Electric power project budget calculation method and system

Abstract

The invention discloses an electric power project budget calculation method and system, and relates to the technical field of data analysis. Acquiring historical data of the target area and preprocessing to obtain target historical data; performing feature extraction on the target historical data to obtain feature data, obtaining spatial data of the target region, and performing construction according to the feature data and the spatial data to obtain a target database; updating the preset model according to the target database to obtain a target model, obtaining an electric power project task, and inputting the electric power project task into the target model to obtain a final project scheme; and if the final project scheme completes the electric power project task, determining an electric power project budget. Obtaining target area load data, performing preprocessing and feature extraction, constructing a target database in combination with the spatial data, and updating and optimizing the preset model; the final project scheme is obtained after the electric power project task is input into the model, the electric power project budget is determined, the data complexity is reduced, and the calculation efficiency and precision are improved.

Description

Technical Field

[0001] This invention belongs to the field of data analysis technology, specifically relating to a method and system for budget calculation of power projects. Background Technology

[0002] With the rapid development of the social economy, power projects, as a key area of ​​infrastructure construction, have undergone a long process of transformation from traditional power grid construction to smart grids. Early power project budget calculations were done manually, but with the increasing number of power projects, manual calculations and existing methods have become insufficient to meet the demands.

[0003] Application No. CN117993854A discloses a data analysis-based method for cost management of power engineering projects in industrial parks; it includes investment budgeting for the main and distribution networks of power engineering projects under construction in industrial parks; investment accounting for power engineering projects under construction in industrial parks, including real-time monitoring of cost and progress information of power engineering projects under construction in industrial parks, judging whether there are abnormal deviation data based on the project data and issuing early warnings for deviation data, performing cost accounting based on the cost information of power engineering projects under construction in industrial parks and generating project reports; and investment evaluation and analysis of power engineering projects under construction in industrial parks.

[0004] In the existing technology, the calculation of power project budgets requires processing multiple data dimensions, resulting in low computational efficiency. Traditional models are unable to handle multiple variables, leading to poor project simulation accuracy and large budget errors. Summary of the Invention

[0005] The purpose of this invention is to solve the problem that the calculation of power project budgets involves many data dimensions, which leads to low computational efficiency. Traditional models are unable to cope with multiple variables, resulting in poor project simulation accuracy and large budget errors. Therefore, this invention proposes a method and system for calculating power project budgets.

[0006] In a first aspect of this invention, a method for calculating the budget of a power project is first proposed, the method comprising: Historical data of the target area is obtained, and the historical data is preprocessed to obtain target historical data; Feature data is obtained by extracting features from the target historical data, spatial data of the target area is obtained, and a target database is constructed based on the feature data and the spatial data. The target model is obtained by updating the preset model according to the target database, and the power project task is obtained. The power project task is then input into the target model to obtain the final project plan. If the final project plan completes the power project task, the power project budget is determined based on the final project plan.

[0007] Optionally, feature data is obtained by extracting features from the target historical data, including: A sample set is determined based on the target historical data, and the weighted distance of each sample in the sample set is calculated; the sample set contains multiple samples; each sample consists of a feature vector and a class label; The target probability is calculated based on the weighted distance of the target sample, and the classification probability is calculated based on the target probability; the target sample is any one of the sample sets. If the target probability is greater than the classification threshold, the target sample is retained; otherwise, the target sample is removed.

[0008] Optionally, the target model is obtained by updating the preset model based on the target database, including: The parameter combination of the preset model is determined, and the population is initialized to obtain the target population; the parameter combination includes: penalty coefficient, function parameters, and parameter search range; Calculate the fitness of each chromosome in the target population, and determine the optimal chromosome based on the value of the fitness. The target population is iteratively updated based on the optimal chromosome to obtain an optimal set of individuals, and a search region set is determined based on the optimal set of individuals; the optimal set of individuals contains the optimal individuals of each chromosome during the iteration process; the search region set contains multiple search regions; and there is a one-to-one correspondence between the optimal individuals and the search regions. The search is iteratively updated based on each search region until the preset conditions are met, at which point the optimal parameter combination is output. The preset model is then updated based on the optimal parameter combination to obtain the target model.

