A method for evaluating rationality of power grid investment planning scheme considering low-carbon target

Abstract

This invention relates to a method for evaluating the rationality of power grid investment planning schemes that consider low-carbon objectives. The method includes: collecting relevant data and indicators related to power grid investment planning; dividing the power grid load forecasting area into regions based on the relevant data and indicators; using a seasonal decomposition method for time series data to decompose the power system load data into three components; constructing prediction models for different components to obtain load forecasting results; using the k-means algorithm to cluster daily wind and solar power output and load data throughout the year; selecting typical days from each quarter for short-term operation simulations; constructing a planning scheme evaluation model with the objective of minimizing operating costs; simulating the matching relationship between the output and load curves of various power source types; optimizing the system's peak-shaving capacity and economic efficiency, as well as the power system's investment and planning schemes; and solving the multi-objective model using the ocean predator algorithm. If all planning schemes meet the system's peak-shaving requirements, the planning results are considered reasonable.

Description

Technical Field

[0001] This invention relates to the field of power grid planning and evaluation technology, and in particular to a method for evaluating the rationality of power grid investment planning schemes that take into account low-carbon objectives. Background Technology

[0002] With the severe challenges of global climate change and the increasing urgency of sustainable development goals, power grid investment planning has shifted from simply pursuing economic benefits and power supply security to a more comprehensive consideration, with low-carbon objectives becoming a key guiding principle. Traditional power grid planning often focuses on meeting the ever-increasing electricity demand, optimizing the grid structure, and reducing costs, but it often fails to adequately consider environmental impacts, especially carbon emissions. With the rapid development of renewable energy, the transformation of the energy structure, and the fulfillment of international emission reduction commitments, power grid investment must balance economic and environmental benefits, ensuring that while meeting future load growth, it also promotes the efficient use of low-carbon energy and a significant reduction in carbon emissions.

[0003] Problems with existing technology

[0004] Static planning methods: Many existing power grid planning methods fail to fully consider the dynamics of energy transition and carbon emission constraints, resulting in planning outcomes that may not be adaptable to future adjustments to low-carbon policies and rapid changes in renewable energy.

[0005] Incomplete carbon emission assessment: When assessing power grid investment projects, carbon emission analysis is often limited to direct emissions, neglecting indirect emissions throughout the entire life cycle, such as equipment manufacturing, transportation, construction and decommissioning.

[0006] Data and model limitations: Existing models face difficulties in acquiring data and make overly simplistic assumptions when dealing with complex systems, such as grid stability under high-proportion renewable energy access and cross-regional energy flow and dispatch. This makes it difficult to accurately predict carbon emissions and costs.

[0007] Optimization algorithm efficiency and accuracy: Traditional optimization algorithms suffer from slow convergence speed and low solution quality when dealing with large-scale, multi-objective, and strongly constrained power grid planning problems, making it difficult to find the optimal investment scheme that is both economical and environmentally friendly.

[0008] The patent application (patent) number CN201510078491.6, entitled "A Method for Optimizing Distribution Network Planning Investment Based on Coordination of Reliability and Economy," mentions determining the unit load reliability improvement cost and unit load increase cost, then determining the investment benefits of corresponding zones, and finally establishing a distribution network investment optimization allocation model with the goal of maximizing investment benefits. Compared with this patent, the advantages of this patent are:

[0009] Comprehensive evaluation model: This proposal puts forward a rationality evaluation model for power grid investment planning that comprehensively considers low-carbon goals. By integrating multiple factors such as load forecasting, energy supply and demand balance, carbon emission measurement, and economic cost analysis, a comprehensive and dynamic evaluation system is formed.

[0010] Dynamic optimization framework: It adopts a method that combines rolling planning with scenario analysis to adapt to rapid changes in energy policies and market demands, dynamically adjust investment direction and scale, and ensure the foresight and flexibility of the planning scheme.

[0011] Advanced Algorithm Applications: Combining advanced optimization algorithms, such as genetic algorithms, particle swarm optimization, and deep reinforcement learning, improves the computational efficiency and solution accuracy of the model when dealing with large-scale data and multi-objective optimization problems, and quickly finds the optimal or near-optimal low-carbon investment strategy.

[0012] Integration of policy and market mechanisms: Fully consider external factors such as carbon trading markets, green financial instruments, and subsidy policies, optimize investment cost and benefit analysis, and promote the economic feasibility and market competitiveness of clean energy projects. Summary of the Invention

[0013] To address the problems existing in the prior art, this invention proposes a method for evaluating the rationality of power grid investment planning schemes that take into account low-carbon objectives.

[0014] The technical solution of the present invention is as follows:

[0015] On the one hand, this invention provides a method for evaluating the rationality of power grid investment planning schemes that consider low-carbon goals, including the following steps:

[0016] Relevant data and indicators related to power grid investment planning are collected. Based on these data and indicators, the power grid load forecasting area is divided into regions. A seasonal decomposition method for time series data is used to decompose the power system load data into three parts: real-time variation, trend-stable component, and irregular fluctuation component. The power system load data is preprocessed to obtain a dataset, which is then divided into a training dataset and a test dataset. Power grid load forecasting models are constructed for different components. The training dataset is used to train the load forecasting model. The trained load forecasting model is then used to predict samples in the test dataset to obtain load forecasting results. The load forecasting results are inversely normalized to obtain the final load forecasting results. The error between the actual load value and the final load forecasting results is compared. The forecasting model is evaluated based on this error, and the final load assessment result is obtained.

[0017] The k-means algorithm was used to cluster the daily wind and solar power output and load data throughout the year, and typical days of each quarter were selected for short-term operation simulation.

[0018] Construct a planning scheme evaluation model with the goal of minimizing operating costs, simulate the matching relationship between the output and load curves of various power sources, and optimize the peak-shaving capacity and economy of the power system, as well as the investment and planning schemes of the power system based on the matching relationship.

[0019] The complex constraints in the planning scheme evaluation model are solved by the marine predator algorithm. If all planning schemes meet the system peak-shaving requirements, the planning results are considered reasonable.

[0020] Preferably, the relevant data includes: local statistical yearbooks, national economic and social development plans, urban and rural master plans, land use master plans, power grid operation and basic statistical data, and planning results of higher-level power grids;

[0021] The indicators include: regional GDP, population, power grid asset size, geographical area, and load density;

[0022] Cluster analysis was used to divide the power grid load forecasting area into regions, combining relevant data and indicators. The specific steps included:

[0023] Assume c is the number of clusters. It is a set of vectors in the feature space, where v i (i = 1, 2, ...) represents the cluster prototype vector of the i-th class; μ ik ∈U, where U is a c×n fuzzy matrix that satisfies the following condition:

[0024]

[0025] The objective function of the improved fuzzy clustering algorithm is defined as follows:

[0026] max{J(U, V)}

[0027]

[0028] in μ ik For data point x k Membership degree of cluster i; μ ik For data point x k The membership degree V of cluster j is the clustering result;

[0029] Given the number of clusters c (2 ≤ c ≤ n, where n is the number of data points); set the iteration stopping threshold ε and the fuzzy index m; initialize the clustering prototype pattern v. (0) Set the iteration counter b = 0;

[0030] For any i, if k exists d ik b >0, then we have

[0031]

[0032] If there exists i, r such that d ik b =0, then μ ir ( b ) = 1, and for j ≠ i, μ ir ( b ) = 0;

[0033]

[0034] If ||V (b+1) -V (b) If || < ε, the improved fuzzy clustering algorithm stops and outputs the partition matrix U and the cluster prototype V; otherwise, b = b + 1, and the improved fuzzy clustering algorithm is executed iteratively.

[0035] Determine the optimal classification tree C m Calculate the value of J(C). If the condition J(Cm) = max{J(C)} is satisfied, then Cm is considered to be the optimal number of classifications. The corresponding output partition matrix U and cluster prototype are obtained to form the final clustering result.

