A method and system for unmanned unit dispatching optimization of a pumped storage power station

Abstract

The application belongs to the field of smart grid, and discloses a kind of unmanned unit scheduling optimization method and system of pumped storage power station.The application predicts the real-time working state faced by pumped storage power station in the future by processing the historical operation data of pumped storage power station through machine learning model, and combines the corresponding power grid constraint condition to optimize the unmanned unit scheduling, calls mathematical optimization toolkit solver to solve the planning model constructed by the constraint condition set for the real-time working state of the pumped storage power station and the corresponding power grid constraint condition to realize dynamic scheduling optimization.The application considers the dynamic change of pumped storage power station, and the objective function model constructed after obtaining the real-time working state of pumped storage power station can better obtain the best unit scheduling optimization method according to the constraint condition, so as to quickly and reliably solve by using the solver, and can improve the intelligent scheduling degree of the whole power grid.

Description

Technical Field

[0001] This invention belongs to the field of smart grid technology, and in particular relates to a method and system for optimizing the scheduling of unmanned aerial vehicles (UAVs) in pumped storage power stations. Background Technology

[0002] Hydropower, as a highly promoted clean energy source, has already penetrated urban and rural power distribution networks as a distributed power source, severely impacting network quality such as voltage and frequency stability. To cope with the impact of hydropower fluctuations on the power grid, the distribution network needs increased investment, leading to a decline in the efficiency of hydropower integration. This also significantly limits the further development of hydropower in urban and rural power distribution networks. Integrating pumped-storage hydropower stations into the distribution network can increase the proportion of clean energy in the power system, incentivize users to invest in the construction of pumped-storage hydropower stations, and achieve the goals of environmental protection and reduced electricity resource costs. Improving the distribution network's capacity to absorb pumped-storage hydropower and mitigating the impact of hydropower on the distribution network have become challenging issues for the further development of this clean energy source. The distribution network's absorption of hydropower includes two aspects…

[0003] The research aims to improve two key aspects: firstly, to increase the power absorption capacity, and secondly, to enhance the distribution network's ability to maintain voltage and frequency stability. A method and system for optimizing the scheduling of unmanned aerial vehicles (UAVs) at pumped-storage power stations has been developed, thereby improving the distribution network's capacity to absorb hydropower. Energy storage devices can participate in grid dispatch as both controllable loads and controllable power sources. However, the applicant believes that thermal power has mature real-time control technologies and equipment, while energy storage devices lack the necessary conditions for flexible control. Therefore, a machine learning model is used to process historical operating data of pumped-storage power stations to predict their future real-time operating states. Combined with relevant grid constraints, UAV scheduling is optimized. A mathematical optimization toolkit solver is used to solve a planning model constructed based on the constraints set for the real-time operating state of the pumped-storage power station and the corresponding grid constraints to achieve dynamic scheduling optimization.

[0004] This invention takes into account the dynamic changes of pumped storage power stations. After obtaining the real-time operating status of the pumped storage power station, the objective function model constructed can better obtain the optimal unit scheduling optimization method according to the constraints. Thus, the solver can solve the problem quickly and reliably, which can improve the intelligent scheduling level of the entire power grid. Summary of the Invention

[0005] To address the aforementioned technical problems, this invention proposes a method and system for optimizing the scheduling of unmanned aerial vehicles (UAVs) in pumped storage power stations.

[0006] In a first aspect of the present invention, a method for optimizing the scheduling of unmanned aerial vehicle (UAV) units in a pumped storage power station is provided, characterized in that the method includes the following steps:

[0007] S1. Establish an objective function model with the goal of minimizing the power loss of all generating units in the power grid, including pumped storage power stations.

[0008] S2. Use machine learning models to process historical operating data of pumped storage power stations to predict the real-time operating status of pumped storage power stations, and set constraints based on the real-time operating status.

[0009] S3. Construct a dynamic programming model based on the objective function model and the real-time working status;

[0010] S4. Use the solver optimization algorithm tool to solve the dynamic programming model and obtain the optimal scheduling solution for the pumped storage power station.

[0011] Furthermore, before predicting the real-time operating state of the pumped storage power station in the future, the machine learning model is trained using the first historical operating data and the first historical operating state. The machine learning model is a neural network model or an SVM model.

[0012] Furthermore, the first historical operating data includes parameters on the water storage fluctuations of the pumped storage power station during its operation.

[0013] Furthermore, the first historical operating data also includes the peak voltage and pumping power of the pumped storage power station, and the first historical operating status is the corresponding history.

[0014] Furthermore, constraints are constructed by processing the second real-time operation data based on the machine learning model obtained through training to obtain the number of real-time scheduling period periods corresponding to the second real-time working state. The second real-time operation data is a parameter with the same parameter features as the first historical operation data but different feature vector values. The second real-time working state is a parameter with the same parameter features as the first historical working state but different feature vector values.

[0015] Furthermore, the dynamic programming model is constructed based on the objective function model and the real-time working state, using the constraints obtained from the real-time working state and the objective function model.

[0016] Furthermore, the machine learning model is an improved SVM model.