[0009] Optionally, inputting the power project tasks into the target model to obtain the final project solution includes: Based on the power project tasks, a task sequence is determined, and the first task is input into the target model to obtain a first solution; the task sequence includes: a first task, a second task, and a third task; the first task is a global variable optimization, the second task is a discrete variable optimization, and the third task is a continuous variable optimization; Based on the first scheme, a first optimization objective is determined, and the second task and the first optimization objective are input into the target model to obtain a second scheme; Based on the second scheme, a second optimization objective is determined. The third task and the second optimization objective are input into the target model to obtain a third scheme. The third scheme is then used as the final project scheme.

[0010] Optionally, if the final project plan completes the power project task, it further includes: The operating status parameters are obtained by simulation based on the final project plan. If the operating status parameters meet the preset constraints, it is determined that the final project plan has completed the power project task. Otherwise, it is determined that the final project plan has not completed the power project task, and the power project task is input into the target model for training until the preset number of training iterations is reached, at which point training stops.

[0011] In a second aspect of this invention, a power project budget calculation system is provided, comprising: The data preprocessing module is used to acquire historical data of the target area and preprocess the historical data to obtain target historical data. The database construction module is used to extract features from the target historical data to obtain feature data, acquire spatial data of the target area, and construct the target database based on the feature data and the spatial data. The model update module is used to update the preset model according to the target database to obtain the target model, and to obtain the power project task, and input the power project task into the target model to obtain the final project plan. The budget determination module is used to determine the power project budget based on the final project plan if the final project plan completes the power project task.

[0012] Optionally, the database construction module includes: The sample calculation module is used to determine a sample set based on the target historical data and calculate the weighted distance of each sample in the sample set; the sample set contains multiple samples; each sample consists of a feature vector and a class label; The probability calculation module is used to calculate the target probability of a target sample based on a weighted distance, and to calculate the classification probability based on the target probability; the target sample is any one of the samples in the sample set. The probability classification module is used to retain the target sample if the target probability is greater than the classification threshold, and otherwise remove the target sample.

[0013] Optionally, the model update module includes: An initialization module is used to determine the parameter combination of the preset model and initialize the population to obtain the target population; the parameter combination includes: penalty coefficient, function parameters, and parameter search range; The fitness calculation module is used to calculate the fitness of each chromosome in the target population and determine the optimal chromosome based on the value of the fitness. The local optimization module is used to iteratively update the target population based on the optimal chromosome to obtain an optimal set of individuals, and to determine a search region set based on the optimal set of individuals; the optimal set of individuals contains the optimal individuals of each chromosome during the iteration process; the search region set contains multiple search regions; and there is a one-to-one correspondence between the optimal individuals and the search regions. The global optimization module is used to iteratively update each search region until a preset condition is met, then output the optimal parameter combination, and update the preset model according to the optimal parameter combination to obtain the target model.

[0014] Optionally, the model update module further includes: The first execution module is used to determine a task sequence based on the power project task, and input the first task into the target model to obtain a first solution; the task sequence includes: a first task, a second task, and a third task; the first task is a global variable optimization, the second task is a discrete variable optimization, and the third task is a continuous variable optimization; The second execution module is used to determine the first optimization objective based on the first scheme, and input the second task and the first optimization objective into the target model to obtain the second scheme; The third execution module is used to determine the second optimization objective based on the second scheme, input the third task and the second optimization objective into the target model to obtain the third scheme, and use the third scheme as the final project scheme.

[0015] Optionally, the budget determination module further includes: The first judgment module is used to obtain the operating status parameters by simulating the final project plan. If the operating status parameters meet the preset constraints, the final project plan is judged to have completed the power project task. The second judgment module is used to determine otherwise that the final project plan has not completed the power project task, and input the power project task into the target model for training until the preset number of training times is reached and then training stops.