[0036] Preferably, the step of decomposing the power system load data into three parts—real-time variation, trend-stable component, and irregular fluctuation component—includes:

[0037] Decompose the original load sequence in the power system load data into subsequences; divide Y t Represented as the actual time series value for period t, the time series seasonal trend adjustment method will use Y... t It is decomposed into three components: seasonal component, trend component, and irregular component; these three components are combined by an additive or multiplicative model, as shown below:

[0038] The additive decomposition model is defined as follows:

[0039] Y t =Trend t +Seasonal t +Irregular t

[0040] The multiplicative factorization model is defined as follows:

[0041] Y t =Trend t ×Seasonal t ×ιrregular t

[0042] In the above two equations, Trendt The trend component over time period t, Seasonal t The seasonal component of time period t, Irregular t It is the irregular component in time period t;

[0043] A multiplicative model is used when seasonal fluctuations are proportional to the level of the time series, and an additive decomposition model is used when seasonal fluctuations do not depend on the level of the time series.

[0044] Using a feedforward neural network method, prediction models are constructed for the above decomposition models respectively, and finally superimposed to obtain the final load prediction result.

[0045] Preferably, the method involves using the k-means algorithm to cluster the daily wind and solar power output and load data throughout the year, and selecting typical days from each quarter for short-term operational simulation. The specific steps are as follows:

[0046] Data collection and preprocessing: First, collect wind and solar power output data and grid load data for every day of the year. The wind and solar power output data includes wind power generation corresponding to wind speed and photovoltaic power generation corresponding to sunshine duration and intensity. The data is then cleaned, outliers are removed, and the data is standardized.

[0047] Determine the number of clusters: Select the K value, which is predetermined by the elbow rule and the silhouette coefficient. For power system data with obvious seasonality, the K value is set to 4, corresponding to the typical days of each quarter of the year.

[0048] Initialization and Iteration: The K-means algorithm starts by randomly selecting K center points and iterates through the following steps until convergence:

[0049] Assign each data point to the category of the nearest centroid; recalculate the centroid of each category until the change in the centroid is less than a preset threshold or the maximum number of iterations is reached.

[0050] Typical day selection: The center point of each cluster represents a typical day of a quarter, which includes the typical characteristics of wind power, solar power generation and grid load in that quarter; the typical day data is used for simulation and analysis to reduce the amount of data processing during actual operation simulation.

[0051] Preferably, the evaluation model for the planning scheme aimed at minimizing operating costs includes an objective function and constraints.

[0052] The operation simulation layer uses the minimum operating cost as the objective function to verify the system's peak-shaving capability; the operation simulation layer operating cost F2 includes three aspects: fuel cost and coal-fired power unit start-up and shutdown costs F. 1t Energy storage device charging, discharging, and maintenance costs F 2tAnd interruptible load compensation costs and wind / solar curtailment penalty costs F 3t Specifically, as shown in the following formula:

[0053]

[0054] In the formula: t represents the number of hours, T represents the total number of hours of short-term simulation; X RA X CA X KA These represent the number of fuel cell units, energy storage units, and renewable energy units participating in the short-term operation simulation, respectively. These are the fuel cost, start-up cost, and shutdown cost for the i-th fuel unit, respectively. These represent the actual output, startup state variables, and shutdown state variables of the i-th fuel unit in hour t, respectively. These are the unit discharge cost, charging cost, and maintenance cost for the j-th energy storage device, respectively. These are the real-time discharge power, charging power, and actual output of the j-th energy storage device in hour t, respectively. For interruptible load; ρ IL ρ ka These are the interruptible load compensation cost coefficient and the wind and solar curtailment factor, respectively. These are the available power and actual output of the k-th energy storage device in hour t, respectively.

[0055] Mode Fuel costs in China can be expressed as a quadratic function:

[0056]

[0057] In the formula: a i b i c i These are the consumption coefficients of the i-th fuel unit.

[0058] Preferably, the constraints specifically include: power balance constraints, fuel unit output constraints, unit start-up and shutdown time constraints, renewable energy output constraints, various power source ramping constraints, system hot standby constraints, and energy storage constraints.

[0059] The power balance constraint requires that, in a power system, the output of all power sources and the consumption of all loads must be kept in balance, as shown in the following formula:

[0060]

[0061] In the formula: These represent the active power output of the fuel cell unit, energy storage unit, and renewable energy unit in hour t, respectively. Let t be the interruptible load of the system participating in the short-term operation simulation at hour t. The system load demand in hour t;

[0062] The output constraint of the fuel unit is that the real-time output of the fuel unit must meet the upper and lower limit constraints during the operation of the power system, as shown in the following formula:

[0063]

[0064] In the formula: U i，t The generator unit is in start-up or shutdown status. These are the minimum and maximum outputs of the fuel unit, respectively. This represents the actual output of the fuel unit in hour t.

[0065] The unit start-up and shutdown time constraints are specifically shown in the following formula:

[0066]

[0067] In the formula: T A，i (t), T B，i (t) represents the time that the i-th unit needs to remain running or shut down in hour t, which is calculated as follows:

[0068]

[0069] In the formula: I A，i I B，i These are the minimum operating interval between two unit shutdowns and the minimum shutdown interval between two startups, respectively.

[0070] The specific steps for constraining renewable energy output are to set a maximum output limit for renewable energy, as shown in the following formula:

[0071]

[0072] in, The amount of available renewable energy resources;

[0073] The various power supply ramping constraints include the following aspects, as shown in the following formula:

[0074]

[0075] Where R is the unit's ramp rate;

[0076] The system hot standby constraint is specifically shown in the following formula:

[0077]

[0078] Among them, XW The number of hydropower units participating in the short-term operation simulation; These represent the available hydropower resources and actual output, respectively; ρ is the thermal reserve coefficient.

[0079] The energy storage constraints are used to constrain the operating state of energy storage devices during short-term operation simulations to meet system requirements and adjust the balance between system load and power supply. Specific upper and lower limits for charging and discharging power are as follows:

[0080]

[0081] in, These refer to the charging and discharging power of the energy storage device, respectively. These are the rated power of the energy storage device and its input power to the grid, respectively. These represent the charging and discharging states of the energy storage device, respectively; k is the margin coefficient.

[0082] The working state constraints are shown in the following formula:

[0083]

[0084] The energy storage constraints for charging and discharging are shown in the following formula:

[0085]

[0086] Among them, E j，t δ represents the actual amount of electricity stored by the energy storage device in hour t; j The energy self-consumption coefficient of the energy storage device; The energy conversion efficiency of the energy storage device during charging and discharging; These represent the maximum and minimum stored capacity of the energy storage device, respectively; the constraint that the capacity of the energy storage device is equal at the beginning and end of its short-term operating cycle is shown in the following formula:

[0087]

[0088] Preferably, the method of solving the complex constraints in the planning scheme evaluation model using the marine predator algorithm includes a high-speed ratio stage, a medium-speed ratio stage, and a low-speed ratio stage, specifically comprising the following steps:

[0089] The positions of predators and prey are initialized, and an elite matrix is ​​constructed with the predators with the best fitness, and a prey matrix is ​​constructed with the prey with a uniform distribution.

