TCN-LSTM-based energy storage end maximum power supply capability prediction method, system and device

Abstract

The invention provides a TCN-LSTM-based energy storage end maximum power supply capability prediction method, system and equipment, and the method comprises the steps: carrying out the load power prediction of each load node through employing a TCN-LSTM model based on the historical data of each load node of an energy storage end, superposing the prediction results of each load node, obtaining a total load power prediction value, and carrying out the prediction of the maximum power supply capability of the energy storage end. And finally, determining the maximum power supply capacity of the energy storage end by adopting an iterative power flow method. According to the method, the load growth rule can be accurately simulated, and the real actual load condition can be reflected, so that the maximum power supply capacity of the energy storage end can be dynamically predicted, and power resources can be better distributed.

Description

Technical Field

[0001] The present invention belongs to the field of energy storage terminal power distribution planning, and specifically relates to a method, system and device for predicting the maximum power supply capacity of an energy storage terminal based on TCN-LSTM. Background Art

[0002] The energy storage terminal is the link between the transmission system and the power users. It receives the power from the transmission network and provides power to the power users. With the increase of social power load, the construction of the energy storage terminal lags behind the actual load demand. This may cause power shortages for power users during peak hours, thus affecting economic development and normal life. The power supply capacity reflects the automation level and safety level of the energy storage terminal. Therefore, exploring the power supply capacity of the energy storage terminal is of great significance to the planning, construction and safe operation of the energy storage terminal. Mobile energy storage refers to the technology of using movable energy storage equipment to store energy and release it when needed. It has the characteristics of flexibility, portability and dispatchability, and is widely used in emergency power supply, outdoor power supply, electric vehicles, etc. Predicting the power supply capacity of the mobile energy storage terminal is of great significance to solving the problem of imbalance between supply and demand, improving power supply reliability, and reducing power supply costs. However, the current prediction method of the power supply capacity of the energy storage terminal often cannot accurately simulate the load growth law, so it is impossible to make an accurate dynamic prediction of the power supply capacity of the energy storage terminal. Summary of the invention

[0003] The purpose of the present invention is to provide a method, system and device for predicting the maximum power supply capacity of an energy storage terminal based on TCN-LSTM in order to solve the above problems in the prior art.

[0004] To achieve the above objectives, the technical solution of the present invention is as follows:

[0005] In a first aspect, the present invention proposes a method for predicting the maximum power supply capacity of an energy storage terminal based on TCN-LSTM, comprising:

[0006] S1. Based on the historical data of each load node at the energy storage end, the TCN-LSTM model is used to predict the load power of each load node;

[0007] S2. Superimpose the prediction results of each load node to obtain the total load power prediction value;

[0008] S3. Use the iterative power flow method to determine the maximum power supply capacity of the energy storage end.

[0009] The S3 includes:

[0010] S31, based on the total load power forecast value P L(t) Perform power flow calculation to find the critical point where the voltage or power exceeds the limit, record the time t0 when the voltage or power exceeds the limit for the first time, and set a = t0-1, b = t0;

[0011] S32, determine whether ba<ε is satisfied, if so, then set P at this time L (a) as the maximum power supply capacity of the energy storage end; if it is not satisfied, then enter S33, where ε is the set threshold;

[0012] S33, calculate the midpoint of the interval [a,b] And according to Perform power flow calculation and determine whether the system voltage or power exceeds the limit. If not, set Then return to S32 loop; if crossing the line occurs, Then return to S32 loop.

[0013] The S2 includes:

[0014] S21, input the historical load power time series data of each load node at the energy storage end into the TCN layer as the feature variable matrix X, and extract the corresponding features through TCN;

[0015] S22. Input the output of the TCN layer into the LSTM layer to obtain the load power prediction results of each load node.

[0016] In S2, the total load forecast value is calculated according to the following formula:

[0017]

[0018] In the above formula, P L (t) is the total load power forecast value at time t, P Li (t) is the load power forecast value of the i-th load node at time t, N bus is the number of load nodes.

