Operation optimization system and method for gravity energy storage and new energy grid-connected power generation

Through the neural network control system combined with deep neural network and XGBoost model, the operation of gravity energy storage system is optimized, and the problem of low power generation efficiency of coupled solar energy and gravity energy storage is solved, and the grid power demand and solar power generation system are achieved efficiently matched, thus improving the efficiency of solar power generation system. The stability and absorption capacity of the system.

CN119010107APending Publication Date: 2024-11-22XIAN THERMAL POWER RES INST CO LTD
View PDF 0 Cites 2 Cited by

Patent Information

Application Number
CN202411015267.8
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2024-07-26
Publication Date
2024-11-22

AI Technical Summary

Technical Problem

In the prior art, the randomness and uncertainty of solar power generation and grid loads lead to low operating efficiency of coupled power generation between solar and gravity energy storage, especially under the influence of weather and sunshine time, it is difficult to achieve effective load peak shaving and energy storage modes. optimization.

Method used

Adopting a neural network control system, combining deep neural network and XGBoost model, predicts the power demand for power grid and the power generated by solar power generation systems, optimizes the operation of gravity energy storage system through energy storage mode and energy release mode to achieve efficient storage and release of electricity , adjust the operating status of the energy storage system to match the solar power generation power and grid load.

Benefits of technology

Improve the operation efficiency of coupled power generation with solar energy and gravity energy storage, reduce the impact on weather and sunshine time, more accurately control the matching of power grid electricity demand and solar power generation system, and improve the system's absorption capacity and stability .

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN119010107A_ABST
    Figure CN119010107A_ABST
Patent Text Reader

Abstract

The invention relates to the technical field of gravity energy storage power generation, in particular to an operation optimization system and method for gravity energy storage and new energy grid-connected power generation. The operation optimization system for gravity energy storage and new energy grid-connected power generation comprises a power grid, a solar power generation system, a gravity energy storage system and a neural network control system, the solar power generation system is respectively connected to the power grid and the gravity energy storage system and is used for generating and outputting electric energy; the gravity energy storage system is connected with the solar power generation system and the power grid, and is used for receiving and storing the electric energy and releasing and outputting the electric energy; the neural network control system is connected to the gravity energy storage system and used for controlling operation of the gravity energy storage system; the power grid is connected with the gravity energy storage system and the solar power generation system and used for receiving electric energy. According to the system, the operation of the gravity energy storage system is well controlled through the neural network control system, and the problem that the operation efficiency of solar energy and gravity energy storage coupling power generation is low is solved.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present invention relates to the technical field of gravity energy storage power generation, and in particular to an operation optimization system and method for gravity energy storage and new energy grid-connected power generation. Background Art

[0002] my country is accelerating the construction of a new power system, but as the penetration rate of wind power and photovoltaic power generation continues to increase, the stability of the power system is becoming increasingly prominent. Energy storage technology provides an effective solution to maintain the overall stability of the system and achieve a high proportion of renewable energy consumption.

[0003] my country is actively exploring new high-quality, large-scale energy storage technologies. Among them, gravity energy storage technology based on discrete mass blocks has received strong support due to its simple working principle and high energy storage efficiency. A paper published in Science Weekly in 2021 stated that gravity energy storage can be used as an effective alternative to electrochemical energy storage technology. The working principle of gravity energy storage is similar to that of pumped storage. The energy storage medium is replaced by a discrete solid medium from the "water" in pumped storage. It is mainly benchmarked against energy-type energy storage. It is a long-term, large-capacity energy storage that can transfer electricity day and night in the power system and improve the overall stability of the system.

[0004] Energy storage systems are key factors in the rapid development and effective use of renewable energy, and they can smooth out the unstable characteristics of renewable energy. Therefore, energy storage systems are efficient buffers between the power supply side, the grid side, and the user side, and will play an increasingly important role in the future trend of renewable energy development. Among them, the advantages of gravity energy storage systems include: first, they have no terrain / geological dependence, are safe and environmentally friendly, and have no explosion hazards, and can use various terrain conditions to build large-capacity energy storage systems; second, the role of gravity energy storage in the power grid is not only to replace existing energy storage solutions, especially in the three northern regions of my country where water resources are scarce but wind and solar resources are abundant, it also provides a better and more convenient solution for increasing and expanding the capacity of existing energy storage systems.

