CN114757107A - Gravity energy storage power distribution method based on load prediction model - Google Patents

Abstract

A gravity energy storage power distribution method based on a load prediction model relates to a gravity energy storage power distribution technology, in order to solve the problem that the existing load prediction accuracy is low, resulting in unreasonable power distribution of the gravity energy storage. The present invention establishes a load prediction model based on a GAN network; and uses it to predict the load of the gravity energy storage system, and outputs the prediction result; based on the prediction result, the demand value for peak regulation and the demand value for frequency regulation are calculated; combined with the constraints of the gravity energy storage system conditions, and using the particle swarm algorithm to obtain the value of the peak shaving participation factor and the value of the frequency regulation participation factor; the peak shaving participation factor and the frequency regulation participation factor are used to jointly build a power distribution model with the goal of maximizing the economic benefits of the gravity energy storage system ; Use the power distribution model to distribute power to the gravity energy storage system. The beneficial effect is that the power distribution of gravity energy storage is more reasonable.

Description

A Gravity Energy Storage Power Distribution Method Based on Load Forecasting Model

technical field

The invention relates to a gravity energy storage power distribution technology.

Background technique

A basic requirement for the stable operation of the power system is that the power should be balanced in real time; with the change of the load characteristics of the power system, the peak-to-valley difference of the power system gradually increases, and the problem of peak regulation becomes increasingly prominent; at the same time, the short-term difference between the power supply and the load The imbalance of active power and the fluctuation of system frequency have always affected the safe and stable operation of the power grid; as a new type of energy storage system, gravity energy storage technology has a high degree of adaptability to the geographical environment, and is expected to become the future power grid peak regulation and frequency regulation. The load forecasting of the power system is the basis for the implementation of various user-oriented applications, and the accurate load forecasting formulates a reasonable production plan for the power system to avoid waste of resources, ensure the safe and reliable operation of the power grid, and improve economic benefits. It plays an important role; according to the length of the forecast period, load forecasting can be divided into long-term, medium-term, short-term and ultra-short-term load forecasting; among them, the long-term and medium-term load changes tend to be stable, and their research is very mature; while the short-term load changes Random and difficult to predict, it has always attracted the attention of researchers; at the end of the 20th century, statistical methods were used for load forecasting for modeling and forecasting; these methods were constructed for linear relationships, ignoring factors such as climate and date type. The impact of prediction is low, and the prediction accuracy is low; therefore, the peak regulation and frequency regulation capacity of the power system is limited, and the power of gravity energy storage cannot be reasonably allocated.

SUMMARY OF THE INVENTION

The purpose of the present invention is to solve the problem that the existing load prediction accuracy is low, which leads to unreasonable power distribution of gravity energy storage, and proposes a gravity energy storage power distribution method based on a load prediction model.

A method for power distribution of gravity energy storage based on a load prediction model according to the present invention, the power distribution method includes the following steps:

Step 1. Establish a load prediction model based on GAN network;

Step 2, using the load prediction model established in step 1 to predict the load of the gravity energy storage system, and output the prediction result of the load of the gravity energy storage system;

Step 3, calculating the peak regulation demand value and the frequency regulation demand value of the gravity energy storage system based on the load prediction result of the gravity energy storage system output in step 2;

Step 4: Calculate the constraints of the gravity energy storage system;

Step 5: Combine the constraints of the gravity energy storage system calculated in Step 4, and use the particle swarm algorithm to optimize the peak regulation demand value, and obtain the value of the peak regulation participation factor; The optimization solution is obtained to obtain the value of the FM participation factor;

Step 6: Using the peak regulation participation factor and the frequency regulation participation factor obtained in Step 5 to jointly build a power distribution model with the goal of maximizing the economic benefit of the gravity energy storage system;

Step 7: Use the power distribution model constructed in step 6 to distribute power to the gravity energy storage system.

The beneficial effects of the invention are: the power distribution method comprehensively considers the cost and benefit of the gravity energy storage system participating in peak regulation and frequency regulation, and combined with the capacity limitation of the gravity energy storage system, and proposes a technical and economical method of gravity energy storage participating in frequency regulation and peak regulation. The model, by introducing the particle swarm algorithm to solve the demand value of peak regulation and the demand value of frequency regulation, realizes the reasonable distribution of the power and capacity of the gravity energy storage system participating in the auxiliary services of the power system.