[0016] The beneficial effects of this invention are: This invention proposes a method for calculating the budget of power projects. It involves acquiring and preprocessing historical data of a target area to obtain target historical data; extracting features from the target historical data to obtain feature data; acquiring spatial data of the target area; constructing a target database based on the feature data and spatial data; updating a preset model based on the target database to obtain a target model; acquiring power project tasks; and inputting the power project tasks into the target model to obtain the final project plan. If the final project plan completes the power project tasks, the power project budget is determined. This method acquires and preprocesses load data of the target area, performs feature extraction, constructs a target database using spatial data, updates and optimizes the preset model, and inputs the power project tasks into the model to obtain the final project plan and determine the power project budget. This approach reduces data complexity and improves computational efficiency and accuracy. Attached Figure Description

[0017] The invention will now be further described with reference to the accompanying drawings.

[0018] Figure 1 A flowchart of a method for budget calculation of power projects provided in an embodiment of the present invention; Figure 2 A framework diagram of another power project budget calculation system provided in an embodiment of the present invention. Detailed Implementation

[0019] 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. The term "and / or" in this document is merely a description of the relationship between related objects, indicating that three relationships can exist. For example, A and B can represent: A alone, A and B simultaneously, and B alone. Furthermore, descriptions involving "first," "second," etc., in this invention are for descriptive purposes only and should not be construed as indicating or implying their relative importance or implicitly specifying the number of technical features indicated. Therefore, features defined with "first" or "second" can explicitly or implicitly include at least one of those features. Additionally, the technical solutions of the various embodiments can be combined with each other, but this must be based on the ability of those skilled in the art to implement them. When the combination of technical solutions is contradictory or cannot be implemented, it should be considered that such a combination of technical solutions does not exist and is not within the scope of protection claimed by the present invention.

[0020] 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.

[0021] This invention provides a method for budget calculation of power projects. See also... Figure 1 , Figure 1 A flowchart illustrating a method for budget calculation of a power project, provided as an embodiment of the present invention. The method includes the following steps: S101, Obtain historical data of the target area, and preprocess the historical data to obtain the target historical data; S102, extract features from historical data of the target to obtain feature data, acquire spatial data of the target area, and construct a target database based on the feature data and spatial data; S103, Update the preset model according to the target database to obtain the target model, and obtain the power project task. Input the power project task into the target model to obtain the final project plan. S104. If the final project plan completes the power project task, the power project budget shall be determined based on the final project plan.

[0022] The present invention provides a method for calculating the budget of a power project. By acquiring load data of the target area and performing preprocessing and feature extraction, and combining spatial data to construct a target database, the preset model is updated and optimized. After inputting the power project task into the model, the final project plan is obtained, and the power project budget is determined, thereby reducing data complexity and improving calculation efficiency and accuracy.

[0023] In one implementation, the target area can be an area where power generation equipment such as thermal power, hydropower, and wind and solar power are built; historical data includes historical load data and weather characteristics of the target area, such as season, temperature, and date characteristics, such as weekdays / rest days; preprocessing operations include: data cleaning: handling missing values, such as deletion and filling, and outliers, such as identification, correction, or removal, to ensure data integrity and accuracy; data transformation: including standardization / normalization to make different dimensional characteristics comparable, discretization, segmentation of continuous values, and function transformation, such as logarithmic transformation.

[0024] In one implementation, core features such as load fluctuation, temperature factor, and geographical distance are screened through neighborhood component analysis. This can reduce data dimensionality and remove redundant information, which not only reduces computational complexity but also improves the model's ability to focus on key factors, thereby improving the model's accuracy and operating efficiency.

[0025] In one implementation, a database linking attribute data and spatial data is established through a geographic information system (spatial data). Feature data is associated with spatial data of power facilities, such as location coordinates, user distribution areas, and power supply node layout. This enables the database to support optimized analysis of power facility layout, achieve reasonable layout planning for power projects, avoid resource waste and operational risks caused by unreasonable layout, and improve the overall operating efficiency and reliability of the power system.

