview of power grid operation, a summary of abnormal conditions, and a visualization chart of the pie chart).

[0097] S5. Input the power grid status report and load demand data into the quantum computing platform, and use quantum algorithms to calculate the optimal solution from the power generation equipment to the load point to obtain the optimal power supply path and switching strategy.

[0098] Including the following steps,

[0099] Normalize the load demand data, and use a smoothing function to calculate the load demand prediction value for a future period of time. The expression is:

[0100]

[0101] Among them, L(s) represents the load demand prediction value for a future period of time, s represents a future period of time, n represents the time window size of the load demand data, k represents the discrete time point index, α represents the smoothness of controlling the load demand prediction value of the future period s. A larger α value will make the load demand prediction value smoother, while a smaller α value will retain the fluctuations of the load demand prediction value, and B represents the load demand data after normalization processing;

[0102] Use the SCADA power grid management program to obtain the input and output powers of the power generation equipment under different load conditions (such as 50% load, 70% load), and establish an equipment efficiency model by plotting the relationship diagram between the input and output efficiency and the load conditions;

[0103] Determine the power generation equipment efficiency (divide the output power by the input power to obtain the efficiency of the power generation equipment) and plot the relationship diagram between the input and output efficiency and the load conditions (the horizontal axis represents the load conditions, and the vertical axis represents the power generation equipment efficiency), observe the trend between the input and output efficiency and the load conditions, and find the load range with the highest efficiency (usually, the efficiency of the power generation equipment will first increase and then decrease as the load increases, showing a hump-shaped curve, and the load range with the highest efficiency is the highest point of the hump);

[0104] The mathematical model using piecewise linear functions divides the load range with the highest efficiency into several intervals. Within each interval, the relationship between efficiency and load is represented by the linear equation y = kx + b, and the slope and intercept of the hump curve are measured to obtain fitting parameters. Based on these fitting parameters, a normal distribution of device efficiency (μ, σ 2 )(where μ represents the fitting parameter and σ represents the standard deviation of the load point, which can be estimated using the load data of historical devices) is defined for each load condition to form a complete device efficiency model;

[0105] The transmission loss ratio of each power generation device (defined as 1 minus the efficiency of the power generation device) is calculated using the device efficiency model and summed to obtain the total transmission loss of the power generation devices (the transmission loss of a power generation device is obtained by multiplying the output power of this power generation device by its transmission loss ratio. Assuming there are three devices, the transmission losses of each are directly calculated and added to obtain the total transmission loss);

[0106] The grid status report and the predicted load demand values for a future period are input into the quantum computing platform. Combining the total transmission loss of the power generation devices and the connection relationships between each grid node, a quantum algorithm is used to calculate the path optimization index for each power supply path. The expression is:

[0107]

[0108] where, G V represents the path optimization index of the Vth power supply path, m represents the number of grid nodes, H j represents the total transmission loss of the jth power generation device, β represents the influence factor of the total transmission loss of the power generation devices. A larger β value will make the influence of the transmission loss of the power generation devices more significant, and vice versa will reduce its influence. max(B) represents the maximum value of the normalized load demand data;

[0109] Select the power supply path with the highest path optimization index as the optimal power supply path;

[0110] Suppose there are three power supply paths (a, b, c), and the transmission losses of a, b, and c are 5, 8, and 6 respectively, and their normalized load demand data are 0.8, 0.6, and 0.7 respectively. Substitute the above data into the path optimization index formula for calculation, and the path optimization indexes of the three power supply paths a, b, and c are 0.235, 0.137, and 0.181 respectively. Then the path optimization index of the a power supply path is the highest, and a is selected as the optimal power supply path;

[0111] Set the initial time point and priority weight for grid nodes (the initial time point is set according to specific events and periodically operating devices. Specific events include sudden increases in current load, access to new energy, etc. The priority weight is set according to the stability and peak demand of different grid nodes. For example, nodes supplying power to hospitals or data centers should be given higher weights), and introduce a time decay factor to calculate the optimal switching time point for grid nodes. The expression is as follows:

[0112]

