Gravity-based energy storage: grid charge-discharge balancing method, storage medium, and equipment.

Abstract

This invention discloses a grid charge-discharge balancing method, storage medium, and device based on gravity energy storage. The grid charge-discharge balancing method includes: collecting historical variables affecting the grid load in a certain region, calculating the weight ratio of each historical variable to the grid generation load, and retaining historical variables with a weight ratio greater than 0.8; inputting the retained historical variables into a grid load prediction model for training until the deviation between the predicted grid load and the actual grid load is less than a threshold, thus completing the training of the grid load prediction model; collecting input variables affecting the grid load in the region in real time, inputting them into the trained grid load prediction model, and predicting the grid load; and achieving grid charge-discharge balance by coordinating the predicted grid load with the charge-discharge mode of the gravity energy storage system, while reducing the response time of the gravity energy storage system.

Description

Technical Field

[0001] This invention relates to the field of gravity energy storage charging and discharging technology, specifically to a grid charging and discharging balancing method, storage medium, and equipment based on gravity energy storage. Background Technology

[0002] With the introduction of the "dual carbon" target, the scale of renewable energy has continued to expand, and new energy storage technology has developed rapidly in recent years. As a new technology, gravity energy storage offers greater site selection flexibility compared to conventional energy storage systems such as photovoltaic energy storage, wind energy storage, and chemical energy storage. It also features high rotational inertia and high energy conversion efficiency.

[0003] However, gravity energy storage systems have many moving parts and complex operating conditions. For example, tunnel energy storage and mine car sliding energy storage have limited means to adjust to changes in grid load. They cannot meet the requirements of grid load fluctuations for regulating the charging and discharging power of the energy storage system. In order to meet the hourly energy storage cycle and minute-level power generation preparation requirements of gravity energy storage systems, it is urgent to find a suitable grid charging and discharging balance method for gravity energy storage systems. Summary of the Invention

[0004] To address the problems existing in the prior art, this invention provides a grid charge-discharge balancing method, storage medium, and device based on gravity energy storage. By coordinating the predicted grid load with the charge-discharge mode of the gravity energy storage system, the grid charge-discharge balance is achieved, while reducing the response time of the gravity energy storage system.

[0005] To achieve the above objectives, the present invention adopts the following technical solution: a grid charging and discharging balancing method based on gravity energy storage, specifically including the following steps:

[0006] Step 1: Collect historical variables affecting the power grid load in a certain region, calculate the weight ratio of each historical variable to the power grid load using the coefficient calculation formula, and retain historical variables with a weight ratio greater than 0.8;

[0007] Step 2: Construct a power grid load prediction model. Input the retained historical variables into the power grid load prediction model for training until the deviation between the predicted power grid load and the actual power grid load is less than a threshold, thus completing the training of the power grid load prediction model.

[0008] Step 3: Collect the input variables affecting the power grid load in the region in real time, input them into the trained power grid load prediction model, and predict the power grid load.

[0009] Step 4: Based on the predicted power grid load and the charging and discharging mode of the gravity energy storage system, achieve a balance between power grid charging and discharging.

[0010] Furthermore, the historical variables collected that affect the power grid load in a certain region include: environmental variables, time variables, and local power grid operating parameters; the environmental variables include: real-time temperature, maximum temperature, minimum temperature, average temperature, average humidity, precipitation parameters, wind speed, wind volume, and sunshine index; the time variables include: date, hour, and minute; the local power grid operating parameters include: wind power installed capacity and photovoltaic installed capacity.

[0011] Furthermore, the formula for calculating the coefficient is as follows:

[0012]

[0013] in, For weighting percentage, The sampling interval is... The number of historical variables sampled. for index, For the first A historical variable, For the front The average of historical variables, For the first Power generation of the power grid under historical variables For the front The average power generation of the power grid under historical variables.

[0014] Furthermore, the power grid load prediction model is a recurrent neural network, consisting of an input layer, three hidden layers, and an output layer connected in sequence.