[0090] The high-speed ratio stage is the initial stage of algorithm optimization iteration, during which global location information is surveyed. The mathematical model for this stage is as follows:

[0091]

[0092] in, This represents the movement step size; It is a random vector based on a Brownian walk normal distribution; This is a term-by-term multiplication operation; P is a constant, equal to 0.5; The vector is a uniform random vector in [0, 1]; Iter is the current iteration number; Max_Iter is the maximum iteration number; N is the population size;

[0093] The intermediate speed ratio stage represents the middle phase of the algorithm iteration. The population is divided into two parts: prey performs lévy flight, responsible for spatial development, while predators perform Brownian motion, responsible for spatial reconnaissance. The mathematical model expression for this stage is as follows:

[0094]

[0095] in, represents the Lévy motion random vector, and CF represents the adaptive parameter controlling the predator's step size;

[0096] The mid-speed ratio stage, occurring in the later stages of algorithm iteration, involves the predator adopting a levy flight strategy, focusing more on exploiting localized areas. The mathematical model expression for this stage is as follows:

[0097]

[0098] External environmental factors such as fish gathering devices and eddy currents alter predators' foraging strategies to escape local optima, allowing the algorithm to avoid premature convergence. The mathematical model expression is as follows:

[0099]

[0100] Where FADs represent the probability of influence, which is typically taken as 0.2; is a binary vector; r is a random number in [0, 1]; r1 and r2 are random indices of the prey matrix, respectively;

[0101] The Ocean Predator algorithm is used to solve the objective function under complex constraints, including coal-fired power output constraints and power balance constraints, to optimize grid operation, including peak-shaving capacity and economic efficiency. After optimization, the difference between the actual peak-shaving capacity and the expected peak load demand is calculated. If the difference is lower than a pre-set threshold, the power planning and operation strategy is effective and can meet the system's peak-shaving demand, ensuring safe and stable grid operation. Otherwise, it is necessary to readjust the optimization model parameters or increase peak-shaving resources and continue to use the Ocean Predator algorithm for optimization.

[0102] On the other hand, the present invention also provides a rationality evaluation system for power grid investment planning schemes that consider low-carbon goals, comprising:

[0103] The load assessment module collects relevant data and indicators related to power grid investment planning. Based on this data and indicators, it divides the power grid load forecasting area into regions. Using a seasonal decomposition method for time series data, it decomposes the power system load data into three parts: real-time variation, trend-stable components, and irregular fluctuation components. The power system load data is preprocessed to obtain a dataset, which is then divided into a training dataset and a test dataset. Power grid load forecasting models are constructed for each component. The training dataset is used to train the load forecasting model, and the trained model is used to predict samples in the test dataset to obtain load forecasting results. These results are then inversely normalized to obtain the final load forecasting result. The error between the actual load value and the final load forecasting result is compared, and the forecasting model is evaluated based on this error, ultimately yielding the load assessment result.

[0104] The simulation module uses the k-means algorithm to cluster the daily wind and solar power output and load data throughout the year, and selects typical days from each quarter for short-term operation simulation.

[0105] The optimization module constructs a planning scheme evaluation model with the goal of minimizing operating costs, simulates the matching relationship between the output and load curves of various power sources, and optimizes the peak-shaving capacity and economy of the power system, as well as the investment and planning schemes of the power system based on the matching relationship.

[0106] The optimization solution module uses the marine predator algorithm to solve the complex constraints in the planning scheme evaluation model. If all planning schemes meet the system peak-shaving requirements, the planning results are considered reasonable.

[0107] In another aspect, the present invention also provides an electronic device having a computer program stored thereon, which, when executed by a processor, implements a method for evaluating the rationality of a power grid investment planning scheme that takes into account low-carbon objectives, as described in any embodiment of the present invention.

[0108] In another aspect, the present invention also provides a computer-readable medium for storing one or more programs, which, when executed by one or more processors, cause the one or more processors to implement a method for evaluating the rationality of a power grid investment planning scheme that takes into account low-carbon objectives, as described in any embodiment of the present invention.

[0109] The present invention has the following beneficial effects:

[0110] 1. This invention achieves high-precision forecasting of power grid load through comprehensive data collection and analysis. The diverse data sources, including but not limited to regional GDP, population size, power grid asset scale, geographical area, and load density, provide a rich information foundation for load forecasting, ensuring the comprehensiveness and accuracy of the model. By precisely dividing the forecast area and combining it with the seasonal decomposition method of time series analysis, load data is decomposed into three parts: trend, seasonality, and random fluctuation. For each component, a forecasting model is established. This method not only captures the inherent patterns of load changes but also effectively improves the accuracy of the forecast. The combination of refined regional division and component forecasting models makes load forecasting closer to reality, providing a solid foundation for subsequent planning.

[0111] 2. This invention utilizes the K-means clustering algorithm to intelligently cluster annual wind and solar power output and load data, effectively identifying typical days for each quarter. This strategy greatly simplifies the complexity of operational simulations while ensuring their representativeness. Through in-depth analysis and simulation of typical days, the operating status of the power grid under different seasons and weather conditions can be more accurately assessed, providing important basis for optimizing peak-shaving strategies and improving the utilization rate of renewable energy sources such as wind and solar power, thus promoting the flexibility and efficiency of the power system.

[0112] 3. The planning scheme evaluation model constructed in this invention aims to minimize operating costs while considering multiple factors such as peak-shaving capacity, economic efficiency, and carbon emissions, demonstrating systematic and comprehensive characteristics. By simulating the matching relationship between the output and load curves of various power sources, the model optimizes power source configuration and dispatch strategies, not only improving the system's peak-shaving capacity but also effectively reducing operating costs, achieving a dual optimization of economic and environmental benefits. This multi-objective optimization from a global perspective is not available in traditional single-objective planning methods, opening up a new path for achieving sustainable development of the power grid.

[0113] 4. A major highlight of this patented technology is the use of the marine predator algorithm to solve the multi-objective model. Compared to traditional optimization algorithms, the marine predator algorithm, with its unique bio-inspired mechanism, exhibits better global search capabilities and faster convergence speed, effectively addressing the complex and variable factors and constraints involved in power grid planning. Through the introduction of this algorithm, the system can quickly find the optimal or near-optimal solution that meets multiple constraints such as peak-shaving requirements, economic efficiency, and low carbon emissions, ensuring the rationality and feasibility of the planning results.

[0114] 5. The technical framework of this invention emphasizes dynamic adaptability to market changes, policy adjustments, and technological advancements. Through rolling planning and real-time data updates, the model can promptly adjust prediction model parameters, optimize strategies, and make investment decisions, ensuring the foresight and flexibility of the planning scheme. This dynamic optimization mechanism ensures the long-term effectiveness of power grid investment planning and its resilience to uncertainties, which is crucial for achieving stable power grid operation and low-carbon transformation. Attached Figure Description

[0115] Figure 1 This is a schematic diagram of the process of the present invention;

[0116] Figure 2 This is a schematic diagram of the neural network structure according to an embodiment of the present invention;

[0117] Figure 3 This is a schematic diagram of the neural network concept in an embodiment of the present invention;

[0118] Figure 4 This is an optimized schematic diagram of an embodiment of the present invention. Detailed Implementation

[0119] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0120] It should be understood that the step numbers used in the text are for ease of description only and are not intended to limit the order in which the steps are performed.

[0121] It should be understood that the terminology used in this specification is for the purpose of describing particular embodiments only and is not intended to limit the invention. As used in this specification and the appended claims, the singular forms “a,” “an,” and “the” are intended to include the plural forms unless the context clearly indicates otherwise.

[0122] The terms “comprising” and “including” indicate the presence of the described feature, whole, step, operation, element and / or component, but do not exclude the presence or addition of one or more other features, wholes, steps, operations, elements, components and / or collections thereof.

[0123] The term “and / or” refers to any combination of one or more of the associated listed items, as well as all possible combinations, and includes these combinations.

[0124] Example 1:

[0125] See Figure 1-4A method for evaluating the rationality of power grid investment planning schemes that consider low-carbon goals includes the following steps:

[0126] S10. Collect relevant data and indicators related to power grid investment planning. Based on the relevant data and indicators, divide the power grid load forecasting area into regions. Use the seasonal decomposition method of time series to decompose the power system load data into three parts: real-time variation, trend-stable component, and irregular fluctuation component. Preprocess the power system load data to obtain a dataset. Divide the preprocessed dataset into a training dataset and a test dataset. Build power grid load forecasting models for different components. Train the load forecasting model using the divided training dataset. Use the trained load forecasting model to predict samples in the test dataset to obtain load forecasting results. Perform inverse normalization on the load forecasting results to obtain the final load forecasting results. Compare the actual load value and the final load forecasting results to obtain the error between the two. Evaluate the forecasting model based on the error and finally obtain the load assessment results.