Description of drawings

1 is a flowchart of a method for distributing power of gravity energy storage based on a load prediction model according to Embodiment 1;

2 is a schematic diagram of the network structure of a GAN network in the first embodiment;

3 is a structural block diagram of a load prediction model in Embodiment 1;

4 is a schematic structural diagram of a generator in Embodiment 1;

5 is a schematic structural diagram of a discriminator in Embodiment 1;

6 is a flowchart of a specific method for obtaining a peak regulation participation factor in Embodiment 1;

FIG. 7 is a graph showing the comparison between the optimal power distribution and the conventional distribution scheme in the first embodiment.

Detailed ways

Embodiment 1: This embodiment will be described with reference to FIG. 1 to FIG. 7 . The method for distributing gravity energy storage power based on a load prediction model described in this embodiment includes the following steps:

Step 1. Establish a load prediction model based on GAN network;

Step 2, using the load prediction model established in step 1 to predict the load of the gravity energy storage system, and output the prediction result of the load of the gravity energy storage system;

Step 3, calculating the peak regulation demand value and the frequency regulation demand value of the gravity energy storage system based on the load prediction result of the gravity energy storage system output in step 2;

Step 4: Calculate the constraints of the gravity energy storage system;

Step 5: Combine the constraints of the gravity energy storage system calculated in Step 4, and use the particle swarm algorithm to optimize the peak regulation demand value, and obtain the value of the peak regulation participation factor; The optimization solution is obtained to obtain the value of the FM participation factor;

Step 6: Using the peak regulation participation factor and the frequency regulation participation factor obtained in Step 5 to jointly build a power distribution model with the goal of maximizing the economic benefit of the gravity energy storage system;

Step 7: Use the power distribution model constructed in step 6 to distribute power to the gravity energy storage system.

In this embodiment, the load prediction model in step 1 includes a generator and a discriminator;

The generator is used to learn the distribution of the real sample r and generate new samples; wherein, the real sample r is the load data of the gravity energy storage system obtained through actual measurement;

The discriminator, which is used to discriminate whether the input sample comes from the generator.

In this embodiment, after the generator learns the deep relationship implied by the data and reaches a balance, the theoretical load prediction model outputted by the load prediction result of the gravity energy storage system is infinitely close to the real data.

In this embodiment, the specific process of establishing a GAN network-based load prediction model in step 1 is: input random noise n and condition value c into the generator, generate a new sample F(n|c), and then generate a sample F (n|c) and the real sample r are respectively input to the discriminator together with the condition c for discrimination; the discriminator output discrimination results are fed back to the generator and discriminator respectively in the form of a loss function, and the generator and discriminator are then based on the feedback results. Modify its own parameters; among them, the condition value c is the load influencing factor of the gravity energy storage system; the random noise n is a random variable conforming to the Gaussian distribution;

The loss function of the generator in the load forecasting model is defined as: L _F , and the loss function of the discriminator in the load forecasting model is defined as: L _D ;

The form of the loss function _LF of the generator is shown in formula (1):

L _F =-E _n,c (D(F(n|c)|c)) (1)

Among them, En _,c represents the expected value of the distribution of random noise n and condition value c; F represents the output data of the generator; D represents the output data of the discriminator;

The form of the loss function _LD of the discriminator is shown in formula (2):

L _D =-E _r,c (D(r|c))+E _n,c (D(F(n|c)|c)) (2)

Among them, E _{r, c} represents the expected value of the distribution of the real sample r and the conditional value c;

The generator hopes to increase the output value of F(n|c), while the discriminator hopes to reduce the output value of F(n|c) and increase the output value of the measured data r; in the process of load forecasting, the task of the generator is to Generate a predicted value that is as close to the measured load data as possible. After splicing the noise n and the load influencing factor c into the generator, the generator outputs the predicted load data F(n|c); the discriminator not only needs to judge the predicted load data F(n|c) ) and the measured load data r, it is also necessary to judge the fit between the predicted load data F(n|c) and the load influencing factor c; therefore, the loss function L of the load forecasting model is a binary minimax with conditional probability. value game.