[0015] Furthermore, the construction process of the power grid load forecasting model is as follows:

[0016] (1) The retained historical variables are used to form an input vector and input into the input layer of the power grid load prediction model. The vector is then used to enter the hidden layer through the forward propagation algorithm to calculate the hidden state. Output variables after hidden layer :

[0017]

[0018]

[0019] in, t Indicates time, Indicates time as t The input vector at that time, Indicates time as t-The hidden state at time 1, where U represents the weight matrix from the input layer to the hidden layer, W represents the weight matrix of the hidden layer, b is the bias, and V represents the weight matrix from the hidden layer to the output layer. Indicates time as t The hidden state of the third layer, where c is the bias;

[0020] (2) Output variables after the hidden layer In the input / output layer, the power grid load is predicted. :

[0021] .

[0022] Furthermore, step 4 includes the following sub-steps:

[0023] Step 401: Calculate the deviation Δy between the predicted power grid load and the actual power generation demand;

[0024] Step 402: When Δy > 0, that is, the power grid load is insufficient, the power grid is charged by adjusting the number or speed of the mass blocks falling in the gravity energy storage system in the next cycle, so as to achieve the balance of power grid charging and discharging; otherwise, proceed to step 403.

[0025] Step 403: When Δy≤0, that is, the power grid has a surplus of electricity load, the surplus electricity in the power grid is consumed by adjusting the number or speed of mass blocks in the gravity energy storage system in the next cycle, so as to achieve the balance of power grid charging and discharging.

[0026] Furthermore, the specific process of step 402 is as follows: When Δy>0, that is, the power grid load is insufficient, the power generation capacity of the gravity energy storage system is calculated first. = mgV t * N Power generation of a single mass block P 2= mgV t And calculate the power generation deviation ΔP of the gravity energy storage system. When △P≥ P At time 2, in the next cycle, add one mass block with each fall until the maximum number of mass blocks is reached; when ΔP < P 2. If the falling mass has reached the maximum number of falls, calculate the change in the mass's falling velocity. V’= It then determines whether Δy is less than the power generation of the gravity energy storage system; if so, it reduces the falling speed of the mass block in the next cycle. V t+1 = V t - V’Otherwise, increase the falling speed of the mass block in the next cycle. V t+1 = V t + V’ ;

[0027] in, m The mass of the mass block. g Let gravitational acceleration be constant. V t Let be the current falling speed of the mass block. N This represents the current number of falling blocks.

[0028] Furthermore, the specific process of step 403 is as follows: When Δy≤0, that is, the power grid has a surplus of electricity load, the charging power of the gravity energy storage system is first calculated. = mgV t * N t and the charging power of a single mass block P 1= mgV t And calculate the charging power deviation ΔQ of the gravity energy storage system. When △Q≥ P When 1, add one mass block with each lift in the next cycle until the maximum number of mass blocks is reached; when ΔQ < P 1. If the maximum number of mass blocks to be lifted has been reached, calculate the change in the lifting speed of the mass blocks. V’= It also determines whether Δy is less than the charging power of the gravity energy storage system. If so, it reduces the lifting speed of the mass block in the next cycle. V t+1 = V t - V’ Otherwise, increase the mass block boosting speed in the next cycle. V t+1 = V t + V’ ;

[0029] in, m The mass of the mass block. g Let gravitational acceleration be constant. V t The current lift rate of the mass block. N t To increase the number of quality blocks currently.

[0030] Furthermore, the present invention also provides a computer-readable storage medium storing a computer program that causes a computer to execute the described gravity-based power grid charge-discharge balancing method.

[0031] Furthermore, the present invention also provides an electronic device, including: a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein when the processor executes the computer program, it implements the aforementioned grid charging and discharging balancing method based on gravity energy storage.