[0019] In the second aspect, the present invention proposes a maximum power supply capacity prediction system for an energy storage terminal based on TCN-LSTM, including a load node power prediction module, a total load power calculation module, and a maximum power supply capacity solution module;

[0020] The load node power prediction module is used to predict the load power of each load node using the TCN-LSTM model based on the historical data of each load node at the energy storage end;

[0021] The total load power calculation module is used to superimpose the prediction results of each load node to obtain a total load power prediction value;

[0022] The maximum power supply capacity solving module is used to determine the maximum power supply capacity of the energy storage terminal by adopting an iterative power flow method.

[0023] The specific steps of the iterative power flow method include:

[0024] A1. According to the total load power forecast value P L (t) Perform power flow calculation to find the critical point where the voltage or power exceeds the limit, record the time t0 when the voltage or power exceeds the limit for the first time, and set a = t0-1, b = t0;

[0025] A2. Determine whether ba<ε. If so, set P L (a) as the maximum power supply capacity of the energy storage end; if it is not satisfied, then go to A3, where ε is the set threshold;

[0026] A3. Take the midpoint of the interval [a,b] and calculate And according to Perform power flow calculation and determine whether the system voltage or power exceeds the limit. If not, set Then return to A2 loop; if crossing the line occurs, Then return to A2 loop.

[0027] The load node power prediction module includes a TCN layer and an LSTM layer;

[0028] The TCN layer is used to extract the characteristics of the historical load power time series data of each load node at the energy storage end;

[0029] The LSTM layer is used to calculate the load power prediction results of each load node based on the features extracted by the TCN layer.

[0030] The total load forecast value is calculated according to the following formula:

[0031]

[0032] In the above formula, P L (t) is the total load power forecast value at time t, P Li (t) is the load power forecast value of the i-th load node at time t, N bus is the number of load nodes.

[0033] In a third aspect, the present invention proposes a TCN-LSTM-based energy storage terminal maximum power supply capacity prediction device, including a memory and a processor;

[0034] The memory is used to store computer program code and transmit the computer program code to the processor;

[0035] The processor is used to execute the aforementioned method according to the instructions in the computer program code.

[0036] In a fourth aspect, the present invention provides a computer-readable storage medium having a computer program stored thereon, wherein the computer program implements the aforementioned method when executed by a processor.

[0037] Compared with the prior art, the present invention has the following beneficial effects:

[0038] The present invention discloses a method for predicting the maximum power supply capacity of an energy storage terminal based on TCN-LSTM. The method first predicts the load power of each load node based on the historical data of each load node at the energy storage terminal using the TCN-LSTM model, and then superimposes the prediction results of each load node to obtain a total load power prediction value, and finally uses an iterative power flow method to determine the maximum power supply capacity of the energy storage terminal. The method combines the TCN-LSTM model with the iterative power flow method to determine the maximum power supply capacity prediction of the energy storage terminal, and can accurately simulate the load growth law and reflect the actual load situation, thereby dynamically predicting the power supply capacity of the energy storage terminal to better allocate power resources. BRIEF DESCRIPTION OF THE DRAWINGS

[0039] Figure 1 This is a flow chart of the method described in Example 1.

[0040] Figure 2 This is a flowchart of the TCN-LSTM algorithm in Example 1.

[0041] Figure 3 This is a structural block diagram of the system described in Example 2.

[0042] Figure 4 This is a structural block diagram of the device described in Example 3. DETAILED DESCRIPTION

[0043] The present invention is further described in detail below in conjunction with the accompanying drawings and specific implementation methods.

[0044] The present invention provides a method for predicting the maximum power supply capacity of an energy storage terminal based on TCN-LSTM. The method adopts the TCN algorithm to expand the causal convolution to perform calculations according to the historical input sequence, which can ensure the temporal sequence and expand the receptive field, thereby greatly improving the ability to process data. LSTM is used to capture the temporal relationship of data by characterizing the characteristics of the forgetting stage, the selective memory stage and the output stage of each node in the transmission network, thereby realizing long-term memory of important features and forgetting of unimportant features.

[0045] Embodiment 1:

[0046] A method for predicting the maximum power supply capacity of energy storage terminal based on TCN-LSTM, such as Figure 1 As shown, the specific steps are as follows:

[0047] 1. Collect the historical load power sample data of each load node at the energy storage end to form time series data with an hourly time interval. Use it as the input data sample to construct the characteristic variable matrix X:

[0048]

[0049] In the above formula, x jk is the load power of the jth load node in time period k, J and K are the number of load nodes and the number of time periods respectively.