[0026] In one implementation, a support vector machine model (pre-defined model) is constructed, taking features from the database as input and outputting classification or regression results with the objectives of "optimal load rate" and "minimum running error". The key parameters of the pre-defined model are optimized using a particle swarm optimization algorithm, such as the penalty coefficient C and the kernel function parameter γ. The prediction error is minimized through swarm intelligence search, thereby improving the model accuracy.

[0027] In one implementation, once the final project plan is determined, the power project budget, including the cost of the project, can be calculated based on that plan.

[0028] In one embodiment, feature data is obtained by extracting features from target historical data, including: The sample set is determined based on the target's historical data, and the weighted distance between each sample in the sample set is calculated; the sample set contains multiple samples; each sample consists of a feature vector and a class label; The target probability is calculated based on the weighted distance, and the classification probability is calculated based on the target probability; the target sample is any one of the samples in the sample set. If the target probability is greater than the classification threshold, the target sample is retained; otherwise, the target sample is removed.

[0029] In one implementation, the sample set S = {x} i ,y i}(i=1,2,…,N), where xi is a feature vector containing multi-dimensional features such as load data, weather quantification values, and date type; y i The category label for the sample identifies the category to which the sample belongs, such as "high load", "low load", "normal operation", "abnormal operation", etc.; the importance of different features is quantified by the feature weight vector w, and the value of any two samples x is calculated. i With x j The weighted distance is calculated using the following formula: , where D w (x i ,x j ) is the sample x i With x j The weighted distance, w k Let x be the weight of the k-th feature, K be the total number of features, and x be the weight of the k-th feature. ik For sample x i The kth eigenvalue, x jk For sample x j The k-th feature value; if the weight w of a certain feature k A larger value indicates that this feature has a higher weight in the distance calculation, meaning that the difference in this feature has a more significant impact on the "distance" between samples; if w kIf the value is close to 0, the difference in this feature has a negligible impact on the distance and can be considered a redundant feature.

[0030] In one implementation, the target probability is calculated based on the weighted distance of the target sample, and the probability of any sample x is calculated. j Selected as target sample x i The probability of the reference point (the probability of the target point); , where P ij For x j As x i The probability of the reference point, (k(z)=exp(-z / τ) is the kernel function (weighted distance D) w (decreasing function), k(D) w (x i ,x j )) is x i and x j The kernel function value, For all other samples with x i The sum of kernel function values ​​is used to measure the correlation between samples through probability distribution, thereby improving classification accuracy; when two samples x i and x j Weighted distance D w The smaller the value, the more similar the core features, k(D) w The larger the value of ), the better. ij The higher the probability (similar samples are more likely to serve as reference points for each other); when D w The larger the value, the greater the difference in core features, k(D) w The smaller P is, ij The lower the value, the lower the probability that an outlier sample will be selected as a reference point.

[0031] In one implementation, the classification probability is calculated based on the target probability, which is equivalent to calculating the probability of sample x. i The probability of being correctly classified (classification probability). , where when x i With x j When they belong to the same category, y ij =1, otherwise y ij =0; p i The larger the value, the stronger x i The stronger the correlation between the features and similar samples, the more reliable the classification; by optimizing the feature vector and retaining the weights w k Higher characteristics, i.e., for p i For features with a significant impact, redundant features with weights close to 0 are removed to achieve feature dimensionality reduction.

[0032] In one embodiment, updating a preset model based on a target database to obtain a target model includes: Determine the parameter combination of the preset model and initialize the population to obtain the target population; the parameter combination includes: penalty coefficient, function parameters, and parameter search range; Calculate the fitness of each chromosome in the target population, and determine the optimal chromosome based on the value of the fitness. The target population is iteratively updated based on the optimal chromosome to obtain the optimal set of individuals. The search region set is then determined based on the optimal set of individuals. The optimal set of individuals contains the optimal individuals of each chromosome during the iteration process. The search region set contains multiple search regions. There is a one-to-one correspondence between the optimal individuals and the search regions. The system iterates and updates each search region until the preset conditions are met, at which point it outputs the optimal parameter combination. The preset model is then updated based on the optimal parameter combination to obtain the target model.

