A Design and Planning Method for Economic and Energy Indicators Oriented to Source-Grid-Load-Storage

Abstract

The present invention belongs to the technical field of power project construction investment, and specifically relates to a method for designing and planning economic and energy indicators for source-grid-load-storage integration, including the following steps: obtaining the prediction result of the annual energy consumption demand based on the monthly electric energy demand prediction model of the region where the source-grid-load-storage integration project is located, which is based on a full-dimensional dynamic convolution - bidirectional time series convolution network and decomposition error correction, and giving the annual available electric energy capacity model and the correlation model of the annual electric energy production - sales volume; constructing economic benefit and energy efficiency indicators for the whole process input and output of the source-grid-load-storage integration project; giving the evaluation method for the economic benefit and energy efficiency of the source-grid-load-storage integration project in different states; and optimizing the maximization of economic benefit and energy efficiency by using an improved raccoon optimization algorithm and making it meet the supply and demand constraint verification. The present invention can be used for evaluating the investment value of the source-grid-load-storage integration project and also provides support for the optimized utilization of energy after completion.

Description

Technical Field

[0001] The present invention belongs to the technical field of power project construction investment, and particularly relates to a method for designing and planning economic and energy indicators for the source-network-load-storage integration. Background Art

[0002] Efforts are being made to build a clean, low-carbon, safe and efficient energy system, improve the utilization rate of clean energy and the operation efficiency of the power system, and better play the role of the integration of the source-network-load-storage and multi-energy complementarity in ensuring energy security. The construction of systems such as power auxiliary services, medium- and long-term, and spot markets in multiple provincial-level regions in China has introduced market players such as the power generation side, the load side, and electric energy storage, fully liberalized market transactions, promoted the establishment of a market trading mechanism for marketized trading users to participate in and undertake auxiliary services, cultivated the user load management ability, and improved the enthusiasm of the user side for peak shaving.

[0003] However, the source-network-load-storage integration project involves various types of energy such as hydropower, wind power, photovoltaic power, and thermal power, and also involves multiple links such as transmission and distribution construction, load, and storage, and is related to the external sales and external purchases of electric energy, as well as the complementarity and scheduling of different types of energy, and even involves cross-regional power dispatching. On the one hand, it is necessary to achieve the balance between energy production and use, and on the other hand, it involves the full utilization of clean and low-carbon energy. Therefore, the source-network-load-storage integration project is associated with a variety of information. How to scientifically and effectively design and plan the source-network-load-storage integration project is extremely important for the construction party or investor.

[0004] For the design and planning of the above-mentioned source-network-load-storage integration project, on the one hand, from the perspective of the construction party, it is necessary to evaluate the total demand of the load. At present, the measurement methods of load demand have the characteristics of being single and having low practicality for future scenarios, and a relatively accurate prediction method needs to be formed; on the other hand, from the perspective of the investor, it is difficult to weigh the comprehensive value of the construction of the source-network-load-storage integration project subjectively, resulting in low credibility of the value evaluation of the source-network-load-storage integration project, and thus delaying the construction progress of the source-network-load-storage integration project. In addition, for the evaluation results of the source-network-load-storage integration, there is also a lack of an effective means to plan and design various indicators, which is not conducive to the guiding role of the evaluation method for the implementation results.

[0005] In summary, there is an urgent need for a load demand prediction method, an investment value evaluation model and method, and an optimization method applicable to the source-network-load-storage integration project to assist in the multi-angle scientific and effective design and planning of the source-network-load-storage integration project and optimize the corresponding indicators. Summary of the Invention

[0006] Aiming at the technical problem that the current measurement methods of load demand are single and have low practicality for future scenarios, the present invention provides a method for designing and planning economic and energy indicators for the source-network-load-storage integration.

[0007] To solve the above technical problems, the technical solution adopted by the present invention is as follows:

[0008] An economic and energy index design and planning method for source-grid-load-storage integration, comprising the following steps:

[0009] S1. Construct a monthly electricity demand prediction model for the region where the source-grid-load-storage integration project is located based on a full-dimensional dynamic convolution-bi-directional time series convolution network and decomposition error correction. Combine the natural resource conditions of the project location to construct an annual available capacity model for power station power generation. Finally, establish an association model between annual energy consumption demand and annual electricity production-sales volume;

[0010] S2. Uniformly describe the construction cost, later operation and maintenance cost, electricity production, line loss, terminal load, and energy storage link in the construction of the source-grid-load-storage integration project using economic quantities;

[0011] S3. Combine the comprehensive value concerned by the investment in the source-grid-load-storage integration project to construct calculation models for economic benefits and energy efficiency used in the evaluation of the source-grid-load-storage integration project;

[0012] S4. Based on the potential supply-demand relationship of the source-grid-load-storage integration project, calculate various economic and energy efficiency indicators under different supply-demand states, and conduct value evaluations of the source-grid-load-storage integration project for different supply-demand situations respectively;

[0013] S5. Combine the calculation models for economic benefits and energy efficiency corrected by the value evaluation model under different supply-demand situations, and use an improved raccoon optimization algorithm to optimize the maximization of economic benefits and energy efficiency and make it meet the supply-demand constraint verification.

[0014] The method for establishing the association model between annual energy consumption demand and annual electricity production-sales volume in S1 is as follows:

[0015] S11. Collect historical electricity demand data with reduced noise. The specific method is to obtain monthly historical electricity demand data H(t) and I(t); use the singular spectrum analysis algorithm to obtain the first intermediate monthly historical electricity demand sequence F without interference, that is, denoised;

[0016] S12. Construct a matrix for the collected data set. The method is to perform wavelet analysis on the first intermediate monthly electricity demand sequence F to obtain the second intermediate monthly electricity demand sequence F n There are a total of N, where N represents the level of wavelet decomposition; based on the N second intermediate monthly electricity demand sequences F n , construct the first electricity demand matrix f, and perform normalization processing to obtain the second monthly electricity demand matrix f1;

[0017] S13. Divide the second monthly electricity demand matrix f1 based on the time series step length, and obtain the matrix slice of the step length [t - T, t - 1]. Based on the dynamic convolutional neural network, extract features from the input matrix to obtain the comprehensive feature f2, where T > 1;

[0018] S14. Input the comprehensive feature f2 of the existing matrix into the bidirectional time series convolutional network time prediction model to obtain the monthly electricity demand prediction result D at the current time step t; t ; Input the comprehensive feature into the BiTCN convolutional time prediction model to obtain the monthly electricity demand prediction result at time step t; i of the monthly electricity demand;

[0019] d(t) = D t

[0020] where d(t) represents the comprehensive prediction result of the monthly electricity demand at the current time step t, and D t represents the prediction result of the monthly electricity demand at the current time step t;

[0021] S15. Input the monthly electricity demand prediction result D at the current time step t t into the variational mode decomposition model to obtain the first monthly electricity demand prediction component G1, the second monthly electricity demand prediction component G2, and the third monthly electricity demand prediction component G3 at the current time step t; Use the trained error correction model to correct the results of the first monthly electricity demand prediction component, the second monthly electricity demand prediction component, and the third monthly electricity demand prediction component, and then splice them to obtain the final monthly electricity demand prediction result g(t);

[0022] S16. Based on the monthly electricity demand prediction result g(t), evaluate the total annual electricity demand of industries, civil use, and commerce in this region

[0023] S17. Combining the natural conditions and electricity demand of the area planned for the source-network-load-storage integration project, the construction of the source-network-load-storage integration project needs to evaluate the maximum capacity that can be built for various power stations on the source side, and construct the power station construction capacity matrix C:

[0024]

[0025] represent the maximum capacities that can be built for hydropower, wind power, photovoltaic power, thermal power, etc. on the source side in turn, and also represent the maximum capacity that can be provided by the electricity in this region;

[0026] Assume that the available capacity expression of various energy sources in the nth year is:

[0027]

[0028] where ηj j = 1, 2, 3, 4 represent the output capacities of different power generation stations. For thermal power, η j can take the value of 1. For wind power and photovoltaic power, they can be valued according to the local historical meteorological laws. σ j is the annual average attenuation coefficient. Currently, the annual attenuation coefficient of photovoltaic power is roughly between 0.4% and 2%. Thus, the annual (n) available capacity matrix of the power stations on the source side can be determined and described as follows:

[0029] C(n) = [c1(n) c2(n) c3(n) c4(n)]