[0127] S20. Use the k-means algorithm to cluster the daily wind and solar power output and load data throughout the year, and select typical days of each quarter for short-term operation simulation;

[0128] S30. Construct a planning scheme evaluation model with the goal of minimizing operating costs, simulate the matching relationship between the output and load curves of various power sources, and optimize the peak-shaving capacity and economy of the power system, as well as the investment and planning scheme of the power system based on the matching relationship.

[0129] S40. Solve the complex constraints in the planning scheme evaluation model using the marine predator algorithm. If all planning schemes meet the system peak-shaving requirements, the planning results are considered reasonable.

[0130] As a preferred embodiment of this example, in steps S10-S40, relevant data is first collected to predict the power grid load. The load prediction area is accurately divided by combining indicators such as regional GDP, population, power grid asset scale, geographical area, and load density to improve the accuracy of load prediction. Then, the seasonal decomposition method of time series is used to decompose the power system load data into three parts: real-time variation, trend stable component, and irregular fluctuation component. Then, prediction models are constructed for different components to finally generate the load prediction result.

[0131] The main supporting documents and related data for the planning are collected, including: local statistical yearbooks, national economic and social development plans, urban and rural master plans, land use master plans, power grid operation and statistical basic data, and planning results from higher-level power grids. Based on the collected data, regional divisions are conducted using indicators such as regional GDP, population, power grid asset scale, geographical area, and load density to improve the accuracy of load forecasting.

[0132] By using cluster analysis and combining the aforementioned relevant indicators, regions and user types are divided, and load forecasting is carried out in a targeted manner.

[0133] Assume c is the number of clusters. It is a set of vectors in the feature space, where v i (i = 1, 2, ...) represents the cluster prototype vector of the i-th class. ik ∈U, where U is a c×n fuzzy matrix.

[0134] The matrix satisfies the following condition:

[0135]

[0136]

[0137] The objective function of the improved fuzzy clustering algorithm is defined as follows:

[0138] max{J(U, V)}

[0139]

[0140] in μ ik For data point x k Membership degree of cluster i; μ ik For data point x k The membership degree V of cluster j is the clustering result;

[0141] Given the number of clusters c (2≤c≤n, where n is the number of data points); set the iteration stopping threshold ε and the fuzzy index m. Initialize the clustering prototype mode v(0); set the iteration counter b = 0.

[0142] For any i, if k exists d ik b >0, then we have

[0143]

[0144] If there exists i, r such that d ik b =0, then μ ir (b) =1, and for j≠i, μir (b) =0

[0145]

[0146] If ||V (b+1) -V (b) If || < ε, the improved fuzzy clustering algorithm stops. It outputs the partition matrix U and the cluster prototype V; otherwise, it sets b = b + 1 and returns to the previous step.

[0147] Determine the optimal classification tree C m Calculate the value of J(C). If the condition J(Cm) = max{J(C)} is satisfied, then Cm is considered the optimal number of classifications. The corresponding output partition matrix U and cluster prototype are then obtained, representing the final clustering result.

[0148] To forecast power grid load, the seasonal decomposition method of time series is used to decompose the power system load data into three parts: real-time variation, trend-stable component, and irregular fluctuation component. Then, a forecast model is built for each component, and the load forecast result is finally generated.

[0149] The general steps for load forecasting are as follows.

[0150] (1). Load data preprocessing.

[0151] (2). Divide the preprocessed dataset into two parts: training dataset and test dataset.

[0152] (3) Constructing a forecasting model. Select and construct the appropriate load forecasting model according to the requirements of load forecasting.

[0153] (4). Sample training. The load prediction model is trained using the training samples divided in step 2.

[0154] (5). Load forecasting. The load forecasting model trained in step four is used to forecast the samples in the test dataset. The load forecast value is the model output value.

[0155] (6) Forecast Result Processing and Evaluation. The load forecast results from the previous step are denormalized to obtain the final load forecast result. Subsequently, the forecast model is evaluated using the error between the actual load value and the forecast value.

[0156] The seasonal decomposition method for time series data is used to decompose power system load data into three parts: real-time variation, trend-stable components, and irregular fluctuation components. For time series regression, seasonal and trend decomposition is a process that improves the attributes of parameter estimation. Seasonal and trend adjustments often help to better understand time series data. For example, if electricity consumption in June increased by 20% compared to May, decomposing the data makes it easier to see that the increase in electricity consumption is mainly caused by weather-related seasonal effects. Therefore, decomposing the original load series into better-interpretable subsequences can yield better load forecasting results.

[0157] For time series regression, seasonal and trend decomposition is a process that improves the attributes of parameter estimation. Seasonal and trend adjustments often help to better understand time series data. For example, if electricity consumption in June increased by 20% compared to May, decomposing the data makes it easier to see that the increase in electricity consumption is mainly caused by weather-related seasonal effects. Therefore, decomposing the original load series into better-interpreted subsequences can yield better load forecasting results.

[0158] Y t This is represented as the actual time series value for period t. The time series seasonal trend adjustment method will use Y... t It is decomposed into three components: a trend component, a seasonal component, and an irregularity component. These three components are represented by an additive or multiplicative model.

[0159] Composed of various elements.

[0160] The additive decomposition model is defined as follows:

[0161] Y t =Trend t +Seasonal t +Irregular t

[0162] The multiplicative factorization model is defined as follows:

[0163] Y t =Trend t ×Seasonal t ×Irregular t

[0164] In the above two equations, Trend t The trend component over time period t, Seasonal t It refers to the seasonal component within the time period t.

[0165] Irregular t It is the irregular component in the time period t.

[0166] Multiplicative models are suitable when seasonal fluctuations are proportional to the level of the time series. When seasonal fluctuations do not depend on the level of the time series, additive decomposition models are preferred.

[0167] In addition to the two basic decomposition models mentioned above, a decomposition model can also be defined as:

[0168] Y t =Seasonal t +Trend t and Irregular t

[0169] When seasonal components change slowly, the above model can be used to decompose the time series. Therefore, non-seasonal forecasting models can be applied to combinations of trend and irregular components. Similar models are also applicable to multiplicative models.

[0170] Y t =Seasonal t ×Trend t and Irregular t

[0171] Using a feedforward neural network method, prediction models are constructed for the three components mentioned above, and finally superimposed to obtain the final load prediction result.

[0172] Using a feedforward neural network method, prediction models are constructed for the three components mentioned above, and finally superimposed to obtain the final load prediction result.

[0173] Let X1, X2, ..., X n Let Y1, Y2, ..., Y be the input vectors of the BP neural network. m For the output value, w ij and w jk The weights are shown in the diagram below. A typical BP neural network topology is illustrated in the diagram below. Figure 1 As shown.

[0174] When the number of input and output nodes in a BP neural network is n and m respectively, it reflects the mapping relationship between n independent variables and m dependent variables. BP neural network modeling and prediction includes three steps: network structure construction, training, and prediction. The basic workflow is as follows: Figure 3 As shown.

[0175] Suppose that the activation function of each node in the network is a sigmoid function, and the input of the first node i in the network is denoted as net. i The output is denoted as o. i The output of the kth node in the output layer is y k Then the input of the j-th node in the intermediate layer is:

[0176]

[0177] o j =f(net) j )

[0178]

[0179] If we define the network error as the difference between the expected output and the actual output, then we have: If the output layer has i neurons, the squared error between the actual output and the expected output is defined as:

[0180]

[0181] Since the BP algorithm corrects the weights according to the negative gradient of the error E, the modification of the weights can be expressed as:

[0182] W m+1 =w m +Δw m =wm-λg m

[0183] Where m represents the number of iterations.