[0033] In one implementation, the preset model can be a support vector machine; the parameter combination is as follows: 1. Penalty coefficient C: controls the overfitting degree of the SVM model. The larger the value, the heavier the penalty for misclassification; 2. Kernel function parameter γ: controls the nonlinearity of the model and affects the mapping complexity of samples in high-dimensional space; 3. Set the search range of the parameters, for example: C∈[0.01,100], γ∈[0.001,10], as the search space boundary of the particle swarm.

[0034] In one implementation, the optimal set of individuals contains the best individuals from each chromosome during the iteration process. The target population contains multiple chromosomes, and each iteration records the historical optimal position of each particle, i.e., the parameter combination with the highest fitness so far for that particle. The optimal chromosome is the position with the highest fitness selected from all the optimal individual sets of particles, which is taken as the current optimal solution for the entire particle swarm. The fitness formula is based on existing technology, for example: ; Where w is the weight vector of the target model, ||w|| 2 Let ε be the squared magnitude of the weight vector, C be the penalty coefficient, and ε be the weight vector. i is a slack variable, and n is the number of training samples.

[0035] In one implementation, during each iteration, the chromosome with the highest fitness is selected as the optimal chromosome, and each particle randomly explores the search space in the early stages of iteration, considering possible combinations of model parameters, and evaluates the quality of solutions using a fitness function. The individual optimal solution records the "best solution found so far" by the particle. Even if subsequent particles deviate from this region due to global guidance, they can still trace back to historical high-quality solutions through the individual optimal solution, preventing the loss of valuable information. The optimal chromosome guides particles to concentrate in the group's recognized optimal region (global exploration), while the individual optimal solution prompts particles to search deeper near their own historical high-quality regions (local development).

[0036] In one implementation, the following conditions are set: the number of iterations reaches a preset value (e.g., 100 times), and the fitness of the optimal chromosome does not improve significantly for multiple rounds (e.g., 10 rounds); until the preset conditions are met, the optimal parameter combination is output, that is, the parameter combination corresponding to the optimal chromosome at this time is used as the optimal parameter combination of the preset model for updating.

[0037] In one embodiment, inputting power project tasks into a target model to obtain a final project solution includes: The task sequence is determined based on the power project tasks. The first task is input into the target model to obtain the first solution. The task sequence includes: the first task, the second task, and the third task. The first task is for global variable optimization, the second task is for discrete variable optimization, and the third task is for continuous variable optimization. Based on the first scheme, the first optimization objective is determined, and the second task and the first optimization objective are input into the target model to obtain the second scheme; Based on the second scheme, the second optimization objective is determined. The third task and the second optimization objective are input into the target model to obtain the third scheme, which is then used as the final project scheme.

[0038] In one implementation, the task sequence is determined based on the power project tasks. The first task is pre-optimization, the second task is coarse optimization, and the third task is fine optimization. Pre-optimization is as follows: In the pre-optimization stage, the variable type (discrete or continuous) is not distinguished. All core variables affecting the relocation plan (circuit breakers, on-load tap changers, distributed power sources) are globally optimized. The purpose is to quickly determine the approximate scheduling range of each device (such as how much reactive power the distributed power source needs to output and how many circuit breakers need to be put into operation), but the operation limit of discrete devices (such as circuit breakers) is not considered.

[0039] In one implementation, coarse optimization focuses on discrete variables. By fixing the pre-optimized reactive power output of distributed power sources, the scheduling plans of circuit breakers and on-load tap changers are individually modified. For example, the pre-optimization may require the circuit breaker to switch 5 times in 1 hour, but in reality only 3 times are allowed. Coarse optimization will adjust the switching time or number of times to ensure that the operational constraints are met, while minimizing the deviation from the pre-optimization result.