[0113] Where, T i represents the optimal switching time point of the i-th grid node, t represents the current time point of the grid node, W i represents the priority weight of the i-th grid node, t0 represents the initial time point of the grid node, and γ represents the time decay factor, whose function is to reduce the influence of the early state over time and make the calculation pay more attention to recent changes;

[0114] Based on the optimal switching time point of the grid node, perform corresponding switching actions;

[0115] The switching actions include power switching, load transfer, and equipment start / stop;

[0116] Power switching: When the main power supply fails or the load demand exceeds its capacity, detect the status of the current power supply (such as voltage, power output), and disconnect the connection of the current power supply and enable the standby power supply at the optimal switching time point;

[0117] Load transfer: When a certain line or equipment is approaching full load or overload, identify the topology of the power grid (such as tree-shaped, meshed), determine the target node or line for load transfer, and reallocate the current path at the optimal switching time point, disconnecting the switch status of the circuit breaker;

[0118] Equipment start / stop: When the load demand increases and standby equipment is started, it is first necessary to ensure that the equipment start / stop process complies with safety specifications (such as warm-up, cooling time), start the equipment at the optimal switching time point, and gradually increase the output power to avoid impact. When stopping the equipment, it is necessary to gradually reduce the output power to ensure a smooth transition of the equipment state;

[0119] The corresponding switching action is the switching strategy.

[0120] S6. Send the optimal power supply path and switching strategy to the central control center, and use the anomaly detection model to monitor the state changes of the power grid and generate protection action suggestions.

[0121] Including the following steps,

[0122] Convert the optimal power supply path and switching strategy into JSON format, and send the converted optimal power supply path and switching strategy to the central control center using the HTTPS communication protocol;

[0123] In this step, the JSON format converted from the optimal power supply path and switching strategy is more concise than other formats (such as XML), reducing the volume of data transmission. Using HTTPS ensures the security and integrity of data transmission, reducing potential risks caused by data leakage or tampering;

[0124] The central control center parses the converted optimal power supply path and switching strategy to generate a switching time point (for example, if the best switching time for a certain node is 2025-02-11 15:00:00, then generate the instruction: node_1:{switchTime:2025-02-11 15:00:00}), a switching action (taking the action of starting and stopping the device as an example, the generated instruction is: node_3:{action:startDevice,deviceID:generator_1}), and a control instruction for the load distribution ratio (assuming the load distribution ratio of a certain node is 60%, then generate the instruction node_1:{loadDistribution:60%});

[0125] Select Autoencoder as the anomaly detection model (the reason for selection is that Autoencoder can effectively extract important features in high-dimensional data and can learn complex non-linear relationships in the data);

[0126] The application of Autoencoder in the detection of power grid abnormal states is not a simple scenario migration. Instead, through targeted calculations and detections, Autoencoder can capture tiny abnormal signals and calculate the reconstruction error value by inputting control instructions. Based on this reconstruction error value and the set threshold, it quantifies the abnormal degree of the power grid, achieving an accurate judgment of the power grid state, thus greatly improving the sensitivity of Autoencoder detection;

[0127] Input the control instruction into the anomaly detection model to obtain the reconstruction error value;

[0128] After inputting into the anomaly detection model, the input control instruction is compressed into a low-dimensional feature representation by the encoder, the compressed low-dimensional feature representation is decoded into the initial control instruction by the decoder, the mean square error is used to calculate the square of the difference between the low-dimensional feature representation and the initial control instruction, and finally the average value of the squared difference is obtained to get the reconstruction error value;

[0129] Set the grid state error threshold (set according to the characteristics of the reconstruction error distribution, calculated using the mean ± 3 times the standard deviation, 0.02 - 0.08), and determine whether the current grid state is normal according to the interval where the reconstruction error value is within the grid state error threshold;

[0130] When the reconstruction error value is less than or equal to 0.05 and greater than or equal to 0.02, the current grid state is normal. When the reconstruction error value is greater than 0.05 and less than or equal to 0.08, the current grid state is abnormal;

[0131] Generate protection action suggestions according to the judgment results;