[0184] Where λ is the learning step size.

[0185]

[0186] Because it is the output layer, at this time... This is the actual output value, based on e. k From the definition and the squared error, we can obtain:

[0187]

[0188] According to e k From the definition, we can obtain:

[0189]

[0190] According to the above formula We can obtain:

[0191]

[0192] According to the above formula, net k =∑ j w kj o j We can obtain:

[0193]

[0194] The final result is:

[0195]

[0196] Now let the learning error of the output layer be:

[0197] σ k =e k f′(net k )

[0198] have to:

[0199]

[0200] (2) Modification of hidden layer neural unit weights Δw kj :

[0201]

[0202] According to the above formula, net j =∑ i w ji o i and o j =f(net) j We can obtain:

[0203]

[0204] Since we are calculating the changes in the weights of the hidden layers, we should consider the effect of the previous layer. Therefore:

[0205]

[0206] According to net k =∑ j w kj o j It can be known

[0207]

[0208] According to We can obtain:

[0209]

[0210] Bundle Attached Derivation:

[0211]

[0212] Let the learning error of the hidden layer be:

[0213]

[0214] The model is then optimized using a genetic algorithm.

[0215] We further utilized the k-means algorithm to cluster the daily wind and solar power output and load data throughout the year, and selected typical days from each quarter for short-term operation simulation.

[0216] K-means clustering is an iterative clustering analysis method designed to divide a dataset into K clusters such that each data point belongs to the centroid of its nearest cluster. Its main objective is to minimize the sum of squared errors from each data point to its cluster centroid. Below are the basic steps and related formulas of the K-means algorithm:

[0217] Initialization: Randomly select K data points as initial centroids (μ1, μ2, ..., μk).

[0218] Assignment: For each data point x in the dataset, calculate its distance to all centroids and assign it to the cluster containing the nearest centroid.

[0219] Distance calculations typically use Euclidean distance:

[0220]

[0221] Where xx is an n-dimensional data point, μi is the i-th centroid, and xj and μij are their values ​​in the j-th dimension, respectively.

[0222] Update centroids: For each cluster Ci, recalculate the mean of all its members as the new centroids:

[0223]

[0224] Here, |Ci| represents the number of data points in cluster Ci.

[0225] Repeat: Repeat steps 2 and 3 until the centroid no longer changes significantly or the preset maximum number of iterations is reached.

[0226] Result: K clusters are obtained in the end, each cluster consists of its centroid and all data points belonging to that cluster.

[0227] Stopping criteria:

[0228] Centroid position change less than threshold: If the distance moved by all centroids is less than a very small threshold ε, then the iteration stops.

[0229] Maximum Iterations: A maximum number of iterations is preset. The algorithm stops when this number is reached, even if the centroids are still changing. The center point of each cluster represents a "typical day" for a quarter, which contains typical characteristics of wind power, solar power generation, and grid load within that quarter. This typical day data can be used for simulation and analysis, reducing the amount of data processing required during actual simulation operation while maintaining high representativeness.

[0230] Further, an evaluation model for planning schemes aimed at minimizing operating costs will be constructed to simulate the matching relationship between the output and load curves of various power sources, thereby optimizing the system's peak-shaving capacity and economy, and optimizing the investment and planning schemes of the power system.

[0231] 1. Objective function

[0232] The operation simulation layer uses the minimum operating cost as the objective function to verify the system's peak-shaving capability. Considering that the system operates with base load, is stable, and typically does not participate in peak-shaving scheduling, the operating cost F2 in the operation simulation layer includes three aspects: ① fuel costs (mainly coal and natural gas) and the start-up and shutdown costs of coal-fired power units. 1t ② Energy storage device charging, discharging, and maintenance costs F 2t ③ Interruptible load compensation costs and wind / solar curtailment penalty costs F 3t .

[0233]

[0234] In the formula: t represents the number of hours, T represents the total number of hours of short-term simulation; X RA X CA X KA These represent the number of fuel cell units, energy storage units, and renewable energy units participating in the short-term operation simulation, respectively. These are the fuel cost, start-up cost, and shutdown cost for the i-th fuel unit, respectively. These represent the actual output, startup state variables, and shutdown state variables of the i-th fuel unit in hour t, respectively. These are the unit discharge cost, charging cost, and maintenance cost for the j-th energy storage device, respectively. These are the real-time discharge power, charging power, and actual output of the j-th energy storage device in hour t, respectively. For interruptible load; ρ IL ρ ka These are the interruptible load compensation cost coefficient and the wind and solar curtailment factor, respectively. These represent the available power and actual output of the k-th energy storage device in hour t, respectively.

[0235] Mode Fuel costs in China can be expressed as a quadratic function:

[0236]

[0237] In the formula: a i b i c i These are the consumption coefficients of the i-th fuel unit.

[0238] 2. Constraints

[0239] (1) Power balance constraint

[0240] Power balance constraints refer to the requirement that the output of all power sources and the consumption of all loads in a power system must remain in balance. At this stage, the constraint that the sum of the output of all power sources equals the total load is crucial for ensuring the stable operation of the power system. If the output power of a power source is lower than the power consumed by the load, it will lead to power system imbalance, potentially causing grid faults or even power outages.

[0241]

[0242] In the formula: These represent the active power output of the fuel cell unit, energy storage unit, and renewable energy unit in hour t, respectively. Let t be the interruptible load of the system participating in the short-term operation simulation at hour t. Let t be the system load demand in hour t.

[0243] (2) Fuel unit output constraints

[0244] During the operation of a power system, the real-time output of fuel units must meet upper and lower limit constraints.

[0245]

[0246] In the formula: U i，t The generator unit is in start-up or shutdown status. These are the minimum and maximum outputs of the fuel unit, respectively. This represents the actual output of the fuel unit in hour t.

[0247] (3) Unit start-up and shutdown time constraints

[0248] Start-up and shutdown times are related to the unit's flexible adjustment capabilities and play an important role in coping with fluctuations in power demand and improving the consumption of clean energy. Generally, the shorter the start-up and shutdown time, the higher the reliability of the system. However, too frequent start-ups and shutdowns can also affect the lifespan of the generator set.

[0249] Therefore, a time interval should be allowed between each start-up and shutdown of the unit.

[0250]

[0251] In the formula: T A，i (t), T B，i (t) represents the time that the i-th unit needs to remain running or shut down in hour t, which is calculated as follows:

[0252]

[0253] In the formula: I A，i I B，i These are the minimum operating interval between two unit shutdowns and the minimum shutdown interval between two unit startups, respectively.

[0254] (4) Renewable energy output constraints

[0255] The output of renewable energy sources such as wind, solar, and hydropower is affected by weather and the environment, thus their maximum output is also volatile and uncertain. In short-term operation simulations, it is necessary to set maximum output limits for renewable energy sources to ensure the stable operation and safety of the system.

[0256]

[0257] In the formula: This refers to the amount of available renewable energy resources.

[0258] (5) Various power supply ramping constraints

[0259] In short-term power system operation simulations, ramp-up constraints for various power sources are crucial. These constraints refer to the time required for the output of each generator unit in the power system to increase from zero to its rated power. Ramp-up constraints for various power sources typically include the following aspects:

[0260]

[0261] In the formula: R is the unit's ramp rate.

[0262] (6) System hot standby constraints

[0263] System hot standby constraints are an important constraint in short-term power system operation simulations. They are used to ensure the system can quickly dispatch backup generators in the event of sudden load fluctuations or generator failures, thus guaranteeing system reliability. Common hot standby generators include coal-fired units and hydroelectric generators.

[0264]

[0265] In the formula: X W The number of hydropower units participating in the short-term operation simulation; P l W,max , These represent the available hydropower resources and actual output, respectively; ρ is the thermal reserve coefficient.