[0030] c j (n)j = 1, 2, 3, 4 represent the available capacities of hydropower, wind power, photovoltaic power, thermal power, etc. on the source side in sequence. Let K represent the number of hours in a year, K = 365 * 24 = 8760. Then, the maximum annual (n) power supply volume on the source side is

[0031] S18. Construct an association model between the annual energy consumption demand and the annual power production - sales volume (KW.h) matrix B, which is specifically described as follows:

[0032]

[0033] Among them, W(t) is the total annual power demand of industry, civil use, and commerce in this region in S16. b 1j j = 1, 2, 3, 4 represent the power production in this region. b 2j represents the loss in power transmission and distribution. b 3j represents the consumption at the load end. b 4j represents the energy storage capacity, mainly for the energy storage description of photovoltaic power and wind power. If hydropower and thermal power generation stations do not need to configure energy storage devices, this part of the data is zero. For other types of energy that require energy storage, the corresponding elements can be valued according to actual needs. represents the proportion of different energies. In an ideal situation ξ j represents the losses generated by different types of power transmission and distribution and communication networks, etc. The annual power production - sales volume (KW.h) matrix B can be described as:

[0034]

[0035] The method for obtaining the denoised monthly historical power demand sequence in S11 is as follows:

[0036] Obtain the specific data sequence H(t) and the change trend sequence I(t) of the power demand for each month in several years in the area where the target source - grid - load - storage integration project acts;

[0037] Input H(t) and I(t) into singular spectrum analysis including embedding, decomposition, grouping, and reconstruction to extract the seasonal trend, long-term trend, and noise in the time series, realizing the analysis or denoising function of the above non-linear time series, and obtaining the first intermediate monthly historical electricity demand series F after denoising.

[0038] The method of constructing the first monthly electricity demand matrix according to N second intermediate monthly electricity demand series in S12 is as follows:

[0039] Let F n , n ∈ {1, 2, … N} represent the nth second intermediate monthly electricity demand series, then the first monthly electricity demand matrix is expressed as:

[0040]

[0041] where, T represents transpose, and f represents the first electricity demand matrix; after normalization, the second monthly electricity demand matrix f1 is obtained.

[0042] The method of the feature extraction training process in S13 is as follows:

[0043] Divide the second monthly electricity demand matrix by combining the sliding window and slicing to obtain matrix slices with a time step of [t i ―T1, t i ―1], t i ∈ [t ― T + T1, t ― 1];

[0044] Take the matrix slice with a time step of [t i ―T1, t i ―1] as the training sample, and take the real monthly electricity demand data at time step t i as the label of the training sample to generate a training sample set;

[0045] For each training sample, concatenate the matrix slice with a time step of [t i ―T1, t i ―1] with other features at time step t i to obtain the comprehensive feature f2.

[0046] The method of using the trained error correction model to correct the initial monthly electricity demand comprehensive prediction result in S15 is as follows:

[0047] Calculate the deviation between the initial monthly electricity demand prediction result at time step t k ∈ [t ― T2, t ― 1] and the real monthly electricity demand at time step t k to construct the error sequence

[0048] Let the error sequence Pt Input the deviation between the comprehensive prediction result of the initial monthly electricity demand at the current time step t by the LSTM network model and the true monthly electricity demand at the current time step t.

[0049] According to the comprehensive prediction result of the initial monthly electricity demand at the current time step t, it is decomposed by variational mode decomposition into the first monthly electricity demand prediction component G1, the second monthly electricity demand prediction component G2, the third monthly electricity demand prediction component G3, and the deviation between the three initial monthly electricity demand prediction components at the current time step t and the true monthly electricity demand at the current time step t is added to obtain the final monthly electricity demand prediction component at the current time step t. The above final monthly electricity demand prediction components are combined to obtain the final monthly electricity demand prediction result g(t).

[0050] The method for uniformly describing the static construction cost, operation and maintenance cost, and economic means of electricity production, sales and use involved in the source-grid-load-storage integrated project in S2 is as follows:

[0051] S21. For the capital investments in multiple links involved in the construction of the source-grid-load-storage integrated project, specifically including the capital investments in the construction of hydropower, wind power, photovoltaic power, and thermal power on the source side, the capital investments in the construction of power transmission and distribution on the grid side, the capital investments in the supporting equipment and lines for industrial electricity, commercial electricity, and residential electricity on the load side, and the capital investments in batteries and related equipment on the energy storage side, the construction investment cost of the source-grid-load-storage can be described by the matrix A as follows:

[0052]

[0053] where, a j1 、a j2 、a j3 、a j4 j = 1, 2, 3, 4 respectively represent the whole-process construction investment costs of different types of energy construction, that is, the construction investments on the source side, grid side, load side, and energy storage side. Specifically, the first column represents the hydropower investment cost, the second column represents the wind power investment cost, the third column represents the photovoltaic power investment cost, and the fourth column represents the thermal power investment cost;

[0054] S22. According to the uncertain risk investment costs of the equipment and lines for the electricity production, transmission and distribution, and user side of hydropower, wind power, photovoltaic power, and thermal power, establish the operation and maintenance cost matrix D:

[0055]

[0056] In the operation and maintenance cost matrix D, each column represents the operation and maintenance costs on the source side, grid side, load side, and energy storage side respectively. Among them, the first column represents the hydropower operation and maintenance cost, the second column represents the wind power operation and maintenance cost, the third column represents the photovoltaic power operation and maintenance cost, and the fourth column represents the thermal power operation and maintenance cost;

[0057] S23. To facilitate the unified evaluation of the economic value of the investment in the source-grid-load-storage integration project, the electricity production and sales are uniformly converted into the corresponding economic costs, described by the electricity price corresponding to KW.h. The prices of different types of energy may vary in each link of production and sales. A transaction price matrix P is constructed to price the corresponding prices of each part separately. The price matrix P is described as follows:

[0058]

[0059] The first column in matrix B represents the prices of each link of hydropower, the second column represents the prices of each link of wind power, the third column represents the prices of each link of photovoltaic power, and the fourth column represents the prices of each link of thermal power. In the source-grid-load-storage integration project, if the energy transaction price does not involve the power generation side, then p 1j For j = 1, 2, 3, 4, it can be taken as 0;

[0060] S24. Based on the calculation method of the Hadamard product, the electricity production, loss, load consumption, and energy storage are converted into the corresponding economic benefit matrix M, which can be described as:

[0061]

[0062] Among them, ○ represents the Hadamard product calculation, which means that the matrix with the same number of rows and columns multiplies the elements with the same subscripts. m ij = b ij * p ij For i = 1, 2, 3, 4 and j = 1, 2, 3, 4, m 1j represents the income obtained from selling electricity by hydropower, wind power, photovoltaic power, and thermal power on the source side. m 2j represents the economic loss caused by the power loss in the power transmission and distribution process and communication network, etc. m 3j is the economic income generated by the load consuming electricity. m 4j is the economic value corresponding to the energy storage. Similarly, if the energy transaction price does not involve the power generation side, then m 1j For j = 1, 2, 3, 4, it can be taken as 0.

[0063] The method for constructing the economic benefit and energy efficiency calculation model used in the evaluation of the source-grid-load-storage integration project in S3 is as follows:

[0064] S31: Assume the number of operating years is N, and construct the main economic indicators Y = [y1 y2 y3] to evaluate the economic value of the investment in the integration project. Among them, y1 represents the total annual average economic return rate of the source-grid-load-storage in this region; y2 represents the annual average economic return rate of new energy in this region; y3 represents the annual average economic return rate of non-new energy; the specific calculations of each part are as follows:

[0065]

[0066] y3 = y1 - y2;

[0067] S32: Assume the number of operating years is N, and construct the power efficiency Z = [z1 z2 z3], where z1 represents the total annual energy output efficiency of the source-grid-load-storage; z2 represents the proportion of new energy power generation; z3 represents the absorption capacity of new energy. The specific calculations for each part are as follows:

[0068]

[0069] S33: Combine economic indicators and energy efficiency indicators to construct an economic benefit and energy efficiency calculation model for evaluating the source-grid-load-storage integration project, specifically described as:

[0070]

[0071] The method for evaluating the value of the source-grid-load-storage integration project separately for different supply and demand situations in S4 is as follows:

[0072] S41. When it indicates that the local power demand is large, and all types of power generation and supporting devices of the source-grid-load-storage need to operate at full load. Then, the economic benefits and energy efficiency of the source-grid-load-storage integration project adopt the economic benefit and energy efficiency calculation model [Y, Z] of S33 T for evaluation;