[0132] When the current grid state is abnormal, the protection action suggestions are as follows: Pop up a red warning prompt on the monitoring interface of the central control center, and highlight the abnormal grid nodes or areas. Send the warning prompt to the operation and maintenance personnel via text message. The operation and maintenance personnel locate the affected lines according to the grid topology structure and the specific locations of the abnormal grid nodes or areas (for example, if the current of a certain line suddenly fluctuates greatly, then focus on checking the switch equipment of this line), and automatically trigger the circuit breaker or relay to disconnect the faulty line; For critical load areas, find the specific reasons (such as insulation aging, load overlimit, etc.) and switch to the standby power supply to ensure the continuous operation of important equipment.

[0133] This embodiment also provides an adaptive switching system for power supply and distribution, including:

[0134] A data acquisition module that acquires grid operation data and load demand data;

[0135] An algorithm design and training module that designs corresponding quantum algorithms according to the grid scale and grid complexity and conducts training to form a quantum algorithm library;

[0136] A grid state report generation module that loads the quantum algorithm library and grid operation data onto the quantum computing platform, uses the quantum algorithm to perform parallel processing on the grid operation data, analyzes the abnormal situations in the current grid, and generates a grid state report according to the analysis results;

[0137] A scheme planning module that inputs the grid state report and load demand data into the quantum computing platform, uses the quantum algorithm to find the optimal solution from the power generation equipment to the load point, and obtains the optimal power supply path and switching strategy;

[0138] A protection action suggestion module that sends the optimal power supply path and switching strategy to the central control center, and uses the anomaly detection model to monitor the state changes of the grid and generate protection action suggestions.

[0139] This embodiment also provides a computer device, which is applicable to the case of the adaptive switching method for power supply and distribution, including: a memory and a processor; the memory is used to store computer-executable instructions, and the processor is used to execute the computer-executable instructions to implement the adaptive switching method for power supply and distribution as proposed in the above embodiment.

[0140] The computer device may be a terminal, and the computer device includes a processor, a memory, a communication interface, a display screen, and an input device connected through a system bus. Among them, the processor of the computer device is used to provide computing and control capabilities. The memory of the computer device includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system and a computer program. The internal memory provides an environment for the operation of the operating system and the computer program in the non-volatile storage medium. The communication interface of the computer device is used to communicate with an external terminal in a wired or wireless manner, and the wireless manner can be implemented through WIFI, a carrier network, NFC (Near Field Communication), or other technologies. The display screen of the computer device may be a liquid crystal display screen or an electronic ink display screen, and the input device of the computer device may be a touch layer covering the display screen, or a button, a trackball, or a touchpad provided on the outer shell of the computer device, or an external keyboard, a touchpad, or a mouse, etc.

[0141] This embodiment also provides a storage medium, on which a computer program is stored, and when the program is executed by a processor, it implements the adaptive switching method for power supply and distribution as proposed in the above embodiment; the storage medium can be implemented by any type of volatile or non-volatile storage device or a combination thereof, such as Static Random Access Memory (SRAM for short), Electrically Erasable Programmable Read-Only Memory (EEPROM for short), Erasable Programmable Read-Only Memory (EPROM for short), Programmable Read-Only Memory (PROM for short), Read-Only Memory (ROM for short), magnetic memory, flash memory, a magnetic disk, or an optical disc.

[0142] In summary, the present invention: inputs the power grid status report and load demand data into the quantum computing platform, uses the quantum algorithm to calculate the optimal solution from the power generation equipment to the load point, and obtains the optimal power supply path and switching strategy, realizing the optimization of the best path and switching strategy from the power generation equipment to the load point. The quantum algorithm can quickly find the optimal solution when dealing with large-scale complex power grids, thereby effectively reducing the transmission losses of transmission lines and power generation equipment and improving the power supply efficiency of the power grid. In addition, by introducing the time decay factor and priority weight, the switching strategy is further optimized to ensure the efficient operation of the power grid and the reasonable allocation of loads.