[0005] At present, few people have optimized the load forecast, energy release mode and energy storage mode of gravity energy storage system under the coupled power generation of gravity energy storage system and solar power generation unit, especially when the solar power generation is greatly affected by factors such as weather and sunshine time, the solar power generation is also seriously affected. When the sunshine is strong and the wind is strong, solar energy and wind energy will bring huge load impact to the power grid. In this case, the thermal power units participating in peak load regulation and frequency regulation are all operating in deep regulation conditions or even shut down; when the weather conditions are poor or the peak power consumption period is in the evening, solar energy will no longer have any effect, and the units participating in peak load regulation will reach a higher output at this time to meet the power demand. Since the load of the power grid fluctuates at any time and is affected by peak load regulation and frequency regulation, the difficulty of optimizing the operation of the solar energy and gravity energy storage coupled power generation system is increased, making the operation efficiency of the solar energy and gravity energy storage coupled power generation low. Summary of the invention

[0006] The purpose of the present invention is to provide an operation optimization system for gravity energy storage and new energy grid-connected power generation, so as to solve the problem of low operating efficiency of solar energy and gravity energy storage coupled power generation caused by the randomness and uncertainty in the selection of typical scenarios for solar power generation and grid load in the prior art.

[0007] In order to solve the above problems, the technical solution of the operation optimization system of gravity energy storage and new energy grid-connected power generation of the present invention is:

[0008] An operation optimization system for gravity energy storage and new energy grid-connected power generation, comprising: a power grid, a solar power generation system, a gravity energy storage system and a neural network control system;

[0009] The solar power generation system is connected to the power grid and the gravity energy storage system respectively for generating and outputting electric energy;

[0010] The gravity energy storage system is connected to the solar power generation system and to the power grid, and is used to receive and store electrical energy and release and output electrical energy;

[0011] The neural network control system is connected to the gravity energy storage system and is used to control the operation of the gravity energy storage system;

[0012] The power grid is connected to the gravity energy storage system and the solar power generation system and is used to receive electrical energy.

[0013] Optionally, the neural network control system includes:

[0014] The input layer module is used to input interference factors that affect gravity energy storage and renewable energy grid-connected power generation, and transmit them to the hidden layer module;

[0015] The hidden layer module is used to analyze interference factors and output the analysis results to the output layer module;

[0016] The output layer module is used to receive the analysis results and control the operation of the gravity energy storage system according to the analysis results.

[0017] Optionally, the hidden layer module includes:

[0018] A neural network control model, wherein the neural network control model includes a deep neural network model, and the deep neural network model is composed of neural network models of multiple hidden layers in a hidden layer module, and is used to predict power demand of a power grid and power generated by a solar power generation system;

[0019] A neural network prediction model is used to characterize the forecast information of power grid demand and power generated by solar power generation system.

[0020] Optionally, the neural network control model further includes:

[0021] The XGBoost model is used to analyze the impact of interference factors on the operation of the gravity energy storage system.

[0022] Optionally, the operating modes of the gravity energy storage system include an energy storage mode and an energy release mode. The energy storage mode is applied when the power demand of the power grid is lower than the power generated by the solar power generation system, and the gravity energy storage system stores the power; the energy release mode is applied when the power demand of the power grid is higher than the power generated by the solar power generation system, and the gravity energy storage system releases the power and outputs it to the power grid.

[0023] Optionally, the energy storage mode includes a gravity energy storage system operation maintaining module and a gravity energy storage system operation adjusting module:

[0024] The gravity energy storage system operation maintenance module is used to ensure that the gravity energy storage system continues to operate normally at rated power when the solar power generation power is greater than or equal to the charging power of the gravity energy storage system;

[0025] The gravity energy storage system operation adjustment module is used to adjust the system operation state when the solar power generation power is less than the rated charging power of the gravity energy storage system and the gravity energy storage system cannot continue to operate normally at the rated power.

[0026] Optionally, the gravity energy storage system operation adjustment module includes:

[0027] The grid compensation module is used for the grid to provide the charging power shortage of the gravity energy storage system, so that the solar power generation power matches the charging power of the gravity energy storage system;

[0028] The energy release module of the gravity energy storage system is used for releasing energy of the gravity energy storage system to reduce the charging power of the gravity energy storage system, so that the solar power generation power matches the charging power of the gravity energy storage system.