[0050] 2. Input the feature variable matrix X into the TCN layer and extract the corresponding features through TCN, such as Figure 2 shown.

[0051] 3. Input the flat layer output by TCN into the LSTM layer containing four LSTM units to obtain the load power prediction results of each load node.

[0052] 4. Superimpose the prediction results of each load node according to the following formula to obtain the total load power prediction value:

[0053]

[0054] In the above formula, P L (t) is the total load power forecast value at time t, P Li (t) is the load power forecast value of the i-th load node at time t, N bus is the number of load nodes.

[0055] 5. Call the total load power forecast value P L (t) Carry out power flow calculation to find the critical point of voltage or power exceeding the limit of the system, record the time t0 when the voltage or power exceeds the limit for the first time, then the maximum load P that the system can provide m That is, located at P L (t0-1) and P L (t0).

[0056] 6. Let a=t0-1, b=t0, then the interval [t0-1,t0] can be expressed as [a,b].

[0057] 7. Determine whether ba<ε is satisfied. If so, set P at this time L (a) is the maximum power supply capacity of the energy storage end; if it is not satisfied, go to step 8, where ε is the set threshold reflecting the minimum time scale.

[0058] 8. Take the midpoint of the interval [a, b] and calculate the total load power forecast value at that time And according to Perform power flow calculation and determine whether the system voltage or power exceeds the limit. If not, it means P m lie in and P L (b) between, therefore, let Then return to step 7 for the next cycle; if crossing the line occurs, it means P m Located in P L (a) and Therefore, let Then return to step 7 for the next cycle.

[0059] Embodiment 2:

[0060] A maximum power supply capacity prediction system for energy storage based on TCN-LSTM, such as Figure 3 As shown, it includes a load node power prediction module, a total load power calculation module, and a maximum power supply capacity solution module.

[0061] The load node power prediction module is used to predict the load power of each load node based on the historical data of each load node at the energy storage end, using the TCN-LSTM model, and includes a TCN layer and an LSTM layer;

[0062] The TCN layer is used to extract the characteristics of the historical load power time series data of each load node at the energy storage end;

[0063] The LSTM layer is used to calculate the load power prediction results of each load node based on the features extracted by the TCN layer.

[0064] The total load power calculation module is used to superimpose the prediction results of each load node according to the following formula to obtain the total load power prediction value:

[0065]

[0066] In the above formula, P L (t) is the total load power forecast value at time t, P Li (t) is the load power forecast value of the i-th load node at time t, N bus is the number of load nodes.

[0067] The maximum power supply capacity solving module is used to determine the maximum power supply capacity of the energy storage terminal by using an iterative power flow method. The specific steps of the iterative power flow method include:

[0068] A1. According to the total load power forecast value P L(t) Perform power flow calculation to find the critical point where the voltage or power exceeds the limit, record the time t0 when the voltage or power exceeds the limit for the first time, and set a = t0-1, b = t0;

[0069] A2. Determine whether ba<ε. If so, set P L (a) as the maximum power supply capacity of the energy storage end; if it is not satisfied, then go to A3, where ε is the set threshold;

[0070] A3. Take the midpoint of the interval [a,b] and calculate And according to Perform power flow calculation and determine whether the system voltage or power exceeds the limit. If not, set Then return to A2 loop; if crossing the line occurs, Then return to A2 loop.

[0071] Embodiment 3:

[0072] A device for predicting the maximum power supply capacity of energy storage end based on TCN-LSTM, such as Figure 4 As shown, including a memory and a processor;

[0073] The memory is used to store computer program code and transmit the computer program code to the processor;

[0074] The processor is used to execute the method described in embodiment 1 according to the instructions in the computer program code.

[0075] Embodiment 4:

[0076] A computer-readable storage medium having a computer program stored thereon, wherein the computer program implements the method described in Embodiment 1 when executed by a processor.