[0073] S42. When it indicates that the annual power demand is less than the planned annual available power. On the one hand, the proportion of new energy absorption can be maximized to increase the annual average economic return rate Y2 of new energy. On the other hand, it has the value of supplying power to areas outside the source-grid-load-storage integration project plan. The maximum power that can be supplied externally is The corresponding annual external power supply matrix O and price matrix U of the source-grid-load-storage integration project can be described as follows:

[0074]

[0075] Among them, o j j = 1, 2, 3, 4 successively represent the external power supply of hydropower, wind power, photovoltaic power, and thermal power, and u j j = 1, 2, 3, 4 successively represent the external selling prices of hydropower, wind power, photovoltaic power, and thermal power, and Therefore, the calculation methods of the total annual average economic return rate y1 of the source-grid-load-storage and the annual average economic return rate y2 of new energy are adjusted as follows:

[0076]

[0077] Since electrical energy can be supplied outside the interval, new energy output can be given priority. On the one hand, it can reduce the cost of local energy storage and provide space for subsequent new energy storage. On the other hand, it can also reduce the curtailment of new energy. Therefore, the evaluation methods for the corresponding total energy efficiency, new energy efficiency, and their consumption capabilities, as well as the calculation methods for energy efficiency, are adjusted as follows:

[0078]

[0079] Ideally, the actual maximum power supply of new energy is close to the annual planned available power, that is

[0080] Subsequently, the adjusted economic benefits and energy efficiency are input into the calculation model [Y, Z] T for evaluation;

[0081] S43. When and there is no need to supply power outside the interval, it means that there is a surplus of electrical energy supplied by the local source-network-load-storage. Then, the new energy consumption ratio can be given priority to maximize, improve the power generation and consumption capabilities of new energy and the annual average economic return rate Y2, and reduce the supply of other energy sources. In this case, the power generation and consumption ratios of each energy source will change, and its calculation method is as follows:

[0082]

[0083] Among them, λ i is used for i = 2, 3 to adjust the energy storage capacity and the consumption at the load end, 0 < λ i < 1. After calculating the values of each element of the new annual energy consumption demand and the annual electrical energy production-sales (KW·h) matrix B according to the new weight, then evaluate each index of the source-network-load-storage integration project based on step S33, and input it into the calculation model [Y, Z] T for evaluation.

[0084] The method for maximizing economic benefits and energy efficiency using the improved raccoon optimization algorithm in S5 is as follows:

[0085] S51. The initialization strategy for constructing the improved raccoon optimization algorithm is:

[0086]

[0087] x i+1 = ax i (1 - x i )

[0088] Among them, the model [Y, Z] TIt is a calculation model of economic benefit and energy efficiency improved based on the above three supply-demand situations. \(x_0\) is the initial situation of the optimization algorithm. \(a\in(0,4]\), and the larger \(a\) is, the higher the chaos. When \(a = 4\), it is in a completely chaotic state;

[0089] S52. The search strategy for constructing the improved raccoon optimization algorithm is as follows:

[0090]

[0091] Among them, N is the number of individuals, represents the current individual position, is the generated position of the \(j\)-th dimension. \(r\) and \(I\) are randomly selected integers in \((0,1)\) and \((1,2)\) respectively, is the fitness of the random position, \(F\) i is the fitness of the \(i\)-th individual;

[0092] S53. The self-adaptive optimal guidance strategy for constructing the improved raccoon optimization algorithm is as follows:

[0093]

[0094] u = R1 + R2×(1 - (T max - t) / T max ) 2

[0095] Among them, \(i = 1,2,\cdots,N\), \(j = 1,2,3\); N is the number of individuals, represents the current individual position, are the local upper and lower bounds of the \(j\)-th dimension respectively; R1 and R2 are random numbers between \((0,1)\);

[0096] S54. Initialize the parameters of the improved raccoon optimization algorithm; that is, use the economic benefit and energy efficiency in the scheduling model as the population;

[0097] S55. Obtain the optimal individual position of the current population, carry out maximization optimization with this individual and check the supply-demand relationship constraint for the result. If the check passes, the optimal economic benefit and energy efficiency allocation scheme can be obtained, that is, the optimal economic index \(Y = [y_1 y_2 y_3]\) and the electric energy efficiency \(Z = [z_1 z_2 z_3]\). Among them, \(y_1\) represents the total annual average economic return rate of the local source-network-load-storage described in S31, \(y_2\) represents the annual average economic return rate of the local new energy described in S31, \(y_3\) represents the annual average economic return rate of the non-new energy described in S31, \(z_1\) represents the total annual energy output efficiency of the source-network-load-storage in S32; \(z_2\) represents the proportion of new energy power generation in S32; \(z_3\) represents the accommodation capacity of new energy in S32. Otherwise, return to S54.

[0098] Compared with the prior art, the beneficial effects of the present invention are as follows:

[0099] By introducing a singular spectrum noise reduction method for the original monthly electricity demand data in a specified time window, using a dynamic convolutional network to read depth information, and using a convolutional time model to obtain prediction results, the present invention significantly improves the prediction accuracy. On the other hand, by introducing variational mode decomposition, error correction is performed and superimposed from three components. Further, the electricity demand prediction results are combined with the natural resource conditions of the project investment location to establish an annual available capacity model, and then an annual energy consumption demand and annual electricity production-sales correlation model is constructed. To evaluate the economic benefits of the source-grid-load-storage integration project, mathematical models for the construction costs, operation and maintenance costs of multiple links of the source-grid-load-storage, and electricity production, distribution, sales, and storage are respectively constructed, and the construction, operation and maintenance, production, sales of the source-grid-load-storage integration project are associated with the economic scale using the calculation method of the Hadamard product. For the economic benefits and energy efficiency concerned in the investment of the source-grid-load-storage integration project, corresponding evaluation index sets and corresponding calculation methods are designed. Based on the supply-demand balance of the source-grid-load-storage, economic and energy efficiency evaluation methods under different supply-demand states are given, providing an evaluation means for the comprehensive value of the investment in the source-grid-load-storage integration project and the optimal utilization of energy after completion. Finally, based on the supply-demand relationship, an effective evaluation index optimization method is provided.

[0100] In the present invention, the description of the energy types corresponding to the source-grid-load-storage mainly focuses on hydropower, wind power, photovoltaic power, and thermal power. When actually planning and designing a source-grid-load-storage integration project, one or more energy sources can be selected in combination with the local actual situation, or other energy sources can be replaced. The data corresponding to irrelevant energy sources can be replaced or deleted in the evaluation model used, and the basic method is the same, which has positive significance for the construction investment planning and design of the source-grid-load-storage integration project. Description of the Drawings

[0101] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the following will briefly introduce the drawings required for the description of the embodiments or the prior art. Obviously, the drawings in the following description are only exemplary, and for those of ordinary skill in the art, without creative efforts, other implementation drawings can be obtained according to the provided drawings.

[0102] The structures, ratios, sizes, etc. illustrated in this specification are only used to cooperate with the content disclosed in the specification for those familiar with this technology to understand and read, and are not used to limit the limiting conditions for the implementation of the present invention. Therefore, they do not have technical substance significance. Any modification of the structure, change of the ratio relationship, or adjustment of the size should still fall within the scope covered by the technical content disclosed in the present invention without affecting the effects that the present invention can produce and the purposes that can be achieved.

[0103] Figure 1 This is the flowchart of the steps of the present invention. Detailed implementation manners

[0104] To make the objectives, technical solutions and advantages of the embodiments of the present invention clearer, the technical solutions in the embodiments of the present invention will be clearly and completely described below. Obviously, the described embodiments are only a part of the embodiments of the present application, rather than all the embodiments. These descriptions are only for further explaining the features and advantages of the present invention, rather than limiting the claims of the present invention; based on the embodiments in the present application, all other embodiments obtained by those of ordinary skill in the art without creative efforts shall fall within the scope of protection of the present application.

[0105] The following will further describe in detail the specific implementation manners of the present invention with reference to the drawings and embodiments. The following embodiments are used to illustrate the present invention, but are not used to limit the scope of the present invention.

[0106] As Figure 1 shown, a method for designing and planning economic and energy indicators for a source-grid-load-storage integrated project specifically includes the following steps:

[0107] S1: Based on the monthly electricity demand prediction model of the area where the source-grid-load-storage integrated project is located, which is based on ODConv (Omni-Dimensional Dynamic Convolution)-BiTCN (Bidirectional Temporal Convolutional Networks) and decomposition error correction, obtain the prediction result of the annual energy consumption demand, and give the annual available electricity capacity model and the correlation model of the annual electricity production-sales volume.