[0040] In one implementation, fine optimization focuses on continuous variables. The coarse-optimized scheduling plans for circuit breakers and on-load tap changers are fixed, and only the reactive power output of distributed generation (a continuous control variable) is adjusted. The goal is to minimize system network losses and compensate for deviations caused by the adjustment of discrete variables, such as voltage offsets due to the limitation on the number of on-load tap changer operations. The final reactive power output plan of distributed generation is then output and combined with the coarse-optimized discrete variable plan to obtain a complete scheduling scheme.

[0041] In one implementation, pre-optimization involves retaining only basic operational constraints of the equipment (such as the reactive power output of the distributed power source not exceeding the rated value and the voltage not exceeding the limit) to quickly explore the global optimal space; coarse optimization involves adding constraints on the number of operations of discrete equipment to solve the problem that pre-optimization is theoretically feasible but practically inoperable; fine optimization involves constraints on the reactive power output boundary of the distributed power source and the node voltage to ensure the accuracy of continuous variable adjustment.

[0042] In one embodiment, if the final project plan completes the power project task, it further includes: The operating status parameters are obtained by simulation based on the final project plan. If the operating status parameters meet the preset constraints, the final project plan is determined to have completed the power project task. Otherwise, the final project plan is deemed to have failed to complete the power project task, and the power project task is input into the target model for training until the preset number of training iterations is reached, at which point training stops.

[0043] In one implementation, the operating status parameters include: voltage and power: if all node voltages are within range and line power is not overloaded, the system stability meets expectations; if there is voltage deviation or overload, distributed generation reactive power compensation needs to be added during the fine optimization stage; equipment operation: if the number of operations of circuit breakers and on-load tap changers does not exceed the preset upper limit (e.g., on-load tap changers switch ≤ 3 times per day), the equipment loss constraint is met; if the limit is exceeded, the operation number limit needs to be tightened during the coarse optimization stage; load rate: if the load rate curve output by simulation shows that the load rate is between 70% and 90% for more than 90% of the time, it meets expectations; if it exceeds the range (e.g., continuously > 90% or < 70%), the load distribution strategy needs to be optimized.

[0044] In one implementation, if the constraints are met in the first simulation, the effectiveness of the solution is directly confirmed, avoiding potential risks in actual execution. If the constraints are not met, the target model can be continuously optimized for adaptability to the task scenario through targeted training on task data. By limiting the number of training sessions, the optimization effect of the model can be guaranteed while preventing resource waste caused by overtraining, thereby improving the flexibility of training and the quality of the solution.

[0045] Based on the same inventive concept, this invention also provides a power project budget calculation system. See [link to relevant documentation]. Figure 2 , Figure 2 A framework diagram of a power project budget calculation system provided in this embodiment of the invention includes: The data preprocessing module is used to acquire historical data of the target area and preprocess the historical data to obtain the target historical data. The database construction module is used to extract features from historical target data to obtain feature data, acquire spatial data of the target area, and construct the target database based on the feature data and spatial data. The model update module is used to update the preset model according to the target database to obtain the target model, and to obtain the power project task, input the power project task into the target model to obtain the final project plan; The budget determination module is used to determine the power project budget based on the final project plan if the final project plan completes the power project tasks.

[0046] The power project budget calculation system provided by this invention acquires target area load data and performs preprocessing and feature extraction, constructs a target database by combining spatial data, and updates and optimizes the preset model; after inputting the power project task into the model, the final project plan is obtained and the power project budget is determined, thereby reducing data complexity and improving calculation efficiency and accuracy.

[0047] In one embodiment, the database building module includes: The sample calculation module is used to determine the sample set based on the target's historical data and calculate the weighted distance between each sample in the sample set; the sample set contains multiple samples; each sample consists of a feature vector and a class label; The probability calculation module is used to calculate the target probability of a target sample based on the weighted distance, and then calculate the classification probability based on the target probability; the target sample can be any one of the samples in the sample set. The probability classification module is used to retain target samples if the target probability is greater than the classification threshold, and otherwise remove target samples.