[0143] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and not to limit them. Although the present invention has been described in detail with reference to the preferred embodiments, those of ordinary skill in the art should understand that the technical solutions of the present invention can be modified or equivalently replaced without departing from the spirit and scope of the technical solutions of the present invention, and they should all be covered within the scope of the claims of the present invention.

Claims

1. An adaptive switching method for power supply and distribution, characterized in that: include, Collect power grid operation data and load demand data; Design and train corresponding quantum algorithms according to the scale and complexity of the power grid to form a quantum algorithm library; Load the quantum algorithm library and power grid operation data onto the quantum computing platform, use the quantum algorithm to process the power grid operation data in parallel, analyze the abnormal conditions in the current power grid, and generate a power grid status report based on the analysis results; Input the grid status report and load demand data into the quantum computing platform, use quantum algorithms to calculate the optimal solution from the power generation equipment to the load point, and obtain the optimal power supply path and switching strategy; The optimal power supply path and switching strategy are sent to the central control center, and the anomaly detection model is used to monitor the state changes of the power grid and generate protection action recommendations.

2. The adaptive switching method for power supply and distribution according to claim 1, characterized in that: The grid operation data includes voltage, current, grid frequency, temperature and transmission line parameters of the power generation equipment; The load demand data includes load growth rate, peak load, total load demand, time-sharing load demand and user-classified load demand.

3. The adaptive switching method for power supply and distribution according to claim 2, characterized in that: Designing the corresponding quantum algorithm according to the scale and complexity of the power grid includes the following steps: Use SCADA grid management programs to obtain grid size of the power grid; The grid scale of the power grid includes the number of power stations, substations, transformers, circuit breakers, relays and the length of transmission lines of the power grid; QAOA is selected as the quantum algorithm based on the scale of the power grid, a unique identifier is defined for each node in the power grid, and the connection relationship between each node is identified; Use the Pauli-Z matrix as a binary decision variable to represent the node status of the connection relationship between each node; Extract the current parameters and resistance parameters of the transmission line from the power grid database and calculate the transmission loss of the transmission line; The interaction of the node states is defined according to the transmission loss of the transmission line and weighted summed to form a cost function, and the cost function is converted into the form of quantum Hamiltonian to obtain the cost Hamiltonian; Based on the cost Hamiltonian, aluminum is selected as the superconducting material to design a superconducting circuit, and a Josephson junction is introduced into the superconducting circuit to form a superconducting quantum bit; The superconducting quantum bits are initialized to a uniform superposition state, and the driving term of each initialized superconducting quantum bit is defined using the Pauli-X matrix to obtain the driving Hamiltonian.

4. The adaptive switching method for power supply and distribution according to claim 3, characterized in that: Training the quantum algorithm to form a quantum algorithm library includes the following steps: Use graph theory to identify the specific topology of the power grid and obtain the complexity of the power grid; Based on the complexity of the power grid, P layers are selected as the number of layers of QAOA and the angle parameters of QAOA are initialized; Based on the cost Hamiltonian and the driving Hamiltonian, the COBYLA optimizer is used to iteratively adjust the initialized angle parameters to obtain the optimal angle parameters; The optimal angle parameters are saved as structured data files to form a quantum algorithm library.

5. The adaptive switching method for power supply and distribution according to claim 4, characterized in that: The quantum algorithm library and power grid operation data are loaded onto the quantum computing platform, and the power grid operation data are processed in parallel using the quantum algorithm to analyze the abnormal conditions in the current power grid. Based on the analysis results, a power grid status report is generated, including the following steps: Select IBM Quantum as the quantum computing platform and load the quantum algorithm library and power grid operation data onto IBM Quantum; Select the voltage level and current intensity of the power generation equipment and the frequency stability of the power grid as the operation indicators of the power grid node; According to the operation indicators of the power grid nodes, the state categories of the power grid nodes and the specific number of quantum bits required are defined, and corresponding mapping rules are formulated for the state categories of each power grid node to map the state of the power grid node into quantum bit representation; The state categories of the power grid nodes include normal state, low load state and high load state; Use superposition and entanglement of quantum states to perform criticality processing and correlation processing on the states of the mapped power grid nodes to obtain the quantum state of the power grid nodes; Selecting the initial state of the quantum state of the grid node state; The initial state of the power grid node state is adjusted to a superposition state using the Ry parameterized revolving door, and the Rz parameterized revolving door is used to rotate around the z axis to adjust the phase of the initial state of the power grid node state; The superposition state and phase of the adjusted grid node states are combined using the Qiskit computing framework to form a QAOA circuit; Use the Aer simulator to run the QAOA circuit on the quantum computing platform and use the quantum state function to calculate the state probability distribution of the adjusted power grid nodes; Use Pandas to analyze the state probability distribution of the adjusted power grid nodes and find out the abnormal conditions in the current power grid; Based on the analysis results, a power grid status report is generated.