[0108] S11: Collect historical electricity demand data with reduced noise. The specific method is to obtain monthly historical electricity demand data; use the Singular Spectrum Analysis (SSA) algorithm to obtain the first intermediate monthly historical electricity demand sequence without interference, that is, denoised.

[0109] Historical electricity data refers to the specific data sequence H(t) and the change trend sequence I(t) of the electricity demand for each month in several years in the area where the target source-grid-load-storage integrated project acts.

[0110] Singular Spectrum Analysis (SSA) mainly includes four steps: embedding - decomposition - grouping - reconstruction. It is a non - linear time - series data processing method that uses decomposition and reconstruction of the time - series matrix under study to extract seasonal trends, long - term trends, noise, etc. in the time series, so as to analyze or denoise the time series and be used for some other tasks. Input H(t) and I(t) to obtain the first intermediate monthly historical electricity demand series F for denoising.

[0111] S12: Construct a matrix for the collected data set. The method is to perform wavelet analysis on the first intermediate monthly electricity demand series F and the second intermediate monthly electricity demand series F n There are N in total, where N represents the level of wavelet decomposition. Based on the N second intermediate monthly electricity demand series, construct the first electricity demand matrix, and perform normalization processing to obtain the second monthly electricity demand matrix f1.

[0112] Wavelet Transform (WT) is a multi - resolution signal analysis tool based on stretching and translation of frequency and time windows. In theory, it will significantly improve the accuracy of electricity demand prediction. Wavelet analysis extracts feature information from signals, further solving calculation problems and improving accuracy compared with Fourier transform. After the original data is processed by wavelet analysis, N sub - sequences are obtained, and these N sub - sequences are normalized into a matrix for subsequent prediction processing.

[0113] Preferably, constructing the first monthly electricity demand matrix according to the N second intermediate monthly electricity demand series includes:

[0114] Let F n , n ∈ {1, 2, … N} represents the nth second intermediate monthly electricity demand series, then the first monthly electricity demand matrix is expressed as:

[0115]

[0116] where T represents transpose, f represents the first electricity demand matrix, and after normalization processing, the second monthly electricity demand matrix f1 is obtained.

[0117] S13: Based on the step size of the time series, divide the second monthly electricity demand matrix f1, and obtain the matrix slice with the step size [t - T, t - 1]. Based on the Omni - Dimensional Dynamic Convolution (ODConv), perform feature extraction on the input matrix to obtain the comprehensive feature f2, T>1.

[0118] ODConv is an improved convolutional neural network for processing image data. Its basic structure still includes a convolutional layer, a pooling layer, and a fully connected layer. The convolutional layer uses a convolutional kernel to perform a convolutional operation on the input to extract spatial information and features. The pooling layer downsamples the output of the convolutional layer to reduce the size of the feature map, thereby reducing the number of parameters and the amount of computation. The fully connected layer is used for the final classification or regression task.

[0119] In order to extract more implicit features from the original electricity demand sequence, this method is different from the dynamic convolution method using ODConv. It achieves dynamicity in four dimensions: channels, filters, space, and convolutional kernels through SE attention, improving the accuracy of lightweight and large CNNs while maintaining an efficient inference speed, providing better performance while reducing additional parameters. Therefore, ODConv is used to extract the features of the historical electricity demand sequence.

[0120] Preferably, the feature extraction training process includes:

[0121] Divide the second monthly electricity demand matrix according to the combination of a sliding window and slicing to obtain matrix slices with a time step of [t i ―T1, t i ―1], where t i ∈[t ― T + T1, t ― 1].

[0122] Take the matrix slice with a time step of [t i ―T1, t i ―1] as a training sample, and use the true monthly electricity demand data at time step t i as the label of the training sample to generate a training sample set.

[0123] For each training sample, concatenate the matrix slice with a time step of [t i ―T1, t i ―1] with other features at time step t i to obtain the comprehensive feature f2.

[0124] S14: Input the comprehensive feature of the existing matrix into the time prediction model of the Bidirectional Temporal Convolutional Networks (BiTCN) to obtain the predicted result of the monthly electricity demand at the current time step t.

[0125] The BiTCN (Bidirectional Temporal Convolution Network) is a variant of the Convolutional Neural Network (CNN). Convolutional Neural Networks are common in the field of vision, but their application in prediction is still insufficient. The BiTCN aims to use two temporal convolution networks, one responsible for encoding future covariates and the other for encoding past covariates and historical values of the sequence. In this way, the model can learn temporal information from the data, and the use of convolution maintains computational efficiency.

[0126] In specific implementation, BiTCN processes the input data through multiple convolutional layers. Each layer of convolution extracts features through different filters and activation functions, and at the same time uses dilated convolution to increase the receptive field, that is, to expand the coverage of the model for the input data without increasing parameters. In addition, by applying a regular network loss (Dropout) and L2 regularization techniques, the model can suppress overfitting and improve generalization ability.

[0127] Input the comprehensive features into the BiTCN convolutional time prediction model to obtain the monthly electricity demand prediction result at time step t i of the current time step t.

[0128] Preferably, the initial comprehensive prediction result of the monthly electricity demand at the current time step t includes:

[0129] d(t) = D t

[0130] where d(t) represents the comprehensive prediction result of the monthly electricity demand at the current time step t, and D t represents the prediction result of the monthly electricity demand at the current time step t.

[0131] S15: Input the prediction result D of the monthly electricity demand at the current time step t t into the Variational Mode Decomposition (VMD) model to obtain the first monthly electricity demand prediction component, the second monthly electricity demand prediction component, and the third monthly electricity demand prediction component at the current time step t. Use the trained error correction model to correct the results of the first monthly electricity demand prediction component, the second monthly electricity demand prediction component, and the third monthly electricity demand prediction component, and then splice them to obtain the final monthly electricity demand prediction result g(t).

[0132] VMD (Variational Mode Decomposition) is an adaptive and completely non - recursive method for modal variational and signal processing. Its adaptability is manifested in determining the number of modal decompositions of the given sequence according to the actual situation. During the subsequent search and solution process, it can adaptively match the optimal center frequency and finite bandwidth of each mode, and can achieve the frequency - domain division of the signal, thereby obtaining the effective components of the given signal and finally obtaining the optimal solution of the variational problem.

[0133] In the implementation of the present invention, this decomposition method is mainly used to divide the prediction result before correction into three segments for error correction, so as to effectively avoid the influence of volatility and reduce the non - stationarity of the monthly electricity demand prediction result sequence with high complexity and strong non - linearity, thereby ultimately achieving the effect of enhancing error correction.

[0134] Preferably, the correction of the initial comprehensive monthly electricity demand prediction result using the trained error correction model includes:

[0135] S151: Calculate the time step t k ∈[t - T2, t - 1] of the deviation between the initial monthly electricity demand prediction result and the real monthly electricity demand at time step t k to construct the error sequence

[0136] S152: Input the error sequence P t into the LSTM network model to predict the deviation between the initial comprehensive monthly electricity demand prediction result at the current time step t and the real monthly electricity demand at the current time step t.

[0137] S153: According to the initial comprehensive monthly electricity demand prediction result at the current time step t, it is decomposed by variational mode decomposition into the first monthly electricity demand prediction component G1, the second monthly electricity demand prediction component G2, the third monthly electricity demand prediction component G3, and the sum of the deviations between the three initial monthly electricity demand prediction components at the current time step t and the real monthly electricity demand at the current time step t to obtain the final monthly electricity demand prediction component at the current time step t. Combine the above - mentioned final monthly electricity demand prediction components to obtain the final monthly electricity demand prediction result g(t).

[0138] S16: Based on the monthly electricity demand estimation model g(t), the total annual electricity demand of industries, civil uses, and commerce in this region can be evaluated.

[0139] S17: Combining the natural conditions and electricity demand of the region planned for the source - network - load - storage integration project, the construction of the source - network - load - storage integration project needs to evaluate the maximum capacity (unit: KW) that can be built for various power stations on the source side. Therefore, a power station construction capacity matrix C is constructed:

[0140]

[0141] They represent the maximum capacity of hydropower, wind power, photovoltaic power, thermal power, etc. on the source side. It also represents the maximum capacity of electricity that can be provided in the area.