[0266] (7) Energy storage constraints

[0267] In short-term operation simulations, the operating state of energy storage devices needs to be constrained to ensure they can meet system demands and are more efficient in balancing system load and power supply. Upper and lower limits for charge / discharge power constraints:

[0268]

[0269] In the formula: These refer to the charging and discharging power of the energy storage device, respectively. These are the rated power of the energy storage device and its input power to the grid, respectively. These represent the charging and discharging states of the energy storage device, respectively; k is the margin coefficient.

[0270] Working status constraints:

[0271]

[0272] Charge and discharge energy storage constraints:

[0273]

[0274] In the formula: E j,t δ represents the actual amount of electricity stored by the energy storage device in hour t; j The energy self-consumption coefficient of the energy storage device; The energy conversion efficiency of the energy storage device during charging and discharging; These represent the maximum and minimum stored capacity of the energy storage device, respectively. The energy storage device is subject to the constraint that its capacity remains equal throughout its short-term operating cycle.

[0275]

[0276] Finally, the multi-objective model is solved using the marine predator algorithm. If all planning schemes meet the system's peak-shaving requirements, the planning results are considered reasonable.

[0277] The specific solution steps are as follows:

[0278] The marine predator algorithm is a novel metaheuristic optimization algorithm proposed by Afshin Faramarzi et al. in 2020. It simulates the natural law of survival of the fittest in the ocean, in which various marine organisms constantly change their roles as predators and prey, and change their foraging strategies according to different situations, thereby optimizing the process.

[0279] The marine predator algorithm posits that marine predators employ a foraging strategy that alternates between Lévy flight and Brownian walk, choosing between the two strategies based on different scenarios to arrive at the optimal foraging strategy.

[0280] First, the positions of predators and prey are initialized. An elite matrix is ​​constructed with the predators with the best fitness, and a prey matrix is ​​constructed with the prey with a uniform distribution, as shown in the following formulas.

[0281] Based on different speed ratios, the optimization process of MPA is divided into three stages.

[0282] Phase 1: Survey Phase

[0283] This stage is also known as the high-speed ratio stage. In this stage, the prey is much faster than the predator. The predator adopts a stationary strategy, while the prey undergoes Brownian motion. This stage often occurs in the early stages of algorithm optimization iterations and is used to survey global positional information. The mathematical model for this stage is as follows:

[0284]

[0285] Remember to modify SPE and remember that i = 1, n

[0286] in This represents the movement step size; It is a random vector based on a Brownian walk normal distribution; This is a term-by-term multiplication operation; P is a constant, equal to 0.5; is a uniform random vector in [0,1]; Iter is the current iteration number; Max_Iter is the maximum iteration number; N is the population size.

[0287] Phase 2: Medium Speed ​​Ratio Phase

[0288] This stage, also known as the mid-speed ratio stage, involves predators and prey moving at roughly equal speeds, both searching for their prey. This stage typically occurs in the middle of the algorithm's iterations, where the population is divided into two parts: prey performs Lévy flight, responsible for spatial development, while predators perform Brownian motion, responsible for spatial reconnaissance. The mathematical model for this stage is as follows:

[0289]

[0290]

[0291] in, Let represent the Lévy motion random vector, and CF represent the adaptive parameter controlling the predator's step size.

[0292] Phase 3: Development Phase

[0293] This stage, also known as the levy stage, occurs when the predator's speed exceeds that of the prey. It primarily happens in the later stages of the algorithm's iteration, where the predator adopts a levy flight strategy, focusing more on exploiting localized areas. The mathematical model for this stage is as follows:

[0294]

[0295] In addition, the algorithm also considers external environmental factors such as fish gathering devices and eddy currents, altering the predator's foraging strategy to escape local optima and avoid premature convergence. Its mathematical model expression is as follows:

[0296]

[0297] Where FADs represent the probability of influence, typically taken as 0.2; is a binary vector; r is a random number within [0,1]; r1 and r2 are random indices of the prey matrix, respectively.

[0298] The Marine Predators Algorithm (MPA) is used to solve the objective function under complex constraints, including coal-fired power output and power balance constraints, aiming to optimize power grid operation, particularly peak-shaving capacity and economic efficiency. After obtaining the optimal solution, the gap between the actual peak-shaving capacity and the expected peak load demand (peak-shaving deficit) is calculated. If this value is lower than a pre-set threshold, it indicates that the power planning and operation strategy is effective and can meet the system's peak-shaving demand, ensuring safe and stable power grid operation; otherwise, the optimization model parameters need to be readjusted or peak-shaving resources increased, and optimization continues.

[0299] Example 2:

[0300] The load assessment module collects relevant data and indicators related to power grid investment planning. Based on this data and indicators, it divides the power grid load forecasting area into regions. Using a seasonal decomposition method for time series data, it decomposes the power system load data into three parts: real-time variation, trend-stable component, and irregular fluctuation component. The power system load data is preprocessed to obtain a dataset, which is then divided into a training dataset and a test dataset. Power grid load forecasting models are constructed for different components. The training dataset is used to train the load forecasting model. The trained load forecasting model is then used to predict samples in the test dataset to obtain load forecasting results. These results are then inversely normalized to obtain the final load forecasting result. The error between the actual load value and the final load forecasting result is compared, and the forecasting model is evaluated based on this error to obtain the final load assessment result. This module implements the function of step S10 in implementation one above, and its specific details will not be elaborated further.

[0301] The simulation module uses the k-means algorithm to cluster the daily wind and solar power output and load data throughout the year, and selects typical days of each quarter for short-term operation simulation; this module is used to implement the function of step S20 in implementation one above, and the details will not be elaborated further.

[0302] The optimization module constructs a planning scheme evaluation model with the goal of minimizing operating costs, simulates the matching relationship between the output and load curves of various power sources, and optimizes the peak-shaving capacity and economy of the power system, as well as the investment and planning scheme of the power system based on the matching relationship. This module is used to implement the function of step S30 in the above implementation one, and the details will not be elaborated further.

[0303] The optimization solution module solves the complex constraints in the planning scheme evaluation model using the marine predator algorithm. If all planning schemes meet the system peak-shaving requirements, the planning results are considered reasonable. This module is used to implement the function of step S40 in the above implementation, and the details will not be elaborated further.

[0304] Example 3:

[0305] This embodiment provides an electronic device that stores a computer program. When the computer program is executed by a processor, it implements a method for evaluating the rationality of a power grid investment planning scheme that considers low-carbon goals, as described in any embodiment of the present invention.

[0306] Example 4:

[0307] This embodiment provides a computer-readable medium for storing one or more programs, which, when executed by one or more processors, cause the one or more processors to implement a method for evaluating the rationality of a power grid investment planning scheme that considers low-carbon objectives, as described in any embodiment of the present invention.

[0308] In this application embodiment, "at least one" refers to one or more, and "more than one" refers to two or more. "And / or" describes the relationship between related objects, indicating that three relationships can exist. For example, A and / or B can represent the existence of A alone, A and B simultaneously, or B alone. A and B can be singular or plural. The character " / " generally indicates that the preceding and following related objects are in an "or" relationship. "At least one of the following" and similar expressions refer to any combination of these items, including any combination of singular or plural items. For example, at least one of a, b, and c can represent: a, b, c, a and b, a and c, b and c, or a and b and c, where a, b, and c can be single or multiple.

[0309] Those skilled in the art will recognize that the units and algorithm steps described in the embodiments disclosed herein can be implemented using electronic hardware, computer software, or a combination of electronic hardware and software. Whether these functions are implemented in hardware or software depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementation should not be considered beyond the scope of this application.

[0310] Those skilled in the art will understand that, for the sake of convenience and brevity, the specific working processes of the systems, devices, and units described above can be referred to the corresponding processes in the foregoing method embodiments, and will not be repeated here.