Claims

1. A method for predicting the maximum power supply capacity of an energy storage terminal based on TCN-LSTM, characterized in that: The method comprises: S1. Based on the historical data of each load node at the energy storage end, the TCN-LSTM model is used to predict the load power of each load node; S2. Superimpose the prediction results of each load node to obtain the total load power prediction value; S3. Use the iterative power flow method to determine the maximum power supply capacity of the energy storage end.

2. According to claim 1, a method for predicting the maximum power supply capacity of an energy storage terminal based on TCN-LSTM is characterized in that: The S3 includes: S31, based on the total load power forecast value P L (t) Perform power flow calculation to find the critical point where the voltage or power exceeds the limit, record the time t0 when the voltage or power exceeds the limit for the first time, and set a = t0-1, b = t0; S32, determine whether ba<ε is satisfied, if so, then set P at this time L (a) as the maximum power supply capacity of the energy storage end; if it is not satisfied, then enter S33, where ε is the set threshold; S33, calculate the midpoint of the interval [a,b] And according to Perform power flow calculation and determine whether the system voltage or power exceeds the limit. If not, set Then return to S32 loop; if crossing the line occurs, Then return to S32 loop.

3. A method for predicting the maximum power supply capacity of an energy storage terminal based on TCN-LSTM according to claim 1 or 2, characterized in that: The S2 includes: S21, input the historical load power time series data of each load node at the energy storage end into the TCN layer as the feature variable matrix X, and extract the corresponding features through TCN; S22. Input the output of the TCN layer into the LSTM layer to obtain the load power prediction results of each load node.

4. A method for predicting the maximum power supply capacity of an energy storage terminal based on TCN-LSTM according to claim 1 or 2, characterized in that: In S2, the total load forecast value is calculated according to the following formula: In the above formula, P L (t) is the total load power forecast value at time t, P Li (t) is the load power forecast value of the i-th load node at time t, N bus is the number of load nodes.

5. A TCN-LSTM-based energy storage terminal maximum power supply capacity prediction system, characterized in that: The system includes a load node power prediction module, a total load power calculation module, and a maximum power supply capacity solution module; The load node power prediction module is used to predict the load power of each load node using the TCN-LSTM model based on the historical data of each load node at the energy storage end; The total load power calculation module is used to superimpose the prediction results of each load node to obtain a total load power prediction value; The maximum power supply capacity solving module is used to determine the maximum power supply capacity of the energy storage terminal by adopting an iterative power flow method.

6. According to claim 5, a TCN-LSTM-based energy storage terminal maximum power supply capacity prediction system is characterized in that: The specific steps of the iterative power flow method include: A1. Based on the total load power forecast value P L (t) Perform power flow calculation to find the critical point where the voltage or power exceeds the limit, record the time t0 when the voltage or power exceeds the limit for the first time, and set a = t0-1, b = t0; A2. Determine whether ba<ε. If so, set P L (a) as the maximum power supply capacity of the energy storage end; if it is not satisfied, then go to A3, where ε is the set threshold; A3. Take the midpoint of the interval [a,b] and calculate And according to Perform power flow calculation and determine whether the system voltage or power exceeds the limit. If not, set Then return to A2 loop; if crossing the line occurs, Then return to A2 loop.

7. A TCN-LSTM-based energy storage terminal maximum power supply capacity prediction system according to claim 5 or 6, characterized in that: The load node power prediction module includes a TCN layer and an LSTM layer; The TCN layer is used to extract the characteristics of the historical load power time series data of each load node at the energy storage end; The LSTM layer is used to calculate the load power prediction results of each load node based on the features extracted by the TCN layer.

8. A TCN-LSTM-based energy storage terminal maximum power supply capacity prediction system according to claim 5 or 6, characterized in that: The total load forecast value is calculated according to the following formula: In the above formula, P L (t) is the total load power forecast value at time t, P Li (t) is the load power forecast value of the i-th load node at time t, N bus is the number of load nodes.

9. A device for predicting the maximum power supply capacity of an energy storage terminal based on TCN-LSTM, characterized in that: The device comprises a memory and a processor; The memory is used to store computer program code and transmit the computer program code to the processor; The processor is configured to execute the method according to any one of claims 1 to 4 according to instructions in the computer program code.

10. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the method according to any one of claims 1 to 4 is implemented.

Images