6. The adaptive switching method for power supply and distribution according to claim 5, characterized in that: The grid status report and load demand data are input into the quantum computing platform, and the optimal solution from the power generation equipment to the load point is calculated using the quantum algorithm to obtain the optimal power supply path and switching strategy, including the following steps: Normalize the load demand data and use the smoothing function to calculate the load demand forecast value for a period of time in the future; Use SCADA power grid management program to obtain the input and output power of power generation equipment under different load conditions, and establish equipment efficiency model by plotting the relationship between input and output efficiency and load conditions; Use the equipment efficiency model to calculate the transmission loss ratio of each power generation equipment and sum them up to obtain the total transmission loss of the power generation equipment; The grid status report and the load demand forecast for a period of time in the future are input into the quantum computing platform. Combined with the total transmission loss of the power generation equipment and the connection relationship between each grid node, the quantum algorithm is used to calculate the path optimization index of each power supply path. Select the power supply path with the highest path optimization index as the optimal power supply path; Set the initial time point and priority weight of the grid node, introduce the time decay factor, and calculate the optimal switching time point of the grid node; Based on the optimal switching time point of the power grid node, the corresponding switching action is executed; The corresponding switching action is a switching strategy.

7. The adaptive switching method for power supply and distribution according to claim 6, characterized in that: The optimal power supply path and switching strategy are sent to the central control center, and the anomaly detection model is used to monitor the state changes of the power grid and generate protection action suggestions, including the following steps: Convert the optimal power supply path and switching strategy into JSON format, and send the converted optimal power supply path and switching strategy to the central control center using the HTTPS communication protocol; The central control center analyzes the optimal power supply path and switching strategy after the conversion format, and generates control instructions for switching time point, switching action and load distribution ratio; Select Autoencoder as the anomaly detection model; Input the control instruction into the anomaly detection model to obtain the reconstruction error value; A grid state error threshold is set, and whether the current grid state is normal is determined based on the interval within which the reconstruction error value is within the grid state error threshold; Based on the judgment results, a protection action suggestion is generated.

8. An adaptive switching system for power supply and distribution, based on the adaptive switching method for power supply and distribution according to any one of claims 1 to 7, characterized in that: include, Data acquisition module, collecting power grid operation data and load demand data; Algorithm design and training module: designs and trains corresponding quantum algorithms according to the scale and complexity of the power grid to form a quantum algorithm library; The power grid status report generation module loads the quantum algorithm library and power grid operation data onto the quantum computing platform, uses quantum algorithms to process the power grid operation data in parallel, analyzes the abnormal conditions in the current power grid, and generates a power grid status report based on the analysis results; The solution planning module inputs the grid status report and load demand data into the quantum computing platform, uses quantum algorithms to find the optimal solution from the power generation equipment to the load point, and obtains the optimal power supply path and switching strategy; The protection action recommendation module sends the optimal power supply path and switching strategy to the central control center, and uses the anomaly detection model to monitor the state changes of the power grid and generate protection action recommendations.

9. A computer device comprising a memory and a processor, wherein the memory stores a computer program, wherein: When the processor executes the computer program, the steps of the adaptive switching method for power supply and distribution described in any one of claims 1 to 7 are implemented.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the steps of the adaptive switching method for power supply and distribution according to any one of claims 1 to 7 are implemented.

Images