[0142] As different energy generation equipment and materials increase in operating time and are affected by local climate conditions, it is difficult to ensure that each energy source can always work at its designed capacity, and the power generation capacity is attenuated and the output is insufficient, especially for photovoltaic power generation, which will attenuate to a certain extent every year. At the same time, photovoltaic, wind power and hydropower will all be affected by annual meteorological conditions, and the output will also be different. In order to accurately evaluate the investment benefits of the source-grid-load-storage integration project in the future, the available capacity expression of each type of energy in the nth year is assumed to be:

[0143]

[0144] Among them, η j j=1, 2, 3, 4 represent the output capacity of different power stations, for thermal power η i It can be 1. Wind power and photovoltaic power can be set according to local historical meteorological laws. i is the annual average attenuation coefficient, and the annual attenuation coefficient of photovoltaic power generation is currently approximately 0.4% to 2%. Therefore, the annual (n) available capacity matrix of the power station on the source side can be determined, as described below:

[0145] C(n)=[c1(n) c2(n) c3(n) c4(n)]

[0146] c j (n)j=1, 2, 3, 4 represent the available capacity of hydropower, wind power, photovoltaic power, thermal power, etc. on the source side respectively. Let K represent the number of hours per year, K=365*24=8760, then the annual maximum power supply on the source side (n) is

[0147] S18: Due to the influence of geographical environment and climatic conditions, when new energy is able to generate sufficient electricity, it needs to be supplied to the load side in a timely manner. On the other hand, it can also be used for energy storage to reduce the abandonment rate of new energy and wind power, improve the consumption of new energy, and improve the balance of power supply. For this purpose, a correlation model between annual energy consumption demand and annual electricity production and sales (KW.h) matrix B is constructed, which is described as follows:

[0148]

[0149] Where W(t) is the total annual electricity demand for industry, civil and commerce in the region of S16, b 1jj = 1, 2, 3, 4 represents the local electric energy production volume, b 2j represents the loss in power transmission and distribution, b 3j represents the consumption at the load end, b 4j represents the energy storage volume, mainly for the energy storage description of photovoltaic and wind energy. If hydropower and thermal power power stations do not need to configure energy storage devices, the data in this part is zero. For other types of energy that require energy storage, the corresponding elements can be valued according to actual needs. Proportion of different energies, ideally To improve the consumption capacity of new energy, its proportion is usually the maximum consumption ratio on the condition of ensuring grid smoothness. According to the planned capacity and output capacity of new energy, the annual proportion of new energy can be roughly determined. At present, most domestic integrated source-network-load-storage projects are planned and constructed with the consumption of new energy electricity not less than 40% of the total electricity consumption. ξ j represents the losses generated by different types of power transmission and distribution and communication networks, etc. The annual electric energy production-sales (KW.h) matrix B can be described as:

[0150]

[0151] The above main calculations are for calculating the electric energy demand-sales model of different energy situations of the integrated source-network-load-storage project. In actual situations, values need to be taken and calculated according to the specific circumstances of project implementation. The matrix B provides a calculation model with the same structure for the subsequent evaluation of the electric energy efficiency and economic benefits of the integrated source-network-load-storage project.

[0152] S2: Uniformly describe the construction costs, later operation and maintenance costs, and electric energy production, line losses, terminal loads, energy storage and other links involved in the construction of the integrated source-network-load-storage project in economic quantities.

[0153] S21: For the capital investments in multiple links involved in the construction of the integrated source-network-load-storage project, specifically including the capital investments in the construction of hydropower, wind power, photovoltaic power, thermal power, etc. on the source side, the capital investments in power transmission and distribution construction on the grid side, the capital investments in the supporting equipment and lines for industrial electricity, commercial electricity, and residential electricity on the load side, and the capital investments in batteries and related equipment on the energy storage side. Then the construction investment cost (unit: ten thousand yuan) of the integrated source-network-load-storage can be described by the matrix A as:

[0154]

[0155] Among them, a j1 、a j2 、a j3 、a j4j = 1, 2, 3, 4 represent the total construction investment costs of different types of energy construction, namely the construction investments on the source side, grid side, load side, and energy storage side. Specifically, the first column represents the investment cost of wind power, the second column represents the investment cost of thermal power, the third column represents the investment cost of photovoltaic power, and the fourth column represents the investment cost of hydropower.

[0156] S22: Based on the uncertain risk input costs (in ten thousand yuan) of equipment and line operation and maintenance on the power generation, transmission, distribution, and user sides of hydropower, wind power, photovoltaic power, and thermal power, establish the operation and maintenance cost matrix D:

[0157]

[0158] In the operation and maintenance cost matrix D, each column represents the operation and maintenance costs on the source side, grid side, load side, and energy storage side respectively. Among them, the first column represents the operation and maintenance cost of hydropower, the second column represents the operation and maintenance cost of wind power, the third column represents the operation and maintenance cost of photovoltaic power, and the fourth column represents the operation and maintenance cost of thermal power.

[0159] S23: To facilitate the unified evaluation of the economic value of the integrated source-grid-load-energy storage project investment, convert the electricity production, sales, and purchase into the corresponding economic costs, described by the electricity price corresponding to KW.h. Considering that the prices of different types of energy may vary in each link of production, sales, and purchase, in order to accurately and effectively evaluate the overall economic benefits, construct the transaction price matrix P, and price each part separately. The price matrix P is described as follows:

[0160]

[0161] There are differences in the operation modes after the integrated source-grid-load-energy storage project is completed. Therefore, the values of each element in the price matrix P need to be determined in combination with the future operation mode, which helps to accurately evaluate the economic benefits of the integrated source-grid-load-energy storage project investment. The first column in matrix B represents the prices of each link of hydropower, the second column represents the prices of each link of wind power, the third column represents the prices of each link of photovoltaic power, and the fourth column represents the prices of each link of thermal power. In the integrated source-grid-load-energy storage project, if the energy transaction price does not involve the power generation end, then p 1j j = 1, 2, 3, 4 can take 0.

[0162] S24: Based on the calculation method of the Hadamard product, the electricity production, loss, load consumption, and energy storage can be converted into the corresponding economic benefit matrix M, which can be described as:

[0163]

[0164] Among them, ○ represents the Hadamard product calculation, which means a matrix with the same number of rows and columns makes the elements with the same subscripts multiply. m ij = b ij * p iji = 1, 2, 3, 4; j = 1, 2, 3, 4. m 1j represents the revenue obtained from selling electric energy on the source side (hydropower, wind power, photovoltaic power, thermal power), m 2j represents the economic loss caused by power loss in the power transmission and distribution process and communication network, etc., m 3j is the economic income generated by the load consuming electric energy, m 4j is the economic value corresponding to energy storage. Similarly, if the energy trading price does not involve the power generation side, then m 1j j = 1, 2, 3, 4 can be taken as 0.

[0165] S3: Combining the comprehensive value concerned by the investment in the source-grid-load-storage integration project, construct the economic benefit and energy efficiency index sets of the whole-process input and output of the source-grid-load-storage integration project, and constitute an evaluation model for describing the investment value of the source-grid-load-storage integration project.

[0166] S31: Assume the number of operating years is N, and construct the main economic indicators Y = [y1 y2 y3] to evaluate the economic value of the integrated project investment. Among them, y1 represents the total annual average economic return rate of the source-grid-load-storage in this region. y2 represents the annual average economic return rate of new energy in this region. y3 represents the annual average economic return rate of non-new energy. The specific calculations for each part are as follows:

[0167]

[0168] y3 = y1 - y2;

[0169] Other economic indicators can be calculated based on the above model. For example, when only considering the economic benefits of the source-grid-load-storage of a certain type of energy, the irrelevant elements in matrices A, D, and M can be taken as zero.

[0170] S32: Assume the number of operating years is N, and construct the power efficiency Z = [z1 z2 z3], which is used to describe the energy efficiency of the integrated project construction, the energy dispatch between regions, and the new energy consumption situation. Among them, z1 represents the total annual energy output efficiency of the source-grid-load-storage, reflecting the power supply and demand balance situation in this region, and also reflecting the utilization rate of the source-grid-load-storage in this region. z2 represents the proportion of new energy power generation, reflecting the new energy output capacity in this region. z3 represents the new energy consumption capacity. The specific calculations for each part are as follows:

[0171]

[0172] Other energy efficiency indicators can also be calculated based on the above model. For example, when only considering the production efficiency or energy storage efficiency of a certain type of energy, the irrelevant elements in matrices B and C(n) can be taken as zero.