The loss function L of the load forecasting model is a binary minimax game with conditional probability, which defines

minmaxL=E _r,c (lnD(r|c))+E _n,c (ln(1-D(F(n|c)|c))) (3)

Among them, min max L represents the binary minimax game value of the loss function L;

At the same time, the load prediction model uses the L1 norm as the loss function, so that the obtained prediction result is more accurate; the loss function of the L1 norm is:

L _L1 =E _r,c,n (||rF(n|c)||) (4)

Among them, L _L1 is the loss function of the L1 norm; E _r,c,n represents the expected value of the distribution of the real sample r, the condition value c and the random noise n;

In Figure 2, r is the real sample, which is the measured load data; c is the condition value, which is the historical load data and other influencing factors; n is the random noise; F(n|c) is the generated predicted load data; L is the The loss function is used as the feedback of the load forecasting model; the random noise n and the condition value c are input into the generator to generate the sample F(n|c), and then the generated sample F(n|c) and the real sample r are combined with the condition c respectively Input the discriminator to discriminate; the discriminator's discrimination result is fed back to the generator and discriminator in the form of a loss function, and the generator and discriminator modify their own parameters according to the feedback result to improve their respective generating and discriminating abilities; in order to further improve The accuracy of the short-term load forecasting of the load forecasting model is measured by using the hidden layer of the discriminator to measure the eigenvalue deviation between the forecast data generated by the generator and the real data; in this way, the network of the entire load forecasting model is continuously iteratively optimized to achieve the generation of The purpose of the prediction result of the device is more accurate.

In this embodiment, the block diagram structure of the load prediction model based on GAN network is shown in Figure 3, which includes three parts: data set construction, model training and prediction results; wherein, data set construction mainly refers to the use of historical load data and other influencing factors, Construct training and test data sets, and perform data preprocessing to determine the input and output variables of the load forecasting model; load influencing factors are conditional data, including historical load data, climate data, and date type data; measured load data is the target of the model data; in the process of load prediction model training, GAN combined with feature loss function is used for short-term load prediction; the probability of the discriminator outputting real data; the generator and discriminator optimize their own weights through continuous adversarial training, so that the entire load prediction model Finally, use the trained prediction model for load prediction, and compare the prediction results; when using the load prediction model constructed in this embodiment, input the contents of each group of test data sets into the generator in turn, and output each The daily prediction results are compared with the daily real values to obtain the prediction accuracy of the load prediction model based on the GAN network.

In actual training, when GAN faces data with a large amount of features, not only the network is not easy to train stably, but also there are many problems such as slow convergence speed, mode collapse, and the quality of generated samples needs to be improved; in order to solve the above problems, the introduction of volume CNN is used to construct the internal structure of the generator and the discriminator; CNN not only has the powerful ability of multi-hidden layer feature extraction, but also can share the convolution kernel, which has no pressure on high-dimensional data processing. The introduction of CNN can improve the performance of GAN. Stability, convergence speed and the quality of the data generated by the generator; the two-dimensional convolution model is conducive to model extraction, and the structure of the two-dimensional convolution model is used to build a prediction model, which can enhance the overall generalization ability and robustness of the model; Part of the hidden layer of the discriminator calculates the feature deviation to optimize the network structure, which can further improve the prediction accuracy.

In this embodiment, the generator includes a 3-layer convolutional neural network;

The 3-layer convolutional neural network is: input layer, convolutional layer C1 and convolutional layer C2;

The input layer is composed of 32 5×5 convolution kernels, the convolution layer C1 is composed of 64 5×5 convolution kernels, and the convolution layer C2 is composed of a 5×5 convolution kernel. The sliding step size of the neural network is both 2.

In order to allow the network to learn a more suitable spatial sampling method, the structure of the generator does not use spatial pooling in CNN, but adopts strided convolution, and adopts batch normalization operations between layers to accelerate convergence and slow down overfitting. Make the gradient propagation deeper; and use the ReLU activation function in the remaining layers except the output layer, and the output layer uses the tanh activation function, and finally generate the prediction data.