[0029] In order to solve the above problems, the technical solution of a method for optimizing the operation of gravity energy storage and new energy grid-connected power generation proposed in the present invention is:

[0030] A method for optimizing the operation of gravity energy storage and new energy grid-connected power generation, comprising the following steps:

[0031] S1, generating electricity through a solar power generation system;

[0032] S2. The neural network control system predicts the power demand of the power grid and the power generated by the solar power generation system, compares the two, and then controls the operation of the gravity energy storage system:

[0033] If the power demand of the grid is equal to the power generated by the solar power generation system, go to S3;

[0034] If the power demand of the power grid is less than the power generated by the solar power generation system, then go to S4;

[0035] If the power demand of the power grid is greater than the power generated by the solar power generation system, then go to S5;

[0036] S3, the gravity energy storage system does not operate, the solar power generation system outputs electric energy to the grid, and the grid receives electric energy;

[0037] S4, the solar power generation system outputs electric energy to the power grid, and the power grid receives the electric energy; at the same time, the solar power generation system outputs excess electric energy to the gravity energy storage system to achieve electric energy storage;

[0038] S5. The solar power generation system outputs electric energy to the power grid, and at the same time the gravity energy storage system releases electric energy and outputs it to the power grid, which receives the electric energy.

[0039] Optionally, in S2, the neural network control system includes an input layer module, a hidden layer module and an output layer module, the hidden layer module includes a neural network control model and a neural network prediction model, the neural network control model includes a deep neural network model, and the prediction of power grid electricity demand and electric energy generated by the solar power generation system by the neural network control system specifically includes:

[0040] Through the deep neural network model, the interference factors that affect the gravity energy storage and new energy grid-connected power generation input by the input layer module are analyzed, the forecast information of the power demand of the power grid and the electric energy generated by the solar power generation system is obtained, and a neural network prediction model is established.

[0041] Optionally, the neural network control model further includes an XGBoost model, and the predicting of power demand of the power grid and electric energy generated by the solar power generation system by the neural network control system further includes:

[0042] Based on the prediction of the deep neural network model and the prediction of the XGBoost model, a DNN-XGBoost model combination model based on the inverse error method is established, and the prediction results of the two models are weighted combined to obtain the prediction value of the combination model. The mathematical expression of the inverse error method is as follows:

[0043] Y D_X,j =w D y Dj +w X y Xj

[0044]

[0045] Among them, Y D_X,j is the predicted value of the combined model, w D With w X are the weight coefficients of the prediction values ​​of the deep neural network model and the XGBoost model, ε D With ε X are the errors of the prediction values ​​of the deep neural network model and the XGBoost model, respectively. Dj With y Xj They are the predicted values ​​of the deep neural network model and the XGBoost model respectively.

[0046] Compared with the prior art, the present application provides an operation optimization system and method for gravity energy storage and new energy grid-connected power generation, which is mainly based on a deep neural network algorithm to predict and optimize the relevant information of coupled solar energy and gravity energy storage coupled power generation. The operation optimization system for gravity energy storage and new energy grid-connected power generation includes a power grid, a solar power generation system, a gravity energy storage system and a neural network control system, wherein the operation of the gravity energy storage system is controlled by the neural network control system. When the power demand of the power grid is lower than the power generated by the solar power generation system, the power supply capacity of new energy is sufficient, the solar power generation system outputs power to the power grid, and the neural network control system controls the operation of the gravity energy storage system. The solar power generation system outputs excess power to the gravity energy storage system. The gravity energy storage system drags the mass block from the stacking platform with low potential energy to the stacking platform with high potential energy, and stores the power as the gravitational potential energy of the mass block to achieve power storage; when the power demand of the power grid is higher than the power generated by the solar power generation system, the power supply capacity of new energy is insufficient, the solar power generation system outputs power to the power grid, and the neural network control system controls the operation of the gravity energy storage system, that is, the peak-shaving operation of the gravity energy storage system, the mass block of the gravity energy storage system will run from the high potential energy position to the low potential energy position, and the generator will be dragged by the transmission device to feed back electricity to the power grid. The above-mentioned gravity energy storage and new energy grid-connected power generation operation optimization system and method can be used for the randomness and uncertainty of the selection of typical scenarios of solar power generation and power grid load, and the operation of the gravity energy storage system can be well controlled by the neural network control system, thereby improving the problem of low operation efficiency of solar energy and gravity energy storage coupled power generation.

[0047] The interference factors that affect gravity energy storage and new energy grid-connected power generation are input into the input layer module in the neural network control system, and analyzed through the hidden layer module. The analysis results are output through the output layer module and the operation of the gravity energy storage system is controlled according to the analysis results, which facilitates more accurate control of the operation of the gravity energy storage system.