[0173] S33: Combine economic indicators and energy efficiency indicators to construct a calculation model for economic benefits and energy efficiency used in the evaluation of the source-grid-load-storage integration project, which is specifically described as follows:

[0174]

[0175] S4: Based on the potential supply-demand relationship of the source-grid-load-storage integration project, calculation methods for various economic and energy efficiency indicators under different supply-demand states are given, providing an evaluation method for the potential economic benefits and energy efficiency of the source-grid-load-storage integration project, and further providing support for the optimal utilization of energy after completion.

[0176] Combined with the natural conditions and energy demand scale in different regions, and on the condition of ensuring reliable and stable energy supply and demand, value evaluations of the source-grid-load-storage integration project are carried out respectively for different supply-demand situations, which specifically include the following three cases:

[0177] S41: When it indicates that the local electricity demand is large, and all types of power generation and supporting devices in the source-grid-load-storage need to operate at full load. Then, the economic benefits and energy efficiency of the source-grid-load-storage integration project adopt the economic benefit and energy efficiency calculation models in S33 [Y, Z] T for evaluation.

[0178] S42: When it indicates that the annual electricity demand is less than the planned annual available electricity. On the one hand, the proportion of new energy consumption can be maximized to increase the annual average economic return rate Y2 of new energy. On the other hand, it has the value of supplying electricity to areas outside the source-grid-load-storage integration project area, and the maximum electricity that can be supplied externally is The corresponding annual external power supply matrix O and price matrix U of the source-grid-load-storage integration project can be described as follows:

[0179]

[0180] Among them, o j j = 1, 2, 3, 4 represent the power supply to the outside for hydropower, wind power, photovoltaic power, and thermal power in sequence, and u j j = 1, 2, 3, 4 represent the selling prices to the outside for hydropower, wind power, photovoltaic power, and thermal power in sequence, and (t). Therefore, the calculation methods for the total annual average economic return rate y1 of the source-grid-load-storage and the annual average economic return rate y2 of new energy are adjusted as follows:

[0181]

[0182] Since electrical energy can be supplied outside the interval, new energy output can be given priority. On the one hand, it can reduce the cost of local energy storage and provide space for subsequent new energy storage. On the other hand, it can also reduce the curtailment of new energy. Therefore, the evaluation methods for the corresponding total energy efficiency, new energy efficiency, and their consumption capacity, as well as the calculation method of energy efficiency, are respectively adjusted as follows:

[0183]

[0184] Ideally, the actual maximum power supply of new energy is close to the available power in the annual plan, that is

[0185] Subsequently, the adjusted economic benefits and energy efficiency are input into the calculation model [Y, Z] T for evaluation;

[0186] S43: When and there is no need to supply power outside the interval, it means that there is a surplus of electrical energy supplied by the local source-network-load-storage. Then, the new energy consumption ratio can be given priority to maximize, improve the power generation and consumption capacity of new energy and the annual average economic return rate Y2, and reduce the supply of other energy sources. In this case, the power generation and consumption ratios of each energy will change, and its calculation method is as follows:

[0187]

[0188] Among them, λ i (i = 2, 3) is used to adjust the energy storage capacity and the consumption at the load end, 0 < λ i < 1. For this reason, for the new annual energy consumption demand and the annual electrical energy production-sales (KW.h) matrix B, only after calculating the values of its elements according to the new weights, and then evaluating each index of the source-network-load-storage integration project based on step S33.

[0189] S5: Taking the supply-demand relationship described above for the source-network-load-storage integration project as a constraint, based on the updated economic benefit and energy efficiency models in three cases, using the intelligent search algorithm with the maximization of economic benefit and energy efficiency as the optimization goal to perform optimization, and obtaining the optimal parameters of economic benefit and energy efficiency.

[0190] The intelligent search algorithm adopted in this embodiment is the improved coati optimization algorithm (Improved Coati Optimization Algorithm, ICOA).

[0191] The Coati Optimization Algorithm is a heuristic optimization algorithm proposed by M Dehghani et al. in 2023, which is based on the population behavior of coatis in nature. This algorithm mainly solves optimization problems by simulating the behavior of coatis hunting iguanas and escaping from predators. Compared with other well-known heuristic algorithms, it has obvious advantages by balancing the utilization of global search and local search. The key of the Coati Optimization Algorithm lies in the steps of hunting iguanas or exploration and escaping from predators or exploitation. In the stage of hunting iguanas, the strategy of attacking iguanas is simulated, enabling the algorithm to move to different positions in the search space, which reflects the global search ability in the problem space. In the stage of escaping from predators, the simulation of coatis constantly escaping from predators makes the algorithm approach a safe position near the current location, which is the exploitation ability of the COA in local search.

[0192] The improved Coati Optimization Algorithm adds chaotic mapping, flight strategy, and lens imaging refraction reverse learning strategy on the basis of the original algorithm to solve nonlinear problems and effectively avoid falling into local optima and expand the search range.

[0193] For those skilled in the art, given the constraints of the intelligent search algorithm and the parameters to be updated, according to the characteristics of different intelligent search algorithms, those skilled in the art can undoubtedly obtain how to update the target parameters using the intelligent search algorithm. Common intelligent search algorithms include: Coyote Algorithm, Particle Swarm Optimization, Whale Optimization Algorithm, Condor Search Algorithm, or Ant Colony Algorithm, etc.

[0194] S51: The initialization strategy for constructing the Improved Coati Optimization Algorithm (ICOA) is as follows:

[0195]

[0196] x i+1 = ax i (1 - x i )

[0197] where the model [Y, Z] T is the calculation model of economic benefit and energy efficiency improved based on the above three supply and demand situations, x0 is the initial situation of the optimization algorithm, a ∈ (0, 4], and the higher the value of a, the higher the chaos. When a = 4, it is in a completely chaotic state.

[0198] S52: The search strategy for constructing the Improved Coati Optimization Algorithm (ICOA) is as follows:

[0199]

[0200] where, N is the number of individuals, represents the current individual position, is the generated position of the j-th dimension. r and I are randomly selected integers from (0, 1) and (1, 2) respectively, is the fitness of the random position, F i is the fitness of the i-th individual.

[0201] S53: The self-adaptive optimal guiding strategy for constructing the Improved Coati Optimization Algorithm (ICOA) is as follows:

[0202]

[0203] u = R1 + R2 × (1 - (T max - t) / T max ) 2

[0204] where i = 1, 2, …, N, j = 1, 2, 3. N is the number of individuals, represents the current individual position, are the local upper and lower bounds of the j-th dimension respectively. R1 and R2 are random numbers between (0, 1).

[0205] S54: Initialize the parameters of the improved coati optimization algorithm; that is, take the economic benefits and energy efficiency in the scheduling model as the population.

[0206] S55: Obtain the optimal individual position of the current population, and carry out maximization optimization with this individual and check the supply-demand relationship constraint for the result. If the check passes, the optimal economic benefit and energy efficiency allocation scheme can be obtained, that is, the optimal economic index Y = [y1 y2 y3] and the power efficiency Z = [z1 z2 z3], where y1 represents the total annual average economic return rate of the local source-network-load-storage in S31, y2 represents the annual average economic return rate of the local new energy in S31, y3 represents the annual average economic return rate of the non-new energy in S31, z1 represents the total annual energy output efficiency of the source-network-load-storage in S32. z2 represents the proportion of this new energy power generation in S32. z3 represents the accommodation capacity of the new energy in S32. Otherwise, return to S54.

[0207] The above only elaborates on the preferred embodiments of the present invention in detail. However, the present invention is not limited to the above embodiments. Within the scope of knowledge possessed by those of ordinary skill in the art, various changes can be made without departing from the purpose of the present invention, and all such changes should be included within the protection scope of the present invention.