In this embodiment, the discriminator includes a 3-layer convolutional neural network, and its hidden layer uses the LeakyReLU function as an activation function, and the second hidden layer and the third hidden layer apply a feature loss function; and use fully connected and sigmoid The activation function performs true and false judgments, so that the result is mapped between (0, 1);

The 3-layer convolutional neural network is the first convolutional layer, the second convolutional layer and the third convolutional layer;

The first convolution layer consists of 32 5×5 convolution kernels, the second convolution layer consists of 64 5×5 convolution kernels, and the third convolution layer consists of 128 5×5 convolution kernels , and the sliding step size of each layer of convolutional neural network is 2.

In this embodiment, the specific formula for calculating the peak-shaving demand value of the gravity energy storage system in step 3 is:

P _D =P _peak -P _valley (5)

Among them, _PD represents the peak load demand value of the gravity energy storage system, P _peak represents the predicted load peak value of the day output by the load prediction model, and P _valley represents the predicted load valley value of the day output by the load prediction model.

In this embodiment, the specific formula for calculating the frequency regulation demand value of the gravity energy storage system in step 3 is:

P _F =ΔP _L -P _plan -P _line (6)

Among them, _{PF represents the frequency regulation demand value of the gravity energy storage system; ΔP L represents} the load change value of the gravity energy storage system; P _plan represents the frequency value of the frequency regulation unit power generation plan of the gravity energy storage system; P _line represents the connection line of the gravity energy storage system Adjust the planned frequency value.

In this embodiment, the constraints of the gravity energy storage system in step 4 are specifically:

P _G1 =αP _D ≤P _G , P _G1 ≥0 (7)

P _G2 =βP _F ≤P _G , P _G2 ≥0 (8)

Among them, _PG is the power capacity of the gravity energy storage system; _PG1 is the power capacity allocated by the gravity energy storage system that can participate in the peak regulation of the system; _PG2 is the power capacity allocated by the gravity energy storage system that can participate in the frequency regulation of the system; α is The value of the peak modulation participation factor; β is the value of the frequency modulation participation factor.

In this embodiment, the specific method for obtaining the peak shaving participation factor in step 5 is:

Step 51: Initialize the peak shaving demand value of the gravity energy storage system as the original population to generate the first generation population;

Step 52: Calculate the economic benefits of the gravity energy storage system on the primary population generated in the step 51, and determine the economic benefits of the gravity energy storage system for different individual particles;

Step 53: Select the position of the individual particle when the economic benefit of the gravity energy storage system is the largest, and obtain the optimal point and the global optimal point of the individual particle;

Step 54: Update the individual particle velocity, position, aggregation degree and weight;

Step 55: Determine whether the updated individual particle velocity, position, aggregation degree and weight meet the constraints of the gravity energy storage system; if the constraints are met, go to Steps 56; otherwise, return to Step 52;

Step 56: Output the updated position of the individual particle, and use the updated position of the individual particle as the peak regulation participation factor;

At the same time, the specific method for obtaining the FM participation factor in step 5 is as follows:

Step 1: Initialize the frequency regulation demand value of the gravity energy storage system as the original population to generate the first generation population;

Step II, calculating the economic benefits of the gravity energy storage system for the primary population generated in the step I, and determining the economic benefits of the gravity energy storage system for different individual particles;

Step III: Select the position of the individual particle when the economic benefit of the gravity energy storage system is the largest, and obtain the optimal point and the global optimal point of the individual particle;

Step IV, update individual particle velocity, position, aggregation degree and weight;

Step 5. Determine whether the updated individual particle velocity, position, aggregation degree and weight meet the constraints of the gravity energy storage system; if the constraints are met, go to Steps 5 and 6; otherwise, return to Step 5 and 2;

Step VI, output the updated position of the individual particle, and use the updated position of the individual particle as the frequency modulation participation factor.