[0048] The hidden layer module includes a neural network control model and a neural network prediction model. The deep neural network model in the neural network control model is used to predict the power demand of the power grid and the electricity generated by the solar power generation system, and the prediction information is used to construct a neural network prediction model, which optimizes the operation system of gravity energy storage and new energy grid-connected power generation, reduces its influence by weather and sunshine time, more accurately judges the size of the power demand of the power grid and the electricity generated by the solar power generation system, and more accurately controls the operation of the gravity energy storage system.

[0049] By weighted combination of the deep neural network model and the XGBoost model, the prediction accuracy of the neural network control model can be further improved, and the importance of each factor in the neural network control model on the optimization of energy storage system operation can be analyzed, which is conducive to the power grid proposing different improvement measures to enhance the system's absorption capacity based on the importance of each factor.

[0050] The inverse error method is used to perform a weighted combination of the prediction results of the two models to obtain the final prediction value. The inverse error method will give the model with a smaller error a larger weight coefficient, so that the error of the entire combined model tends to decrease, and a prediction value with a smaller error is obtained, thereby achieving the effect of improving the overall prediction accuracy.

[0051] The operation modes of the gravity energy storage system include energy storage mode and energy release mode. By controlling the operation of energy storage mode and energy release mode, the operation optimization system of gravity energy storage and new energy grid-connected power generation can be more accurately applied when the power demand of the power grid is lower or higher than the electricity generated by the solar power generation system.

[0052] The energy storage mode includes a gravity energy storage system operation maintenance module and a gravity energy storage system operation adjustment module. The energy storage mode is adjusted by comparing the solar power generation power and the gravity energy storage system charging power to fully ensure the operation of the energy storage mode in the gravity energy storage system.

[0053] The operation adjustment module of the gravity energy storage system includes a power grid compensation module and a gravity energy storage system energy release module. By considering the new energy electricity price, the power grid benchmark electricity price and the charging efficiency of the gravity energy storage system, a more accurate adjustment module is selected to better match the new energy power generation power with the gravity energy storage system charging power. BRIEF DESCRIPTION OF THE DRAWINGS

[0054] Figure 1 It is a structural schematic diagram of an operation optimization system for gravity energy storage and new energy grid-connected power generation according to an embodiment of the present application;

[0055] Figure 2 It is a structural schematic diagram of a neural network control model in an operation optimization system of gravity energy storage and new energy grid-connected power generation according to an embodiment of the present application;

[0056] Figure 3 A flow chart of a combined model of a deep neural network model and an XGBoost model in an operation optimization system for gravity energy storage and new energy grid-connected power generation according to an embodiment of the present application;

[0057] Figure 4 The present invention is a flow chart of energy flow path selection of the gravity energy storage system operation adjustment module in the energy storage mode in the operation optimization system of gravity energy storage and new energy grid-connected power generation according to an embodiment of the present application. DETAILED DESCRIPTION

[0058] As cited in the background technology, the randomness and uncertainty of the selection of typical scenarios of solar power generation and grid load in the prior art lead to low operating efficiency of solar energy and gravity energy storage coupled power generation. Therefore, the present invention provides an operation optimization system for gravity energy storage and new energy grid-connected power generation, including a power grid, a solar power generation system, a gravity energy storage system and a neural network control system; the solar power generation system is used for the generation and output of electric energy, and is respectively connected to the power grid and the gravity energy storage system to form two operation routes of the operation optimization system for gravity energy storage and new energy grid-connected power generation; one operation route is that the solar power generation system is connected to the power grid to directly supply power to the power grid; the other operation route is that the solar power generation system is connected to the gravity energy storage system, which is used to realize that the gravity energy storage system stores the electric energy generated by the solar power generation system; at the same time, the gravity energy storage system continues to be connected to the power grid, which is used to realize that the gravity energy storage system releases electric energy and supplies power to the power grid. The neural network control system is separately connected to the gravity energy storage system and is used to accurately control the operation of the gravity energy storage system, so that the gravity energy storage system can receive and store electrical energy and release and output electrical energy, thereby improving the low operating efficiency of the solar energy and gravity energy storage coupled power generation and solving the problem of the difficulty in optimizing the operation of the solar energy and gravity energy storage coupled power generation system.

[0059] Embodiments of the present application are described in detail below, and examples of the embodiments are shown in the accompanying drawings, wherein the same or similar reference numerals throughout represent the same or similar elements or elements having the same or similar functions. The embodiments described below with reference to the accompanying drawings are exemplary and are intended to be used to explain the present application, and should not be construed as limiting the present application.

[0060] The following describes the operation optimization system and method of gravity energy storage and new energy grid-connected power generation in an embodiment of the present application with reference to the accompanying drawings.