[0032] Compared with existing technologies, the present invention has the following beneficial effects: The present invention uses gravity energy storage as a minute-level energy storage system. Due to the time interval between the charging and discharging modes of the generator in the energy storage system, and considering the unstable fluctuations in the power load of the new energy grid, when gravity energy storage is used as a single distribution and storage system, it is necessary to adjust the mode of the energy storage system itself in advance to cope with changes in grid load. By adjusting the positive or negative deviation between the predicted grid load and the actual power generation demand, the charging and discharging mode of the gravity energy storage system is adjusted to achieve grid charging and discharging balance. At the same time, during the charging and discharging process of the gravity energy storage system, by first adjusting the number of moving mass blocks and then adjusting the speed of the moving mass blocks, when the load change value is less than the energy release capacity of a single mass block, the falling speed of the existing mass blocks is matched to adjust the energy release curve of the mass blocks, so as to better match the grid load and further optimize the grid charging and discharging balance process. Attached Figure Description

[0033] Figure 1 This is a flowchart of the grid charge-discharge balance method based on gravity energy storage according to the present invention;

[0034] Figure 2 This is a schematic diagram illustrating how to achieve grid charge-discharge balance by coordinating the charging and discharging methods of a gravity energy storage system with the predicted grid load. Detailed Implementation

[0035] The technical solution of the present invention will be further explained and described below with reference to the accompanying drawings.

[0036] like Figure 1 This is a flowchart of the grid charge-discharge balancing method based on gravity energy storage according to the present invention. The grid charge-discharge balancing method includes the following steps:

[0037] Step 1: Collect historical variables affecting the power grid load in a certain region. The historical variables collected in this invention include: environmental variables, time variables, and local power grid operating parameters. Environmental variables include: real-time temperature, maximum temperature, minimum temperature, average temperature, average humidity, precipitation parameters, wind speed, wind volume, and sunshine index; time variables include: date, hour, and minute; local power grid operating parameters include: wind power installed capacity and photovoltaic installed capacity. The weight ratio of each historical variable to the power grid load is calculated using the coefficient calculation formula.

[0038] The formula for calculating the coefficients in this invention is as follows:

[0039]

[0040] in, For weighting percentage, The sampling interval is... The number of historical variables sampled. for index, For the first A historical variable, For the front The average of historical variables, For the first Power generation of the power grid under historical variables For the front The average power generation of the power grid under historical variables.

[0041] The weighting percentage of each historical variable in relation to the power grid load in this invention The numerical range of is between [0, 1]. The value ∈ [0, 0.2) indicates that the correlation between this historical variable and the power grid load is extremely weak; The range ∈ [0.2, 0.4) indicates that the correlation between this historical variable and the power grid load is relatively weak; The range ∈ [0.4, 0.6) indicates that the correlation between this historical variable and the power grid load is moderate; The range ∈ [0.6, 0.8) indicates a strong correlation between this historical variable and the power grid load; The value ∈ [0.8, 1] indicates that the historical variable has a very strong correlation with the power grid load. In order to establish a more accurate power grid load prediction model, this invention retains historical variables with a weight ratio greater than 0.8 as inputs to the power grid load prediction model.

[0042] Step 2: Construct a power grid load prediction model. Input the retained historical variables into the power grid load prediction model for training until the deviation between the predicted power grid load and the actual power grid load is less than a threshold, thus completing the training of the power grid load prediction model. In this invention, the power grid load prediction model is a recurrent neural network, consisting of an input layer, three hidden layers, and an output layer connected sequentially. Compared with traditional backpropagation neural networks, recurrent neural networks have timeliness in propagation and are suitable for the time-series requirements of power grid load. At the same time, to avoid the gradient explosion or gradient vanishing problems in the hidden layers of recurrent neural networks, the recurrent neural network of this invention uses three hidden layers.

[0043] The construction process of the power grid load forecasting model in this invention is as follows:

[0044] (1) The retained historical variables are used to form an input vector and input into the input layer of the power grid load prediction model. The vector is then used to enter the hidden layer through the forward propagation algorithm to calculate the hidden state. Output variables after hidden layer :

[0045]

[0046]

[0047] in, t Indicates time, Indicates time as t The input vector at that time, Indicates time as t- The hidden state at time 1, where U represents the weight matrix from the input layer to the hidden layer, W represents the weight matrix of the hidden layer, b is the bias, and V represents the weight matrix from the hidden layer to the output layer. Indicates time as t The hidden state of the third layer, where c is the bias;

[0048] (2) Output variables after the hidden layer In the input / output layer, the power grid load is predicted. :

[0049] .