In this embodiment, the expression of the power distribution model constructed in step 6 is:

P _G =P _G1 +P _G2 =αP _D +βP _F (9)

Among them, _PG is the power capacity of the gravity energy storage system; _PG1 is the power capacity allocated by the gravity energy storage system that can participate in the peak regulation of the system; _PG2 is the power capacity allocated by the gravity energy storage system that can participate in the frequency regulation of the system; α is The value of the peak modulation participation factor; β is the value of the frequency modulation participation factor.

In this embodiment, taking into account the demand for peak regulation and frequency regulation of the system, and making full use of the capacity of the gravity energy storage system, the peak regulation participation factor and frequency regulation participation factor of the gravity energy storage system are introduced. system power distribution;

The allocation scheme based on the optimal economic benefit is adopted, that is, on the basis of taking into account the cost and benefits of the gravity energy storage system participating in system peak regulation and frequency regulation, a power allocation model aiming at maximizing the economic benefit of the gravity energy storage system is constructed. To ensure the dispatch economy of the gravity energy storage system.

(1) Insufficient scheduling penalty cost:

Because the allocation of the peak shaving power and frequency regulation power capacity of the gravity energy storage system will lead to the phenomenon of insufficient peak shaving power and frequency regulation capacity of the system, a penalty factor is introduced.

C _lack = c _E E _lack +Σc _P P _lack (10)

Among them: c _E is the unit penalty cost of insufficient peak shaving power; c _P is the unit penalty cost of insufficient frequency regulation power; E _lack is the amount of insufficient power to participate in peak regulation according to the assigned gravity energy storage system; P _lack P _lack is the frequency regulation Missing power.

(2) Peak regulation income and frequency regulation income:

The frequency regulation revenue of the gravity energy storage system comes from providing paid frequency regulation services for the power system;

I _F =Σi ₁ P _G2 +i ₂ E _F (11)

Among them: i ₁ is the unit compensation income of frequency regulation power provided by the gravity energy storage system; i ₂ is the unit benefit of frequency regulation electricity; E _F is the frequency regulation electricity.

The peak shaving revenue of the gravity energy storage system comes from the peak and valley electricity price revenue and the peak shaving subsidy revenue.

I _D =i ₃ E _D +i ₄ ηE _D (12)

Among them: i ₃ is the unit electricity subsidy for peak regulation; i ₄ is the comprehensive price difference of the power grid; E _D is the peak regulation electricity; η is the energy conversion efficiency;

Therefore, the economic and technical model of the gravity energy storage system is:

I _total = _IF + _ID -C _lack (13)

Among them, I _total is the economic benefit of the gravity energy storage system.

Simulation:

Taking the power grid data of a certain area as an example, short-term load prediction is performed, and the power of the gravity energy storage system is economically distributed to verify the effectiveness of the power distribution method of this embodiment.

The results predicted by the GAN model are compared with the actual values as shown in Table 1.

Table 1 Comparison of load forecast results with actual values

Unit: MW

It can be seen from Table 1 that the load forecasting model described in this implementation, by enhancing the processing capability of high-dimensional data, can consider more influencing factors at the same time, and further improve the forecasting accuracy.

Based on this prediction result, several gravity energy storage systems in the region are allocated the optimal power economy in the simulation system, and compared with the conventional scheme's capacity-proportional allocation income, the results are shown in Figure 7; through comparative analysis It can be seen that the benefits obtained by using the power allocation method described in this embodiment are higher than those obtained by the conventional scheme, and with the increase of time, the income gap between the two gradually widens, which shows that the power allocation method described in this embodiment is in progress. Efficiency and economy of power distribution in gravity energy storage systems.

The above are only specific embodiments of the present invention, but the protection scope of the present invention is not limited thereto. Any person skilled in the art can easily think of changes or substitutions within the technical scope disclosed by the present invention. should be included within the protection scope of the present invention. Therefore, the protection scope of the present invention should be based on the protection scope of the claims.