[0061] Combine the following Figure 1 , the operation optimization system of gravity energy storage and new energy grid-connected power generation provided in this application is described in detail. Figure 1 It is a structural schematic diagram of the operation optimization system of gravity energy storage and new energy grid-connected power generation according to an embodiment of the present application.

[0062] In this embodiment, Figure 1 As shown, the operation optimization system of gravity energy storage and new energy grid-connected power generation includes a power grid, a solar power generation system, a gravity energy storage system and a neural network control system. The operation optimization system of gravity energy storage and new energy grid-connected power generation forms two operation paths, one is that the solar power generation system is connected to the power grid through a DC / AC inverter; the other is that the solar power generation system is connected to the gravity energy storage system through a DC / AC inverter, and the gravity energy storage system is connected to the power grid through a transformer. Among them, the operation modes of the gravity energy storage system include energy storage mode and energy release mode.

[0063] When the solar power generation system receives sunlight, the solar power generation system generates electricity and passes through the DC / AC inverter. Under the action of the inverter, the direct current is converted into alternating current, and the alternating current is then transmitted to the power grid to provide electrical energy for the power grid. When the power demand of the power grid is less than the power generated by the solar power generation system, the solar power generation system generates excess power. At this time, the neural network control system controls the energy storage mode of the gravity energy storage system to start running. The solar power generation system transmits the excess power to the gravity energy storage system after passing through the DC / AC inverter. The gravity energy storage system drags the mass block from the stacking platform with low potential energy to the stacking platform with high potential energy, and stores the power as the gravitational potential energy of the mass block. The gravity energy storage system realizes energy storage; when the power demand of the power grid is greater than the power generated by the solar power generation system, the power generated by the solar power generation system is insufficient to match the power demand of the power grid. At this time, the neural network control system controls the energy release mode of the gravity energy storage system to start running. The gravity energy storage system operates in peak-shaving mode. The mass block of the gravity energy storage system will run from the high potential energy position to the low potential energy position, and the power will be transmitted to the power grid through the transformer. The gravity energy storage system realizes energy release.

[0064] Among them, the neural network or artificial neural network (ANN) in the neural network control system is a type of machine learning algorithm that can decode the implicit relationship in a set of data by imitating the process of how the brain works. Artificial neural networks have superior characteristics that traditional algorithms do not have, such as autonomous learning, knowledge reasoning, and adaptive characteristics, so they are regarded as one of the most promising methods in the field of deep learning. Neural network or artificial neural network (ANN) algorithm can decode the implicit relationship in a set of data by imitating the process of how the brain works.

[0065] Specifically, Figure 2 As shown, the neural network control system includes: an input layer module, which includes an input layer, for inputting interference factors affecting gravity energy storage and new energy grid-connected power generation, and transmitting them to the hidden layer module, wherein the interference factors mainly include: sunshine time, sunshine intensity, electric energy of solar power generation system, power grid power demand, gravity energy storage system, etc.; a hidden layer module, which includes a hidden layer, for receiving and analyzing the above interference factors, and sending the analysis results to the output layer module. Among them, the hidden layer in the hidden layer module can be only one layer, but when processing multi-dimensional nonlinear data, one hidden layer cannot effectively display the nonlinear relationship between data, so in this embodiment, preferably, there are multiple hidden layers in the hidden layer module; an output layer module, which includes an output layer, for receiving the analysis results, and controlling the operation of the gravity energy storage system according to the analysis results. Among them, the input layer and the hidden layer are composed of multiple neurons (solid circles), each neuron contains multiple weights and an activation function. When the input interference factor data is transmitted to the neuron, it is weighted and summed, and then the summation result is transmitted to the activation function to generate an output.

[0066] In a specific embodiment of the present application, the hidden layer module includes: a neural network control model, the neural network control model includes a deep neural network model (DNN), the deep neural network model is composed of multiple hidden layer neural network models in the hidden layer module, and is used to predict the power demand of the power grid and the electric energy generated by the solar power generation system, thereby realizing the prediction of the energy storage mode and energy release mode of the gravity energy storage system, and realizing accurate control of the operation of the gravity energy storage system.

[0067] Specifically: a deep neural network model is used to predict solar power generation and grid power load demand, training set and test set data samples of the neural network model with multiple hidden layers are selected, the number of neurons in each hidden layer and the number of hidden layers are determined, the optimal loss function and optimization algorithm are selected, and a neural network prediction model is established; the neural network prediction model is used to characterize the prediction information of grid power demand and the electric energy generated by the solar power generation system, so as to more accurately control the operation of the gravity energy storage system through the prediction information.