[0050] Step 3: Collect the input variables affecting the power grid load in the region in real time, input them into the trained power grid load prediction model, and predict the power grid load.

[0051] Step 4: When the grid load is constrained, but remains constant or continues to increase, the gravity energy storage system discharges to balance the grid load. Specifically, the charging and discharging method of the gravity energy storage system is coordinated with the predicted grid load to achieve grid charging and discharging balance. By predicting the power generation load, the future power generation load of the new energy grid can be judged, and the internal dynamic mechanism equipment of the gravity energy storage system can be adjusted in a timely manner to avoid the phenomenon of wind and electricity curtailment in the new energy system. Figure 2 Specifically, it includes the following sub-steps:

[0052] Step 401: Calculate the deviation Δy between the predicted power grid load and the actual power generation demand. If Δy > 0, it means that the actual power generation of the power grid is small and the output is insufficient. In this case, the gravity energy storage system needs to increase its power generation to eliminate Δy. Conversely, if Δy ≤ 0, it means that the actual power generation of the power grid is sufficient. In this case, the power generation of the gravity energy storage system should be reduced.

[0053] Step 402: When Δy > 0, indicating insufficient grid load, the grid is charged by adjusting the number or speed of falling mass blocks in the gravity energy storage system for the next cycle, thus achieving grid load balance. When the gravity energy storage system generates electricity, if Δy is greater than the power generated by a single falling mass block, the number of mass blocks is preferentially increased or decreased. If Δy < the power generated by a single mass block, the falling speed of existing mass blocks is adjusted to match grid load fluctuations. Otherwise, proceed to step 403. The specific process is as follows:

[0054] When Δy > 0, meaning the power grid load is insufficient, the power generation capacity of the gravity energy storage system should be calculated first. = mgV t * N Power generation of a single mass block P 2= mgV t And calculate the power generation deviation ΔP of the gravity energy storage system. When △P≥ P At time 2, in the next cycle, one more mass block is dropped each time until the maximum number of mass blocks dropped is reached. The drop of a single mass block represents the minimum power generation within the gravity energy storage system. Adjusting the number of individual mass blocks in each cycle can avoid the impact on the power grid caused by large fluctuations in the power of the energy storage system itself; when ΔP < P 2. If the falling mass has reached its maximum number of falls, the power P=FV can also be transformed into P=mg*V in the vertical motion relationship. When the mass of the mass cannot be changed, the power can be changed by adjusting the falling speed of the mass. Calculate the change in the falling speed V'= It then determines whether Δy is less than the power generation of the gravity energy storage system; if so, it reduces the falling speed of the mass block in the next cycle. Vt+1 = V t - V’ Otherwise, increase the falling speed of the mass block in the next cycle. V t+1 = V t + V’ ;in, m The mass of the mass block. g Let gravitational acceleration be constant. V t Let be the current falling speed of the mass block. N This represents the number of mass blocks currently falling.

[0055] Step 403: The power grid commonly experiences wind and electricity curtailment, with excess power generation going unused. When Δy≤0, indicating a power grid load surplus, the gravity energy storage system switches to charging mode, acting as a load on the grid to absorb the surplus power. The surplus power in the grid is consumed by adjusting the number or speed of mass blocks lifted in the gravity energy storage system in the next cycle, thus achieving grid load balance. The specific process is as follows: When Δy≥0, indicating a power grid load surplus, the charging power of the gravity energy storage system is first calculated. = mgV t * N t and the charging power of a single mass block P 1= mgV t And calculate the charging power deviation ΔQ of the gravity energy storage system. When △Q≥ P When 1, add one more mass block with each lift in the next cycle until the maximum number of mass blocks is reached; when ΔQ < P 1. If the maximum number of mass blocks to be lifted has been reached, calculate the change in the lifting speed of the mass blocks. V’= It also determines whether Δy is less than the charging power of the gravity energy storage system. If so, it reduces the lifting speed of the mass block in the next cycle. V t+1 = V t - V’ Otherwise, increase the mass block boosting speed in the next cycle. V t+1 = V t + V’ ;

[0056] in, m The mass of the mass block.g Let gravitational acceleration be constant. V t The current lift rate of the mass block. N t To increase the number of quality blocks currently.