Claims

1. A design and planning method for economic and energy indicators for source-grid-load-storage, characterized in that, It includes the following steps: S1. Construct a monthly electricity demand prediction model for the area where the source-grid-load-storage integration project is located based on a full-dimensional dynamic convolution - bidirectional time series convolution network and decomposition error correction. Combine the natural resource conditions of the project area to construct an annual available capacity model for power station generation, and finally establish an association model between annual energy consumption demand and annual electricity production - sales volume; The method for establishing the association model between annual energy consumption demand and annual electricity production - sales volume in S1 is as follows: S11. Collect historical electricity demand data with reduced noise. The specific method is to obtain monthly historical electricity demand data H(t) and I(t); use the singular spectrum analysis algorithm to obtain the first intermediate monthly historical electricity demand sequence F without interference, that is, denoised; S12. Construct a matrix for the collected data set. The method is to perform wavelet analysis on the first intermediate monthly power demand sequence F to obtain the second intermediate monthly power demand sequence F n There are N in total, where N represents the level of wavelet decomposition; based on the N second intermediate monthly power demand sequences F n , construct the first power demand matrix f, and perform normalization processing to obtain the second monthly power demand matrix f1; S13. Based on the time series step size, divide the second monthly electricity demand matrix f1, and obtain the matrix slice with the step size [t - T, t - 1]. Based on the dynamic convolution neural network, perform feature extraction on the input matrix to obtain the comprehensive feature f2, where T > 1; S14. Input the comprehensive feature f2 of the existing matrix into the two-way time series convolutional network time prediction model to obtain the monthly electricity demand prediction result D at the current time step t t ; Input the comprehensive feature into the BiTCN convolutional time prediction model to obtain the time step t i of the monthly electricity demand prediction result; d(t) = D t Among them, d(t) represents the comprehensive prediction result of the monthly electricity demand at the current time step t, and D t represents the prediction result of the monthly electricity demand at the current time step t; S15. Input the monthly electricity demand prediction result D at the current time step t t into the variational mode decomposition model to obtain the first monthly electricity demand prediction component G1, the second monthly electricity demand prediction component G2, and the third monthly electricity demand prediction component G3 at the current time step t; use the trained error correction model to correct the results of the first monthly electricity demand prediction component, the second monthly electricity demand prediction component, and the third monthly electricity demand prediction component, and then splice them to obtain the final monthly electricity demand prediction result g(t); S16. Based on the monthly electricity demand forecast result g(t), evaluate the total annual electricity demand of industries, households and businesses in this region S17. Combine the natural conditions and electricity demand in the planning area of the source-grid-load-storage integration project. The source-grid-load-storage integration project construction needs to evaluate the maximum capacity that can be built for various power stations on the source side, and construct a power station construction capacity matrix C: They represent the maximum capacity of hydropower, wind power, photovoltaic power, thermal power, etc. on the source side. It also represents the maximum capacity of electrical energy that can be provided in the area; Let the expression for the available capacity of various energy sources in the nth year be: Among them, η j j = 1, 2, 3, 4 represents the output capacity of different power generation stations. For thermal power, η j can take 1. For wind power and photovoltaic power, the values can be determined according to the local historical meteorological laws. σ j is the annual average attenuation coefficient. Currently, the annual attenuation coefficient of photovoltaic power is roughly between 0.4% and 2%. Thus, the annual (n) available capacity matrix of the power stations on the source side can be determined and is described as follows: C(n) = [c1(n) c2(n) c3(n) c4(n)] c j (n) j = 1, 2, 3, 4 represent the available capacities of hydropower, wind power, photovoltaic power, thermal power, etc. on the source side in sequence. Let K represent the number of hours per year, K = 365 * 24 = 8760, then the annual (n) maximum power supply from the source side is S18. Construct an association model between the annual energy consumption demand and the annual electricity production - sales volume (KW.h) matrix B, which is specifically described as follows: Among them, W(t) is the total annual electricity demand of local industry, civil use and commerce in S16, b 1j j = 1, 2, 3, 4 represents the local electricity production volume, b 2j represents the loss in power transmission and distribution, b 3j represents the consumption at the load end, b 4j represents the energy storage amount, mainly for the energy storage description of photovoltaic and wind energy. If energy storage devices do not need to be configured for hydropower and thermal power power stations, the data in this part is zero. For other types of energy that require energy storage, the corresponding elements can be valued according to actual needs, j = 1, 2, 3, 4 represents the proportion of different energy sources. Ideally, ξ j represents the losses generated by different types of power transmission and distribution and communication networks, etc. The annual electricity production - sales (KW.h) matrix B can be described as: S2. Uniformly describe the construction cost, later operation and maintenance cost, and electricity production, line loss, terminal load, and energy storage links in the construction of the source-grid-load-storage integration project with economic quantities; S3. Combine the comprehensive value concerned by the investment in the source-grid-load-storage integration project to construct calculation models for economic benefits and energy efficiency used in the evaluation of the source-grid-load-storage integration project; S4. Based on the potential supply - demand relationship of the source-grid-load-storage integration project, calculate various indicators of economic and energy efficiency under different supply - demand states, and conduct value evaluation of the source-grid-load-storage integration project for different supply - demand situations respectively; S5. Combine the calculation models for economic benefits and energy efficiency corrected by the value evaluation model under different supply - demand situations, and use the improved raccoon optimization algorithm to optimize the maximization of economic benefits and energy efficiency and make it meet the supply - demand constraint verification.

2. The design and planning method of economic and energy indicators for source-network-load-storage according to claim 1, characterized in that The method for obtaining the denoised monthly historical electricity demand sequence in S11 is as follows: Obtain the specific data sequence H(t) and the change trend sequence I(t) of the electricity demand for each month in the area affected by the target source-grid-load-storage integration project over several years; Input H(t) and I(t) into the singular spectrum analysis including embedding, decomposition, grouping, and reconstruction to extract the seasonal trend, long-term trend, and noise in the time series, realize the denoising function, and obtain the denoised first intermediate monthly historical electricity demand sequence F.

3. A design and planning method for economic and energy indicators for source-network-load-storage, according to claim 1, characterized in that, The method for constructing the first monthly electricity demand matrix according to N second intermediate monthly electricity demand sequences in S12 is as follows: Let F n , n ∈ {1, 2, … N} represent the n-th second intermediate monthly electricity demand sequence, then the first monthly electricity demand matrix is expressed as: Among them, T represents transpose, and f represents the first electricity demand matrix; after normalization, the second monthly electricity demand matrix f1 is obtained.

4. A design and planning method for economic and energy indicators for source-network-load-storage, according to claim 1, characterized in that The method for the feature extraction training process in S13 is as follows: Divide the second monthly electricity demand matrix according to the combination of sliding window and slicing to obtain matrix slices with a time step of [t i -T1, t i -1], where t i ∈ [t - T + T1, t - 1]; Take the matrix slice with a time step of [t i -T1, t i -1] as the training sample, and use the true monthly electricity demand data at time step t i as the label of the training sample to generate a training sample set; For each training sample, the matrix slice with time steps [t i - T1, t i - 1] is concatenated with other features at time step t i to obtain the comprehensive feature f2.

5. A design and planning method for economic and energy indicators for source-grid-load-storage, according to claim 1, characterized in that The method for correcting the initial comprehensive monthly electricity demand prediction result using the trained error correction model in S15 is as follows: Calculate time step t k The deviation between the initial monthly electricity demand prediction result within [t - T2, t - 1] and the true monthly electricity demand at time step t k is used to construct a composition error sequence The error sequence P t is input into the LSTM network model to predict the deviation between the comprehensive prediction result of the initial monthly electricity demand at the current time step t and the true monthly electricity demand at the current time step t; Based on the initial comprehensive monthly electricity demand prediction result at the current time step t, it is decomposed by variational mode decomposition into the first monthly electricity demand prediction component G1, the second monthly electricity demand prediction component G2, the third monthly electricity demand prediction component G3, and the deviation between the three initial monthly electricity demand prediction components at the current time step t and the true monthly electricity demand at the current time step t is added to obtain the final monthly electricity demand prediction component at the current time step t. The above final monthly electricity demand prediction components are combined to obtain the final monthly electricity demand prediction result g(t).