[0068] In a possible implementation of the embodiment of the present application, the neural network control model also includes: an XGBoost model for analyzing the influence of interference factors on the operation of the gravity energy storage system. The XGBoost model and the deep neural network model are used in combination. Figure 3 This is a flow chart of the combined model of the deep neural network model and the XGBoost model. As the hidden layer in the neural network control model, the deep neural network model and the XGBoost model are applied to the analysis of the power demand of the power grid, the gravity energy storage system and the solar power generation forecast and operation optimization. The deep neural network model focuses on the accuracy of the neural network control model, while the XGBoost model focuses more on the interpretability of the neural network control model. Through the XGBoost model, the influence of various input interference factors on the operation of the gravity energy storage system can be analyzed concisely and effectively. Therefore, by combining the deep neural network model with the XGBoost model, the accuracy of the model prediction can be further improved, and the importance of each factor in the model on the optimization of the operation of the energy storage system can be analyzed, which is conducive to the power grid to propose different improvement measures to improve the system's absorption capacity based on the importance of each factor.

[0069] In a specific embodiment of the present application, first, based on the prediction of the power grid demand and solar power generation based on the deep neural network model and the prediction of the influence of various interference factors on the operation of the gravity energy storage system based on the XGBoost model, the corresponding prediction values ​​are obtained, and the two prediction values ​​are trained and output; secondly, the prediction values ​​of the two models are weighted and combined using the inverse error method to obtain a DNN-XGBoost combination model based on the inverse error method, and the final prediction value is obtained and output by weighting the DNN model weight and the XGBoost model weight. Among them, the mathematical expression of the inverse error method is as follows:

[0070] Y D_X,j =w D y Dj +w X y Xj

[0071]

[0072] Among them, Y D_X,j is the predicted value of the combined model, w D With w X are the weight coefficients of the prediction values ​​of the deep neural network model and the XGBoost model, ε D With ε X are the errors of the prediction values ​​of the deep neural network model and the XGBoost model, respectively. Dj With y Xj They are the predicted values ​​of the deep neural network model and the XGBoost model respectively.

[0073] In a specific embodiment of the present application, Figure 4As shown, the energy storage mode in the gravity energy storage system includes a gravity energy storage system operation maintenance module and a gravity energy storage system operation adjustment module, that is, when the gravity energy storage system is in charging condition, the operation state of the gravity energy storage system in the energy storage mode is adjusted by monitoring the power generated by the new energy. For the new energy power generation system, when the power generated by the new energy is greater than or equal to the charging power of the gravity energy storage system, the gravity energy storage system can still continue to operate normally at the rated power, and the price of new energy electricity is relatively low, and the additional power will not be included in the gravity energy storage system. At this time, there is no need to adjust the number of mass blocks, that is, there is no need to adjust the operating state of the gravity energy storage system. When the power generated by renewable energy is less than the rated charging power of the gravity energy storage system, the actual charging power of the gravity energy storage system is reduced. The power generated by renewable energy cannot drag the number of mass blocks corresponding to the rated power of the gravity energy storage system to a high potential energy position, and cannot drive the rated number of mass blocks to operate. At this time, it is necessary to formulate the most appropriate energy flow path selection based on parameters such as benchmark electricity price, average renewable energy electricity price, corresponding power of a single mass block, power shortage of the new energy system, power fluctuation time of the new energy system, and time required for the first-order mass block to enter the stack to match the renewable energy power generation power and the gravity energy storage charging power. The first-order mass block here is the mass block closest to the upper stack field.

[0074] The above solution is applied on the new energy side as the source end. When there is power fluctuation on the source end, there is no need to adjust the output of the new energy side (photovoltaic, etc.) during the entire process. Instead, the energy flow path of the gravity energy storage system is selected to better match the new energy generation power and the gravity energy storage system charging power.

[0075] Specifically, see Figure 4 , the energy flow path selection process of the gravity energy storage system operation adjustment module of the gravity energy storage system in the energy storage mode of the embodiment of the present invention. There are two main energy flow paths under charging conditions. The first is to directly rely on the power grid to compensate for the shortage of new energy power fluctuations, and the second is to deal with new energy power fluctuations by unloading mass blocks. The specific analysis is as follows:

[0076] 1) Energy flow path 1: When the power generated by renewable energy is less than the charging power of the gravity energy storage system, the power generation can no longer meet the rated charging power of the ramp-type gravity energy storage system. At this time, the power grid can be used to continuously provide the charging power shortage of the gravity energy storage system, so that the rated number of mass blocks can stably operate to the high potential energy stack, thereby matching the power generated by renewable energy with the charging power of the gravity energy storage system. Among them, the power shortage coefficient M has a threshold value M0, and the energy flow path is selected by the threshold value, that is, when M is less than or equal to the threshold value M0, the energy flow path 1 is selected or maintained.