Claims (10)

1. A gravity energy storage power distribution method based on a load prediction model is characterized by comprising the following steps:

step one, establishing a load prediction model based on a GAN network;

predicting the load of the gravity energy storage system by using the load prediction model established in the step one, and outputting a prediction result of the load of the gravity energy storage system;

thirdly, calculating a peak regulation demand value and a frequency modulation demand value of the gravity energy storage system based on the load prediction result of the gravity energy storage system output in the second step;

step four, calculating constraint conditions of the gravity energy storage system;

combining the constraint conditions of the gravity energy storage system calculated in the step four, and performing optimal solution on the peak regulation required value by adopting a particle swarm algorithm to obtain a value of a peak regulation participation factor; simultaneously, performing optimized solution on the frequency modulation required value by adopting a particle swarm algorithm to obtain a value of a frequency modulation participation factor;

Step six, adopting the peak regulation participation factors and the frequency modulation participation factors obtained in the step five to jointly construct a power distribution model with the maximum economic benefit of the gravity energy storage system as a target;

and step seven, performing power distribution on the gravity energy storage system by using the power distribution model constructed in the step six.

2. The gravity energy storage power distribution method based on the load prediction model as claimed in claim 1, wherein the load prediction model in the first step comprises a generator and a discriminator;

a generator for learning the distribution of the real samples r and generating new samples; the real sample r is load data of the gravity energy storage system obtained through actual measurement;

a discriminator for discriminating whether the input sample is from the generator.

3. The gravity energy storage power distribution method based on the load prediction model according to claim 2, wherein the specific process of establishing the load prediction model based on the GAN network in the first step is as follows: inputting the random noise n and the condition value c into a generator to generate a new sample F (n | c), and inputting the generated sample F (n | c) and the real sample r into a discriminator together with the condition c for discrimination; the discrimination result output by the discriminator is respectively fed back to the generator and the discriminator in the form of a loss function, and the generator and the discriminator revise the parameters thereof according to the feedback result; the condition value c is a load influence factor of the gravity energy storage system; the random noise n is a random variable conforming to Gaussian distribution;

The loss function of the generator in the load prediction model is defined as: l is a radical of an alcohol_FThe loss function of the discriminator in the load prediction model is defined as: l is a radical of an alcohol_D；

Loss function L of the generator_FIs of the form shown in equation (1):

L_F＝-E_n,c(D(F(n|c)|c)) (1)

wherein E is_n,cExpressing the expected values of the distribution of random noise n and the condition value c; fRepresenting the output data of the generator; d represents output data of the discriminator;

loss function L of the discriminator_DIs of the form shown in equation (2):

L_D＝-E_r,c(D(r|c))+E_n,c(D(F(n|c)|c)) (2)

wherein E is_r,cRepresenting the expected values of the distribution of the real sample r and the condition value c;

the loss function L of the load prediction model is a binary minimum maximum game containing conditional probability and is defined

minmaxL＝E_r,c(lnD(r|c))+E_n,c(ln(1-D(F(n|c)|c))) (3)

Wherein minmaxL represents a binary minimum maximum game value of the loss function L;

meanwhile, the load prediction model uses an L1 norm as a loss function; the loss function for the L1 norm is:

L_L1＝E_r,c,n(||r-F(n|c)||) (4)

wherein L is_L1A loss function that is the norm of L1; e_r,c,nRepresenting the expected values of the distribution for the true sample r, the condition value c, and the random noise n.

4. The gravity energy storage power distribution method based on the load prediction model according to claim 2, wherein the generator comprises a 3-layer convolutional neural network;

The 3 layers of convolutional neural networks are respectively as follows: an input layer, a convolutional layer C1, and a convolutional layer C2;

the input layer consists of 32 convolution kernels of 5 × 5, the convolutional layer C1 consists of 64 convolution kernels of 5 × 5, the convolutional layer C2 consists of 1 convolution kernel of 5 × 5, and the step size of sliding of each convolutional neural network is 2.

5. The gravity energy storage power distribution method based on the load prediction model according to claim 2, wherein the discriminator comprises a 3-layer convolutional neural network; the 3 layers of convolutional neural networks are respectively a first convolutional layer, a second convolutional layer and a third convolutional layer;

the first convolutional layer is composed of 32 convolution kernels of 5 × 5, the second convolutional layer is composed of 64 convolution kernels of 5 × 5, the third convolutional layer is composed of 128 convolution kernels of 5 × 5, and the step size of each convolutional neural network is 2.