[0048] In one embodiment, the model update module includes: The initialization module is used to determine the parameter combination of the preset model and initialize the population to obtain the target population; the parameter combination includes: penalty coefficient, function parameters, and parameter search range; The fitness calculation module is used to calculate the fitness of each chromosome in the target population and determine the optimal chromosome based on the value of the fitness. The local optimization module is used to iteratively update the target population based on the optimal chromosome to obtain the optimal set of individuals, and to determine the search region set based on the optimal set of individuals; the optimal set of individuals contains the optimal individuals of each chromosome during the iteration process; the search region set contains multiple search regions; and there is a one-to-one correspondence between the optimal individuals and the search regions. The global optimization module is used to iteratively update the model based on each search region until the preset conditions are met, at which point the optimal parameter combination is output. The preset model is then updated based on the optimal parameter combination to obtain the target model.

[0049] In one embodiment, the model update module further includes: The first execution module is used to determine the task sequence based on the power project tasks, and input the first task into the target model to obtain the first solution; the task sequence includes: the first task, the second task, and the third task; the first task is global variable optimization, the second task is discrete variable optimization, and the third task is continuous variable optimization; The second execution module is used to determine the first optimization objective based on the first scheme, and input the second task and the first optimization objective into the target model to obtain the second scheme; The third execution module is used to determine the second optimization objective based on the second scheme, input the third task and the second optimization objective into the target model to obtain the third scheme, and use the third scheme as the final project scheme.

[0050] In one embodiment, the budget determination module further includes: The first judgment module is used to obtain the operating status parameters by simulating the final project plan. If the operating status parameters meet the preset constraints, the final project plan is judged to have completed the power project task. The second judgment module is used to determine if the final project plan has not completed the power project task, and input the power project task into the target model for training until the preset number of training times is reached, at which point training stops.

[0051] The foregoing has described one embodiment of the present invention in detail, but this is merely a preferred embodiment and should not be considered as limiting the scope of the invention. All equivalent variations and modifications made within the scope of the claims of this invention should still fall within the scope of the claims.

Claims

1. A method for budgeting power projects, characterized in that, The method includes: Historical data of the target area is obtained, and the historical data is preprocessed to obtain target historical data; Feature data is obtained by extracting features from the target historical data, spatial data of the target area is obtained, and a target database is constructed based on the feature data and the spatial data. The target model is obtained by updating the preset model according to the target database, and the power project task is obtained. The power project task is then input into the target model to obtain the final project plan. If the final project plan completes the power project task, the power project budget is determined based on the final project plan.

2. The method for budget calculation of power projects according to claim 1, characterized in that, Feature data is obtained by extracting features from the target historical data, including: A sample set is determined based on the target historical data, and the weighted distance of each sample in the sample set is calculated; the sample set contains multiple samples; each sample consists of a feature vector and a class label; The target probability is calculated based on the weighted distance of the target sample, and the classification probability is calculated based on the target probability; the target sample is any one of the sample sets. If the target probability is greater than the classification threshold, the target sample is retained; otherwise, the target sample is removed.

3. The method for budget calculation of power projects according to claim 1, characterized in that, The target model is obtained by updating the preset model based on the target database, including: The parameter combination of the preset model is determined, and the population is initialized to obtain the target population; the parameter combination includes: penalty coefficient, function parameters, and parameter search range; Calculate the fitness of each chromosome in the target population, and determine the optimal chromosome based on the value of the fitness. The target population is iteratively updated based on the optimal chromosome to obtain an optimal set of individuals, and a search region set is determined based on the optimal set of individuals; the optimal set of individuals contains the optimal individuals of each chromosome during the iteration process; the search region set contains multiple search regions; and there is a one-to-one correspondence between the optimal individuals and the search regions. The search is iteratively updated based on each search region until the preset conditions are met, at which point the optimal parameter combination is output. The preset model is then updated based on the optimal parameter combination to obtain the target model.