[0057] This invention provides a grid charge-discharge balance method based on gravity energy storage. By adjusting the charge-discharge mode of the gravity energy storage system according to the positive or negative deviation between the predicted grid load and the actual power generation demand, the grid charge-discharge balance is achieved. At the same time, during the charge-discharge process of the gravity energy storage system, the grid charge-discharge balance process is further optimized by prioritizing the adjustment of the number of moving mass blocks and then adjusting the speed of the moving mass blocks.

[0058] In the embodiments disclosed in this application, a computer storage medium may be a tangible medium that may contain or store programs for use by or in conjunction with an instruction execution system, apparatus, or device. The computer storage medium may include, but is not limited to, electronic, magnetic, optical, electromagnetic, infrared, or semiconductor systems, apparatus, or devices, or any suitable combination of the foregoing. More specific examples of computer storage media include electrical connections based on one or more wires, portable computer disks, hard disks, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM or flash memory), optical fibers, and portable compact disc read-only memory (CD). ROM), optical storage devices, magnetic storage devices, or any suitable combination of the foregoing.

[0059] Those skilled in the art will recognize that the units and algorithm steps of the various examples described in conjunction with the embodiments disclosed in this application can be implemented in electronic hardware or a combination of computer software and electronic hardware. Whether these functions are implemented in hardware or software depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementation should not be considered beyond the scope of this application.

[0060] The above are merely preferred embodiments of the present invention. The scope of protection of the present invention is not limited to the above embodiments. All technical solutions falling within the scope of the present invention's concept are within the scope of protection of the present invention. It should be noted that for those skilled in the art, any improvements and modifications made without departing from the principles of the present invention should be considered within the scope of protection of the present invention.

Claims

1. A grid charge-discharge balancing method based on gravity energy storage, characterized in that, Specifically, the steps include the following: Step 1: Collect historical variables affecting the power grid load in a certain region, calculate the weight ratio of each historical variable to the power grid load using the coefficient calculation formula, and retain historical variables with a weight ratio greater than 0.8; Step 2: Construct a power grid load prediction model. Input the retained historical variables into the power grid load prediction model for training until the deviation between the predicted power grid load and the actual power grid load is less than a threshold, thus completing the training of the power grid load prediction model. Step 3: Collect the input variables affecting the power grid load in the region in real time, input them into the trained power grid load prediction model, and predict the power grid load. Step 4: Based on the predicted grid load and the charging / discharging mode of the gravity energy storage system, achieve grid charging / discharging balance; this includes the following sub-steps: Step 401: Calculate the deviation Δy between the predicted power grid load and the actual power generation demand; Step 402: When Δy > 0, that is, the power grid load is insufficient, the power grid is charged by adjusting the number or speed of the mass blocks falling in the gravity energy storage system in the next cycle, so as to achieve the balance of power grid charging and discharging; otherwise, proceed to step 403. Step 403: When Δy≤0, that is, the power grid has a surplus of electricity load, the surplus electricity in the power grid is consumed by adjusting the number or speed of mass blocks in the gravity energy storage system in the next cycle, so as to achieve the balance of power grid charging and discharging.

2. The grid charge-discharge balancing method based on gravity energy storage according to claim 1, characterized in that, The historical variables collected that affect the power grid load in a certain region include: environmental variables, time variables, and local power grid operating parameters; the environmental variables include: real-time temperature, maximum temperature, minimum temperature, average temperature, average humidity, precipitation parameters, wind speed, wind volume, and sunshine index; the time variables include: date, hour, and minute; the local power grid operating parameters include: wind power installed capacity and photovoltaic installed capacity.