[0077] 2) Energy flow path 2: The same as the energy flow path 1 scenario, but this path first uses the power grid to provide this part of the missing charging power, and a certain number of mass blocks arrive at the high potential energy stack, and then adjusts the number of mass blocks running on the slope, that is, unloading the mass blocks. At this time, the charging power of the slope gravity energy storage system will be slightly less than the power generation power on the new energy side, completing the power matching between the two. Among them, when M is greater than the threshold M0, energy flow path 2 is selected.

[0078] Due to the discrete characteristics of the mass block of the gravity energy storage system, it can generally only be adjusted based on a single mass block. In order to give the judgment limit of the above two energy flow path selections, the ratio of the new energy charging power shortage to the corresponding power of a single mass block is defined as the power shortage coefficient, which is expressed as:

[0079]

[0080] Where: ΔP is the power deviation, P0 is the power design value, and M is the power shortage coefficient.

[0081] Using the above-mentioned gravity energy storage and new energy grid-connected power generation operation optimization system, the present invention also provides a gravity energy storage and new energy grid-connected power generation operation optimization method. The gravity energy storage and new energy grid-connected power generation operation optimization method comprises the following steps: first, generating electric energy through a solar power generation system; second, predicting the power demand of the power grid and the electric energy generated by the solar power generation system through a neural network control system, comparing the two, and then controlling the operation of the gravity energy storage system: if the power demand of the power grid is equal to the electric energy generated by the solar power generation system, the gravity energy storage system does not operate, the solar power generation system outputs electric energy to the power grid, and the power grid receives electric energy; if the power demand of the power grid is less than the electric energy generated by the solar power generation system, the solar power generation system outputs electric energy to the power grid, and the power grid receives electric energy; at the same time, the solar power generation system outputs excess electric energy to the gravity energy storage system to achieve electric energy storage; if the power demand of the power grid is greater than the electric energy generated by the solar power generation system, the solar power generation system outputs electric energy to the power grid, and at the same time, the gravity energy storage system releases electric energy and outputs it to the power grid, and the power grid receives electric energy.

[0082] Here, those skilled in the art can understand that the specific operations of each step of the above-mentioned method for optimizing the operation of gravity energy storage and renewable energy grid-connected power generation have been described in the above reference. Figures 1 to 4 The operation optimization method system of gravity energy storage and new energy grid-connected power generation has been introduced in detail, so its repeated description will be omitted.

[0083] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. The patent protection scope of the present invention shall be based on the claims. All equivalent structural changes made using the contents of the description and drawings of the present invention should also be included in the protection scope of the present invention.

Claims

1. An operation optimization system for gravity energy storage and new energy grid-connected power generation, characterized in that: include: Power grids, solar power generation systems, gravity energy storage systems, and neural network control systems; The solar power generation system is connected to the power grid and the gravity energy storage system respectively for generating and outputting electric energy; The gravity energy storage system is connected to the solar power generation system and to the power grid, and is used to receive and store electrical energy and release and output electrical energy; The neural network control system is connected to the gravity energy storage system and is used to control the operation of the gravity energy storage system; The power grid is connected to the gravity energy storage system and the solar power generation system and is used to receive electrical energy.

2. The operation optimization system of gravity energy storage and new energy grid-connected power generation according to claim 1 is characterized in that: The neural network control system comprises: The input layer module is used to input interference factors that affect gravity energy storage and renewable energy grid-connected power generation, and transmit them to the hidden layer module; The hidden layer module is used to analyze interference factors and output the analysis results to the output layer module; The output layer module is used to receive the analysis results and control the operation of the gravity energy storage system according to the analysis results.

3. The operation optimization system of gravity energy storage and new energy grid-connected power generation according to claim 2 is characterized in that: The hidden layer module comprises: A neural network control model, wherein the neural network control model includes a deep neural network model, and the deep neural network model is composed of neural network models of multiple hidden layers in a hidden layer module, and is used to predict power demand of a power grid and power generated by a solar power generation system; A neural network prediction model is used to characterize the forecast information of power grid demand and power generated by solar power generation system.

4. The operation optimization system for gravity energy storage and new energy grid-connected power generation according to claim 3 is characterized in that: The neural network control model also includes: The XGBoost model is used to analyze the impact of interference factors on the operation of the gravity energy storage system.