[0311] In the several embodiments provided in this application, any function, if implemented as a software functional unit and sold or used as an independent product, can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of this application, in essence, or the part that contributes to the prior art, or a part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of this application. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0312] The above description is merely an embodiment of the present invention and does not limit the patent scope of the present invention. Any equivalent structural or procedural transformations made based on the content of the present invention's specification and drawings, or direct or indirect applications in other related technical fields, are similarly included within the patent protection scope of the present invention.

Claims

1. A method for evaluating the rationality of power grid investment planning schemes that consider low-carbon objectives, characterized in that, Includes the following steps: Relevant data and indicators related to power grid investment planning are collected. Based on the relevant data and indicators, the power grid load forecasting area is divided into regions. The seasonal decomposition method of time series is used to decompose the power system load data into three parts: real-time variation, trend stability component, and irregular fluctuation component. The power system load data is preprocessed to obtain a dataset. The preprocessed dataset is divided into two parts: a training dataset and a test dataset. Power grid load forecasting models are built for different components. The load forecasting model is trained using the divided training dataset. The trained load forecasting model is used to predict the samples in the test dataset to obtain the load forecasting results. The load forecasting results are inversely normalized to obtain the final load forecasting results. By comparing the actual load value with the final load forecast result, the error between the two is obtained. The forecast model is evaluated based on the error, and the final load assessment result is obtained. The k-means algorithm was used to cluster the daily wind and solar power output and load data throughout the year, and typical days of each quarter were selected for short-term operation simulation. Construct a planning scheme evaluation model with the goal of minimizing operating costs, simulate the matching relationship between the output and load curves of various power sources, and optimize the peak-shaving capacity and economy of the power system, as well as the investment and planning schemes of the power system based on the matching relationship. The complex constraints in the planning scheme evaluation model are solved by the marine predator algorithm. If all planning schemes meet the system peak-shaving requirements, the planning results are considered reasonable. The complex constraints in the planning scheme evaluation model are solved using the marine predator algorithm, including the high-speed ratio stage, the medium-speed ratio stage, and the low-speed ratio stage. The specific steps are as follows: The positions of predators and prey are initialized, and an elite matrix is ​​constructed with the predators with the best fitness, and a prey matrix is ​​constructed with the prey with a uniform distribution. The high-speed ratio stage is the initial stage of algorithm optimization iteration, during which global location information is surveyed. The mathematical model for the high-speed ratio stage is as follows: ； ； ； in, This represents the movement step size; It is a random vector based on a Brownian walk normal distribution; This is a term-by-term multiplication operation; It is a constant, equal to 0.5; It is a uniform random vector in [0, 1]; This represents the current iteration number; This represents the maximum number of iterations. Population size; The intermediate speed ratio stage represents the middle phase of the algorithm iteration. The population is divided into two parts: prey performs lévy flight, responsible for spatial development, while predators perform Brownian motion, responsible for spatial reconnaissance. The mathematical model expression for this stage is as follows: ； ； ； ； ； ； ； in, Represents the Lévy motion random vector. This represents an adaptive parameter that controls the predator's step size. The medium-speed ratio stage is the later stage of algorithm iteration, during which the predator adopts a levy flight strategy and focuses more on developing local areas. The mathematical model expression for this stage is as follows: ； ； ； The external environmental factors of fish gathering devices and eddy current effects alter the predator's foraging strategy to escape local optima, allowing the algorithm to avoid premature convergence. Its mathematical model expression is as follows: ； in, To minimize the impact on probability, we set it to 0.2; It is a binary vector; A random number within the range [0, 1]; , These are random indices of the prey matrix; The Ocean Predator algorithm is used to solve the objective function under complex constraints, including coal-fired power output constraints and power balance constraints, to optimize grid operation, including peak-shaving capacity and economic efficiency. After optimization, the difference between the actual peak-shaving capacity and the expected peak load demand is calculated. If the difference is lower than a pre-set threshold, the power planning and operation strategy is effective and can meet the system's peak-shaving demand, ensuring safe and stable grid operation. Otherwise, it is necessary to readjust the optimization model parameters or increase peak-shaving resources and continue to use the Ocean Predator algorithm for optimization.

2. The method for evaluating the rationality of a power grid investment planning scheme considering low-carbon objectives as described in claim 1, characterized in that: The relevant data include: local statistical yearbooks, national economic and social development plans, urban and rural master plans, land use master plans, power grid operation data, basic statistical data, and planning results of higher-level power grids; The indicators include: regional GDP, population, power grid asset size, geographical area, and load density; Cluster analysis was used to divide the power grid load forecasting area into regions, combining relevant data and indicators. The specific steps included: Assume c is the number of clusters. It is a set of vectors in the feature space, where Indicates the first Cluster prototype vector of the class; It is A fuzzy matrix that satisfies the following conditions: ； ； ； The objective function of the improved fuzzy clustering algorithm is defined as follows: ； in ; For data points The membership degree of cluster i; For data points The membership degree V of cluster j is the clustering result; Given the number of clusters , This refers to the number of data points; setting the iteration stop threshold. and fuzzy index Initialize the clustering prototype pattern Set the iteration counter ; For any , If it exists Then we have: ； If it exists Make Then there is And for , ; ； like The improved fuzzy clustering algorithm then stops, and the partition matrix is ​​output. and clustering prototype Otherwise And iteratively execute the improved fuzzy clustering algorithm; Determine the best classification tree ;calculate The value of , when the condition is met: Then it is believed This represents the optimal number of categories.

3. The method for evaluating the rationality of power grid investment planning schemes considering low-carbon objectives according to claim 1, characterized in that, The steps for decomposing power system load data into three parts—real-time variation, trend-stable component, and irregular fluctuation component—include: Decompose the original load sequence in the power system load data into subsequences; Represented as The actual time series value for the period, adjusted by the seasonal trend method for time series. It is decomposed into three components: seasonal component, trend component, and irregular component; these three components are combined by an additive or multiplicative model, as shown below: The additive decomposition model is defined as follows: ； The multiplicative factorization model is defined as follows: ； In the above two equations, During the time period The trend component, During the time period Seasonal portion During the time period Irregular components; A multiplicative model is used when seasonal fluctuations are proportional to the level of the time series, and an additive decomposition model is used when seasonal fluctuations do not depend on the level of the time series. Using a feedforward neural network method, prediction models are constructed for the above decomposition models respectively, and finally superimposed to obtain the final load prediction result.

4. The method for evaluating the rationality of a power grid investment planning scheme considering low-carbon objectives as described in claim 1, characterized in that, The method involves using the k-means algorithm to cluster daily wind and solar power output and load data throughout the year, and selecting typical days from each quarter for short-term operational simulations. The specific steps are as follows: Data collection and preprocessing: First, collect wind and solar power output data and grid load data for every day of the year. The wind and solar power output data includes wind power generation corresponding to wind speed and photovoltaic power generation corresponding to sunshine duration and intensity. The data is then cleaned, outliers are removed, and the data is standardized. Determine the number of clusters: Select the K value, which is predetermined by the elbow rule and the silhouette coefficient. For power system data with obvious seasonality, the K value is set to 4, corresponding to the typical days of each quarter of the year. Initialization and Iteration: The K-means algorithm starts by randomly selecting K center points and iterates through the following steps until convergence: Assign each data point to the category of the nearest centroid; recalculate the centroid of each category until the change in the centroid is less than a preset threshold or the maximum number of iterations is reached. Typical day selection: The center point of each cluster represents a typical day of a quarter, which includes the typical characteristics of wind power, solar power generation and grid load in that quarter; the typical day data is used for simulation and analysis to reduce the amount of data processing during actual operation simulation.