6. The gravity energy storage power distribution method based on the load prediction model as claimed in claim 1, wherein the specific formula for calculating the peak shaver requirement value of the gravity energy storage system in the third step is as follows:

P_D＝P_peak-P_valley (5)

wherein, P_DRepresenting the peak shaver requirement value, P, of the gravity energy storage system_peakPredicted load peak of the day, P, representing the output of the load prediction model_valleyRepresenting the current day's predicted load trough output by the load prediction model.

7. The gravity energy storage power distribution method based on the load prediction model according to claim 1, wherein the specific formula for calculating the frequency modulation requirement value of the gravity energy storage system in the third step is as follows:

P_F＝ΔP_L-P_plan-P_line (6)

wherein, P_FRepresenting the frequency modulation requirement value of the gravity energy storage system; delta P_LRepresenting the load change value of the gravity energy storage system; p is_planRepresenting a power generation planning frequency value of a frequency modulation unit of the gravity energy storage system; p is_lineA tie-line adjustment planning frequency value representing the gravity energy storage system.

8. The gravity energy storage power distribution method based on the load prediction model according to claim 1, wherein the constraint conditions of the gravity energy storage system in the fourth step are specifically:

P_G1＝αP_D≤P_G，P_G1≥0 (7)

P_G2＝βP_F≤P_G，P_G2≥0 (8)

wherein, P_GIs the power capacity of the gravity energy storage system; p_G1The capacity of the power which can participate in system peak regulation and is distributed for the gravity energy storage system; p_G2The capacity of the power which can participate in system frequency modulation and is distributed for the gravity energy storage system; alpha is the value of the peak regulation participation factor; beta is the value of the frequency modulation participation factor.

9. The gravity energy storage power distribution method based on the load prediction model according to claim 1, wherein the specific method for obtaining the peak regulation participation factor in the step five is as follows:

fifthly, initializing the peak regulation demand value of the gravity energy storage system as an original population to generate an initial population;

Step two, calculating the economic benefits of the gravity energy storage system of the primary population generated in the step one to determine the economic benefits of the gravity energy storage system of different individual particles;

fifthly, selecting the position of the individual particle when the economic benefit of the gravity energy storage system is the maximum to obtain an individual particle optimal point and a global optimal point;

fifthly, updating the speed, the position, the aggregation degree and the weight of the individual particles;

fifthly, judging whether the updated individual particle speed, position, aggregation degree and weight meet the constraint conditions of the gravity energy storage system; if the constraint condition is met, executing a fifth step and a sixth step; otherwise, returning to execute the step two;

fifthly, outputting the positions of the updated individual particles, and taking the positions of the updated individual particles as peak regulation participation factors;

meanwhile, the concrete method for obtaining the frequency modulation participation factor in the step five is as follows:

step I, initializing a frequency modulation demand value of a gravity energy storage system as an original population to generate an initial generation population;

step II, calculating the economic benefit of the gravity energy storage system of the primary population generated in the step I, and determining the economic benefit of the gravity energy storage system of different individual particles;

Step III, selecting the position of the individual particle when the economic benefit of the gravity energy storage system is the maximum to obtain an individual particle optimal point and a global optimal point;

step IV, updating the speed, the position, the aggregation degree and the weight of the individual particles;

step V, judging whether the speed, the position, the aggregation degree and the weight of the updated individual particles meet the constraint conditions of the gravity energy storage system; if the constraint condition is met, executing a fifth step and a sixth step; otherwise, returning to execute the step two;

and VI, outputting the position of the updated individual particle, and taking the position of the updated individual particle as a frequency modulation participation factor.

10. The gravity energy storage power distribution method based on the load prediction model according to claim 1, wherein the expression of the power distribution model constructed in the sixth step is as follows:

P_G＝P_G1+P_G2＝αP_D+βP_F (9)

wherein, P_GIs the power capacity of the gravity energy storage system; p_G1The capacity of the power which can participate in system peak regulation and is distributed for the gravity energy storage system; p_G2The capacity of the power which can participate in system frequency modulation and is distributed for the gravity energy storage system; alpha is the value of the peak regulation participation factor; beta is the value of the frequency modulation participation factor.

Images