4. The method for budget calculation of power projects according to claim 1, characterized in that, The final project solution is obtained by inputting the power project tasks into the target model, including: Based on the power project tasks, a task sequence is determined, and the first task is input into the target model to obtain a first solution; the task sequence includes: a first task, a second task, and a third task; the first task is a global variable optimization, the second task is a discrete variable optimization, and the third task is a continuous variable optimization; Based on the first scheme, a first optimization objective is determined, and the second task and the first optimization objective are input into the target model to obtain a second scheme; Based on the second scheme, a second optimization objective is determined. The third task and the second optimization objective are input into the target model to obtain a third scheme. The third scheme is then used as the final project scheme.

5. The method for budget calculation of power projects according to claim 1, characterized in that, If the final project plan completes the power project task, it also includes: The operating status parameters are obtained by simulation based on the final project plan. If the operating status parameters meet the preset constraints, it is determined that the final project plan has completed the power project task. Otherwise, it is determined that the final project plan has not completed the power project task, and the power project task is input into the target model for training until the preset number of training iterations is reached, at which point training stops.

6. A power project budget calculation system, characterized in that, The system includes: The data preprocessing module is used to acquire historical data of the target area and preprocess the historical data to obtain target historical data. The database construction module is used to extract features from the target historical data to obtain feature data, acquire spatial data of the target area, and construct the target database based on the feature data and the spatial data. The model update module is used to update the preset model according to the target database to obtain the target model, and to obtain the power project task, and input the power project task into the target model to obtain the final project plan. The budget determination module is used to determine the power project budget based on the final project plan if the final project plan completes the power project task.

7. A power project budget calculation system according to claim 6, characterized in that, The database construction module includes: The sample calculation module is used to determine a sample set based on the target historical data and calculate the weighted distance of each sample in the sample set; the sample set contains multiple samples; each sample consists of a feature vector and a class label; The probability calculation module is used to calculate the target probability of a target sample based on a weighted distance, and to calculate the classification probability based on the target probability; the target sample is any one of the samples in the sample set. The probability classification module is used to retain the target sample if the target probability is greater than the classification threshold, and otherwise remove the target sample.

8. A power project budget calculation system according to claim 6, characterized in that, The model update module includes: An initialization module is used to determine the parameter combination of the preset model and initialize the population to obtain the target population; the parameter combination includes: penalty coefficient, function parameters, and parameter search range; The fitness calculation module is used to calculate the fitness of each chromosome in the target population and determine the optimal chromosome based on the value of the fitness. The local optimization module is used to iteratively update the target population based on the optimal chromosome to obtain an optimal set of individuals, and to determine a search region set based on the optimal set of individuals; the optimal set of individuals contains the optimal individuals of each chromosome during the iteration process; the search region set contains multiple search regions; and there is a one-to-one correspondence between the optimal individuals and the search regions. The global optimization module is used to iteratively update each search region until a preset condition is met, then output the optimal parameter combination, and update the preset model according to the optimal parameter combination to obtain the target model.

9. A power project budget calculation system according to claim 6, characterized in that, The model update module also includes: The first execution module is used to determine a task sequence based on the power project task, and input the first task into the target model to obtain a first solution; the task sequence includes: a first task, a second task, and a third task; the first task is a global variable optimization, the second task is a discrete variable optimization, and the third task is a continuous variable optimization; The second execution module is used to determine the first optimization objective based on the first scheme, and input the second task and the first optimization objective into the target model to obtain the second scheme; The third execution module is used to determine the second optimization objective based on the second scheme, input the third task and the second optimization objective into the target model to obtain the third scheme, and use the third scheme as the final project scheme.

10. A power project budget calculation system according to claim 6, characterized in that, The budget determination module also includes: The first judgment module is used to obtain the operating status parameters by simulating the final project plan. If the operating status parameters meet the preset constraints, the final project plan is judged to have completed the power project task. The second judgment module is used to determine otherwise that the final project plan has not completed the power project task, and input the power project task into the target model for training until the preset number of training times is reached and then training stops.

Images