5. The operation optimization system of gravity energy storage and new energy grid-connected power generation according to claim 1 is characterized in that: The operation modes of the gravity energy storage system include a storage mode and a release mode. The storage mode is applied when the power demand of the power grid is lower than the power generated by the solar power generation system, and the gravity energy storage system stores the power; the release mode is applied when the power demand of the power grid is higher than the power generated by the solar power generation system, and the gravity energy storage system releases the power and outputs it to the power grid.

6. The operation optimization system for gravity energy storage and new energy grid-connected power generation according to claim 5 is characterized in that: The energy storage mode includes a gravity energy storage system operation maintenance module and a gravity energy storage system operation adjustment module: The gravity energy storage system operation maintenance module is used to ensure that the gravity energy storage system continues to operate normally at rated power when the solar power generation power is greater than or equal to the charging power of the gravity energy storage system; The gravity energy storage system operation adjustment module is used to adjust the system operation state when the solar power generation power is less than the rated charging power of the gravity energy storage system and the gravity energy storage system cannot continue to operate normally at the rated power.

7. The operation optimization system for gravity energy storage and new energy grid-connected power generation according to claim 6 is characterized in that: The gravity energy storage system operation adjustment module includes: The grid compensation module is used for the grid to provide the charging power shortage of the gravity energy storage system, so that the solar power generation power matches the charging power of the gravity energy storage system; The energy release module of the gravity energy storage system is used for releasing energy of the gravity energy storage system to reduce the charging power of the gravity energy storage system, so that the solar power generation power matches the charging power of the gravity energy storage system.

8. A method for optimizing the operation of gravity energy storage and new energy grid-connected power generation, characterized in that: The operation optimization system of gravity energy storage and new energy grid-connected power generation according to any one of claims 1 to 7 comprises the following steps: S1, generating electricity through a solar power generation system; S2. The neural network control system predicts the power demand of the power grid and the power generated by the solar power generation system, compares the two, and then controls the operation of the gravity energy storage system: If the power demand of the grid is equal to the power generated by the solar power generation system, go to S3; If the power demand of the power grid is less than the power generated by the solar power generation system, then go to S4; If the power demand of the power grid is greater than the power generated by the solar power generation system, then go to S5; S3, the gravity energy storage system does not operate, the solar power generation system outputs electric energy to the grid, and the grid receives electric energy; S4, the solar power generation system outputs electric energy to the power grid, and the power grid receives the electric energy; at the same time, the solar power generation system outputs excess electric energy to the gravity energy storage system to achieve electric energy storage; S5. The solar power generation system outputs electric energy to the power grid, and at the same time the gravity energy storage system releases electric energy and outputs it to the power grid, which receives the electric energy.

9. The method for optimizing the operation of gravity energy storage and new energy grid-connected power generation according to claim 8, characterized in that: In S2, the neural network control system includes an input layer module, a hidden layer module and an output layer module, the hidden layer module includes a neural network control model and a neural network prediction model, the neural network control model includes a deep neural network model, and the prediction of power grid power demand and electric energy generated by the solar power generation system by the neural network control system specifically includes: Through the deep neural network model, the interference factors that affect the gravity energy storage and new energy grid-connected power generation input by the input layer module are analyzed, the forecast information of the power demand of the power grid and the electric energy generated by the solar power generation system is obtained, and a neural network prediction model is established.

10. The method for optimizing the operation of gravity energy storage and new energy grid-connected power generation according to claim 9, characterized in that: The neural network control model also includes an XGBoost model, and the prediction of power grid power demand and power generated by the solar power generation system by the neural network control system also includes: Based on the prediction of the deep neural network model and the prediction of the XGBoost model, a DNN-XGBoost model combination model based on the inverse error method is established, and the prediction results of the two models are weighted combined to obtain the prediction value of the combined model. The mathematical expression of the inverse error method is as follows: Y D_X,j =w D y Dj +w X y Xj Among them, Y D_X,j is the predicted value of the combined model, w D With w X are the weight coefficients of the prediction values ​​of the deep neural network model and the XGBoost model, ε D With ε X are the errors of the prediction values ​​of the deep neural network model and the XGBoost model, respectively. Dj With y Xj They are the predicted values ​​of the deep neural network model and the XGBoost model respectively.

Citation Information

Cited By

  • Solar power generation and gravity energy storage combined power supply system and power supply method

    CN121308039A

  • A power supply system and method combining solar power generation and gravity energy storage

    CN121308039B