5. The method for evaluating the rationality of a power grid investment planning scheme considering low-carbon objectives as described in claim 1, characterized in that, The construction of the planning scheme evaluation model with the goal of minimizing operating costs includes the objective function and constraints; The simulation layer is used to verify the system's peak-shaving capability with the objective function of minimizing operating costs; the operating cost of the simulation layer... It includes three aspects: fuel costs and the start-up and shutdown costs of coal-fired power units. Energy storage device charging, discharging, and maintenance costs Interruptible load compensation costs and wind / solar curtailment penalty costs Specifically, as shown in the following formula: In the formula: For hours This represents the total number of hours for short-term simulation. , , These represent the number of fuel cell units, energy storage units, and renewable energy units participating in the short-term operation simulation, respectively. , , The first Fuel costs, start-up costs, and shutdown costs for fuel-powered generator units; , , The first The fuel unit was in Actual output per hour, startup status variables, and shutdown status variables; , , The first The unit discharge cost, charging cost, and maintenance cost of an energy storage device; , , The first The energy storage device in the first Real-time discharge power, charging power, and actual output power per hour; Interruptible load; , These are the interruptible load compensation cost coefficient and the wind and solar curtailment factor, respectively. , The first The energy storage device in the first Available power per hour and actual output; Mode The fuel cost in China can be expressed as a quadratic function: In the formula: , , The first Consumption coefficient of fuel unit.

6. The method for evaluating the rationality of a power grid investment planning scheme considering low-carbon objectives according to claim 5, characterized in that, The constraints specifically include: power balance constraints, fuel unit output constraints, unit start-up and shutdown time constraints, renewable energy output constraints, various power source ramping constraints, system hot standby constraints, and energy storage constraints. The power balance constraint requires that, in a power system, the output of all power sources and the consumption of all loads must be kept in balance, as shown in the following formula: In the formula: , , Fuel cell generators, energy storage generators, and renewable energy generators were respectively in the [number]th [year]. Hours of hard work and effort; For the system in the first The interruptible load of the system participating in short-term operation simulation for one hour; In the first The system's load demand per hour; The output constraint of the fuel unit is that the real-time output of the fuel unit must meet the upper and lower limit constraints during the operation of the power system, as shown in the following formula: In the formula: The generator unit is in start-up or shutdown status. , These are the minimum and maximum outputs of the fuel unit, respectively. For fuel units in the first Actual output per hour; The unit start-up and shutdown time constraints are specifically shown in the following formula: ； In the formula: , The first Taiwanese unit in The required operating or shutdown time per hour is calculated as follows: In the formula: , These are the minimum operating interval between two unit shutdowns and the minimum shutdown interval between two startups, respectively. The specific steps for constraining renewable energy output are to set a maximum output limit for renewable energy, as shown in the following formula: ； in, The amount of available renewable energy resources; The various power supply ramping constraints include the following aspects, as shown in the following formula: ； in, This refers to the unit's ramp rate; The system hot standby constraint is specifically shown in the following formula: ； in, The number of hydropower units participating in the short-term operation simulation; , These are the available hydropower resources and actual output, respectively. This is the hot reserve coefficient; The energy storage constraints are used to constrain the operating state of energy storage devices during short-term operation simulations to meet system requirements and adjust the balance between system load and power supply. Specific upper and lower limits for charging and discharging power are as follows: ； in, , These refer to the charging and discharging power of the energy storage device, respectively. , These are the rated power of the energy storage device and its input power to the grid, respectively. , These represent the charging and discharging states of the energy storage device, respectively. This is the margin coefficient; The working state constraints are shown in the following formula: The energy storage constraints for charging and discharging are shown in the following formula: in, For energy storage devices in the first The actual stored power in hours; The energy self-consumption coefficient of the energy storage device; , The energy conversion efficiency of the energy storage device during charging and discharging; , These represent the maximum and minimum stored capacity of the energy storage device, respectively; the constraint that the capacity of the energy storage device is equal at the beginning and end of its short-term operating cycle is shown in the following formula: 。 7. A rationality evaluation system for power grid investment planning schemes considering low-carbon objectives, characterized in that, include: The load assessment module collects relevant data and indicators related to power grid investment planning. Based on the relevant data and indicators, it divides the power grid load forecasting area into regions. Using the seasonal decomposition method of time series, it decomposes the power system load data into three parts: real-time variation, trend stability component, and irregular fluctuation component. The power system load data is preprocessed to obtain a dataset. The preprocessed dataset is divided into a training dataset and a test dataset. Power grid load forecasting models are built for different components. The training dataset is used to train the load forecasting model. The trained load forecasting model is used to predict the samples in the test dataset to obtain the load forecasting results. The load forecasting results are inversely normalized to obtain the final load forecasting results. By comparing the actual load value with the final load forecast result, the error between the two is obtained. The forecast model is evaluated based on the error, and the final load assessment result is obtained. The simulation module uses the k-means algorithm to cluster the daily wind and solar power output and load data throughout the year, and selects typical days from each quarter for short-term operation simulation. The optimization module constructs a planning scheme evaluation model with the goal of minimizing operating costs, simulates the matching relationship between the output and load curves of various power sources, and optimizes the peak-shaving capacity and economy of the power system, as well as the investment and planning schemes of the power system based on the matching relationship. The optimization solution module solves the complex constraints in the planning scheme evaluation model using the marine predator algorithm. If all planning schemes meet the system peak-shaving requirements, the planning results are considered reasonable. The complex constraints in the planning scheme evaluation model are solved using the marine predator algorithm, including the high-speed ratio stage, the medium-speed ratio stage, and the low-speed ratio stage. The specific steps are as follows: The positions of predators and prey are initialized, and an elite matrix is ​​constructed with the predators with the best fitness, and a prey matrix is ​​constructed with the prey with a uniform distribution. The high-speed ratio stage is the initial stage of algorithm optimization iteration, during which global location information is surveyed. The mathematical model for the high-speed ratio stage is as follows: ； ； ； in, This represents the movement step size; It is a random vector based on a Brownian walk normal distribution; This is a term-by-term multiplication operation; It is a constant, equal to 0.5; It is a uniform random vector in [0, 1]; This represents the current iteration number; This represents the maximum number of iterations. Population size; The intermediate speed ratio stage represents the middle phase of the algorithm iteration. The population is divided into two parts: prey performs lévy flight, responsible for spatial development, while predators perform Brownian motion, responsible for spatial reconnaissance. The mathematical model expression for this stage is as follows: ； ； ； ； ； ； ； in, Represents the Lévy motion random vector. This represents an adaptive parameter that controls the predator's step size. The medium-speed ratio stage is the later stage of algorithm iteration, during which the predator adopts a levy flight strategy and focuses more on developing local areas. The mathematical model expression for this stage is as follows: ； ； ； The external environmental factors of fish gathering devices and eddy current effects alter the predator's foraging strategy to escape local optima, allowing the algorithm to avoid premature convergence. Its mathematical model expression is as follows: ； in, To minimize the impact on probability, we set it to 0.2; It is a binary vector; A random number within the range [0, 1]; , These are random indices of the prey matrix; The Ocean Predator algorithm is used to solve the objective function under complex constraints, including coal-fired power output constraints and power balance constraints, to optimize grid operation, including peak-shaving capacity and economic efficiency. After optimization, the difference between the actual peak-shaving capacity and the expected peak load demand is calculated. If the difference is lower than a pre-set threshold, the power planning and operation strategy is effective and can meet the system's peak-shaving demand, ensuring safe and stable grid operation. Otherwise, it is necessary to readjust the optimization model parameters or increase peak-shaving resources and continue to use the Ocean Predator algorithm for optimization.

8. An electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that, When the processor executes the program, it implements a method for evaluating the rationality of a power grid investment planning scheme that takes into account low-carbon objectives, as described in any one of claims 1 to 6.

9. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by the processor, it implements a method for evaluating the rationality of a power grid investment planning scheme that takes into account low-carbon objectives, as described in any one of claims 1 to 6.

Images