6. The design and planning method of economic and energy indicators for source-network-load-storage according to claim 1, characterized in that The method for uniformly describing the static construction cost, operation and maintenance cost, and electricity production, sales and use economic means in the source-grid-load-storage integration project in S2 is as follows: S21. The capital investments in multiple links involved in the construction of the source-grid-load-storage integration project, specifically including the construction capital investments in hydropower, wind power, photovoltaic power, thermal power, etc. on the source side, the capital investments in transmission and distribution construction on the grid side, the capital investments in supporting equipment and lines for industrial electricity, commercial electricity, and residential electricity on the load side, and the capital investments in batteries and related equipment on the energy storage side. Then, the construction investment cost of the source-grid-load-storage can be described by the matrix A as follows: Among them, a j1 、a j2 、a j3 、a j4 j = 1, 2, 3, 4 respectively represent the whole-process construction investment costs of different types of energy construction, that is, the construction investments on the source side, grid side, load side, and energy storage side. Specifically, the first column represents the investment cost of hydropower, the second column represents the investment cost of wind power, the third column represents the investment cost of photovoltaic power, and the fourth column represents the investment cost of thermal power; S22. According to the uncertain risk investment costs of power production, transmission and distribution, and equipment and line operation and maintenance of hydropower, wind power, photovoltaic power, and thermal power, a matrix D of operation and maintenance costs is established: In the matrix D of operation and maintenance costs, each column represents the operation and maintenance costs of the source side, grid side, load side, and energy storage side respectively. Among them, the first column represents the operation and maintenance cost of hydropower, the second column represents the operation and maintenance cost of wind power, the third column represents the operation and maintenance cost of photovoltaic power, and the fourth column represents the operation and maintenance cost of thermal power; S23. To facilitate the unified evaluation of the economic value of the investment in the source-grid-load-storage integration project, the electricity production and sales are uniformly converted into the corresponding economic costs, described by the electricity price corresponding to KW.h. The prices of different types of energy may be different in each link of production and sales. A transaction price matrix P is constructed to price the corresponding prices of each part respectively. The price matrix P is described as follows: The first column in matrix B represents the prices of various links in hydropower, the second column represents the prices of various links in wind power, the third column represents the prices of various links in photovoltaic power, and the fourth column represents the prices of various links in thermal power. In the integrated project of power source, grid, load and energy storage, if the energy trading price does not involve the power generation side, then p 1j For j = 1, 2, 3, 4, it can be taken as 0; S24. Based on the calculation method of the Hadamard product, the electricity production, loss, load consumption, and energy storage are converted into the corresponding economic benefit matrix M, which can be described as: Among them, ○ represents the Hadamard product calculation, which means a matrix with the same total number of rows and columns, where elements with the same subscript are multiplied, m ij = b ij · p ij i = 1, 2, 3, 4 j = 1, 2, 3, 4, m 1j represents the revenue obtained from selling electric energy generated by hydropower, wind power, photovoltaic power, and thermal power on the source side, m 2j represents the economic loss caused by power loss in the power transmission and distribution process and communication network, etc., m 3j is the economic income generated by the load consuming electric energy, m 4j is the economic value corresponding to energy storage. Similarly, if the energy trading price does not involve the power generation side, then m 1j j = 1, 2, 3, 4 can be taken as 0.

7. The design and planning method of economic and energy indicators for source-network-load-storage according to claim 1, characterized in that The method for constructing the economic benefit and energy efficiency calculation models used in the evaluation of the source-grid-load-storage integration project in S3 is as follows: S31: Assume the number of operating years is N, and construct the main economic indicators Y = [y1 y2 y3] to evaluate the economic value of the investment in the integration project. Among them, y1 represents the total annual average economic return rate of the source-grid-load-storage in this region; y2 represents the annual average economic return rate of new energy in this region; y3 represents the annual average economic return rate of non-new energy. The specific calculations for each part are as follows: y3 = y1 - y2; S32: Assume the number of operating years is N, and construct the electricity efficiency Z = [z1 z2 z3]. Among them, z1 represents the total annual energy output efficiency of the source-grid-load-storage; z2 represents the proportion of new energy power generation; z3 represents the consumption capacity of new energy. The specific calculations for each part are as follows: S33: Combine economic indicators and energy efficiency indicators to construct calculation models for economic benefits and energy efficiency used in the evaluation of source-grid-load-storage integration projects, specifically described as follows:

8. A design and planning method for economic and energy indicators for source-network-load-storage, characterized in that, The method for evaluating the value of source-grid-load-storage integration projects separately for different supply-demand situations in S4 is as follows: S41. When indicates that the local electricity demand is large and all types of power generation and supporting devices in the source-grid-load-storage need to operate at full load, the economic benefits and energy efficiency of the source-grid-load-storage integration project can be evaluated using the economic benefit and energy efficiency calculation models of S33 [Y, Z] T for assessment; S42. When it indicates that the annual electricity demand is less than the planned annual available electricity. On the one hand, the proportion of new energy consumption can be maximized to increase the annual average economic return rate Y2 of new energy. On the other hand, it has the value of supplying electricity to areas outside the planning of the source-network-load-storage integration project. The maximum electricity that can be supplied externally is The annual electricity supply matrix O and its price matrix U of the corresponding source-network-load-storage integration project to the outside can be described as follows: Among them, o j j = 1, 2, 3, 4 represent the external power supply of hydropower, wind power, photovoltaic power, and thermal power in sequence, and u j j = 1, 2, 3, 4 represent the selling prices of hydropower, wind power, photovoltaic power, and thermal power sold to the outside in sequence, and Therefore, the calculation methods of the total annual average economic return rate y1 of the power source, grid, load, and storage and the annual average economic return rate y2 of new energy are adjusted to: Since electrical energy can be supplied outside the interval, the output of new energy can be given priority. On the one hand, it can reduce the cost of local energy storage and provide space for subsequent new energy storage. On the other hand, it can also reduce the curtailment of new energy. Therefore, the evaluation methods for the total energy efficiency, new energy efficiency, and their consumption capacity, as well as the calculation methods for their energy efficiency, are adjusted as follows: Ideally, the actual maximum power supply of new energy is close to the available power in the annual plan, that is Subsequently, input the adjusted economic benefits and energy efficiency into the calculation model [Y, Z] T and then conduct the evaluation; S43. When and there is no need to supply power outside the interval, it indicates that there is a surplus of power supplied by the local source-network-load-storage. Then, the proportion of new energy consumption can be preferentially maximized to improve the power generation capacity and consumption capacity of new energy and the annual average economic return rate Y2, and reduce the supply of other energy sources. In this case, the power generation and consumption proportions of each energy source j = 1, 2, 3, 4 will change, and its calculation method is as follows: Among them, λ i is for i = 2, 3 to adjust the energy storage capacity and the consumption at the load end, 0 < λ i < 1. After the new annual energy consumption demand and the annual electric energy production - sales volume (KW.h) matrix B only calculate the values of its respective elements according to the new weights, then based on the steps of S33, evaluate each index of the source - grid - load - storage integration project and input it into the calculation model [Y, Z] T and then it can be evaluated.

9. A design and planning method for economic and energy indicators for source-network-load-storage, characterized in that, The method for maximizing economic benefits and energy efficiency using the improved raccoon optimization algorithm in S5 is as follows: S51: The initialization strategy for constructing the improved raccoon optimization algorithm is as follows: x i+1 = ax i (1 - x i ) Among them, the model [Y, Z] T is a calculation model of economic benefits and energy efficiency improved based on the above three supply-demand situations. x0 is the initial situation of the optimization algorithm. a ∈ (0, 4], and the greater the value of a, the higher the degree of chaos. When a = 4, it is in a completely chaotic state; S52: The search strategy for constructing the improved raccoon optimization algorithm is as follows: Among them, j = 1, 2, 3, N is the number of individuals, represents the current individual position, is the generated j-th dimensional position, r and I are randomly selected integers in (0, 1) and (1, 2) respectively, is the fitness of the random position, F i is the fitness of the i-th individual; S53: The adaptive optimal guidance strategy for constructing the improved raccoon optimization algorithm is as follows: u = R1 + R2×(1 - (T max - t) / T max ) 2 where \(i = 1, 2, \ldots, N\), \(j = 1, 2, 3\); \(N\) is the number of individuals, represents the current individual position, are the local upper and lower bounds of the \(j\)-th dimension respectively; \(R1\) and \(R2\) are random numbers between \((0, 1)\); S54: Initialize the parameters of the improved raccoon optimization algorithm; that is, use the economic benefits and energy efficiency in the scheduling model as the population. S55: Obtain the optimal individual position of the current population, perform maximization optimization starting from this individual, and perform a supply-demand relationship constraint check on the results. If the check passes, the optimal configuration plan for economic benefits and energy efficiency is obtained, that is, the optimal economic indicators Y = [y1 y2 y3] and power efficiency Z = [z1 z2 z3], where y1 represents the total annual average economic return rate of the local source-grid-load-storage in S31, y2 represents the annual average economic return rate of the local new energy in S31, y3 represents the annual average economic return rate of non-new energy in S31, z1 represents the total annual energy output efficiency of the source-grid-load-storage in S32; z2 represents the proportion of new energy power generation in S32; z3 represents the consumption capacity of new energy in S32. Otherwise, return to S54.

Images