3. The grid charge-discharge balancing method based on gravity energy storage according to claim 1, characterized in that, The formula for calculating the coefficient is: in, For weighting percentage, The sampling interval is... The number of historical variables sampled. for index, For the first A historical variable, For the front The average of historical variables, For the first Power generation of the power grid under historical variables For the front The average power generation of the power grid under historical variables.

4. The grid charge-discharge balancing method based on gravity energy storage according to claim 1, characterized in that, The power grid load prediction model is a recurrent neural network, consisting of an input layer, three hidden layers, and an output layer connected in sequence.

5. A grid charge-discharge balancing method based on gravity energy storage according to claim 1, characterized in that, The process of constructing the power grid load forecasting model is as follows: (1) The retained historical variables are used to form an input vector and input into the input layer of the power grid load prediction model. The vector is then used to enter the hidden layer through the forward propagation algorithm to calculate the hidden state. Output variables after hidden layer : in, t Indicates time, Indicates time as t The input vector at that time, Indicates time as t- The hidden state at time 1, where U represents the weight matrix from the input layer to the hidden layer, W represents the weight matrix of the hidden layer, b is the bias, and V represents the weight matrix from the hidden layer to the output layer. Indicates time as t The hidden state of the third layer, where c is the bias; (2) Output variables after the hidden layer In the input / output layer, the power grid load is predicted. : 。 6. The grid charge-discharge balancing method based on gravity energy storage according to claim 1, characterized in that, The specific process of step 402 is as follows: When Δy>0, that is, the power grid load is insufficient, the power generation capacity of the gravity energy storage system is calculated first. = mgV t * N Power generation of a single mass block P 2= mgV t And calculate the power generation deviation ΔP of the gravity energy storage system. When △P≥ P At time 2, in the next cycle, add one mass block with each fall until the maximum number of mass blocks is reached; when ΔP < P 2. If the falling mass has reached the maximum number of falls, calculate the change in the mass's falling velocity. V’= It then determines whether Δy is less than the power generation of the gravity energy storage system; if so, it reduces the falling speed of the mass block in the next cycle. V t+1 = V t - V’ Otherwise, increase the falling speed of the mass block in the next cycle. V t+1 = V t + V’ ; in, m For the mass of the mass block, g Let gravitational acceleration be constant. V t Let be the current falling speed of the mass block. N This represents the current number of falling blocks.

7. A grid charge-discharge balancing method based on gravity energy storage according to claim 1, characterized in that, The specific process of step 403 is as follows: When Δy≤0, that is, the power grid has a surplus of electricity load, the charging power of the gravity energy storage system is calculated first. = mgV t * N t and the charging power of a single mass block P 1= mgV t And calculate the charging power deviation ΔQ of the gravity energy storage system. When △Q≥ P When 1, add one mass block with each lift in the next cycle until the maximum number of mass blocks is reached; when ΔQ < P 1. If the maximum number of mass blocks to be lifted has been reached, calculate the change in the lifting speed of the mass blocks. V’= It also determines whether Δy is less than the charging power of the gravity energy storage system. If so, it reduces the lifting speed of the mass block in the next cycle. V t+1 = V t - V’ Otherwise, increase the mass block boosting speed in the next cycle. V t+1 = V t + V’ ; in, m For the mass of the mass block, g Let gravitational acceleration be constant. V t The current lift rate of the mass block. N t To increase the number of quality blocks currently.

8. A computer-readable storage medium storing a computer program, characterized in that, The computer program causes the computer to execute the grid charge-discharge balancing method based on gravity energy storage as described in any one of claims 1-7.

9. An electronic device, characterized in that, include: The memory, the processor, and the computer program stored in the memory and executable on the processor, wherein when the processor executes the computer program, it implements the grid charge-discharge balancing method based on gravity energy storage as described in any one of claims 1-7.

Images