[0017] A system for optimizing the scheduling of unmanned aerial vehicles (UAVs) in a pumped storage power station is also provided, characterized in that the system includes:

[0018] Objective function construction module: To establish an objective function model with the objective of minimizing the power loss of all generating units in the power grid, including pumped storage power stations;

[0019] Model prediction and constraint setting module: It uses machine learning models to process historical operating data of pumped storage power stations to predict the real-time operating status that pumped storage power stations will face in the future, and sets constraints based on the real-time operating status.

[0020] Dynamic programming model module: Constructs a dynamic programming model based on the objective function model and the real-time operating status;

[0021] Optimization and solution module: The dynamic programming model is solved using the solver optimization algorithm tool to obtain the optimal scheduling solution for the pumped storage power station.

[0022] This invention employs a machine learning model to process historical operational data of pumped storage power stations to predict their future real-time operating status. Combined with relevant grid constraints, it optimizes the scheduling of unmanned aerial vehicle (UAV) groups. The invention also uses a mathematical optimization toolkit solver to solve a planning model constructed based on the constraints set for the real-time operating status of the pumped storage power station and the corresponding grid constraints, thereby achieving dynamic scheduling optimization.

[0023] This invention considers the dynamic changes of pumped storage power stations. After obtaining the real-time operating status of the pumped storage power stations, the objective function model constructed can better obtain the optimal unit scheduling optimization method according to the constraints. The time period classification model is set for the pumped storage power station situation, and the constructed dynamic programming model can be solved quickly and reliably using a solver, which can improve the intelligent scheduling level of the entire power grid. Attached Figure Description

[0024] Figure 1 This is a flowchart of an unmanned aerial vehicle (UAV) scheduling optimization method for a pumped storage power station according to the present invention;

[0025] Figure 2 This is a schematic diagram of the unmanned aerial vehicle (UAV) scheduling optimization system for a pumped storage power station according to the present invention.

[0026] Figure 3 This is a schematic diagram of the power distribution network structure in this invention;

[0027] Figure 4 This refers to the output power of the power distribution network of this invention within one day. Detailed Implementation

[0028] The invention will now be further described in conjunction with the accompanying drawings and specific embodiments.

[0029] In a first aspect, to solve the above-mentioned technical problems, this invention provides a method and system for optimizing the scheduling of unmanned aerial vehicles (UAVs) in a pumped storage power station.

[0030] In a first aspect of the present invention, a method for optimizing the scheduling of unmanned aerial vehicle (UAV) units in a pumped storage power station is provided, characterized in that the method includes the following steps:

[0031] S1. Establish an objective function model with the goal of minimizing the power loss of all generating units in the power grid, including pumped storage power stations.

[0032] S2. Use machine learning models to process historical operating data of pumped storage power stations to predict the real-time operating status of pumped storage power stations, and set constraints based on the real-time operating status.

[0033] S3. Construct a dynamic programming model based on the objective function model and the real-time working status;

[0034] S4. Use the solver optimization algorithm tool to solve the dynamic programming model and obtain the optimal scheduling solution for the pumped storage power station.

[0035] Furthermore, the objective function model is as follows:

[0036]

[0037] In the formula, min P is the minimum power loss function for all units, and T p N represents the total number of time periods in the scheduling cycle. n P represents the number of thermal power units in the power grid. rc，t P represents the average power loss of the thermal power unit r in the power grid during time period t. wc，t Let ε be the average power loss of the pumped storage power station w in the power grid during time period t. rc，t Let ε be the power loss during start-up and shutdown of the thermal power unit r in the power grid during time period t. wc，t P represents the power loss during start-up and shutdown of a pumped storage power station w in time period t. or，t P represents the transmission loss power of the thermal power unit r in the power grid during time period t. ow，t The transmission loss power of the pumped storage power station w in the power grid during time period t.

[0038] Furthermore, before predicting the real-time operating state of the pumped storage power station in the future, the machine learning model is trained using the first historical operating data and the first historical operating state. The machine learning model is a neural network model or an SVM model.

[0039] Furthermore, the first historical operational data includes water storage fluctuation parameters of the pumped storage power station during its operation, calculated using the following formula:

[0040]

[0041] In the formula, V is the water storage fluctuation parameter, h is the working range height of the pumped storage power station, and hu h represents the upper limit of water storage height for a pumped storage power station. d This represents the lower limit of the water storage height for pumped storage power stations.

[0042] Furthermore, the first historical operating data also includes the peak voltage and pumping power of the pumped storage power station, and the first historical operating status is the corresponding historical scheduling cycle period.

[0043] Furthermore, constraints are constructed by processing the second real-time operation data based on the machine learning model obtained through training to obtain the number of real-time scheduling period periods corresponding to the second real-time working state. The second real-time operation data is a parameter with the same parameter features as the first historical operation data but different feature vector values. The second real-time working state is a parameter with the same parameter features as the first historical working state but different feature vector values.

[0044] Furthermore, the dynamic programming model is constructed based on the objective function model and the real-time working state, using the constraints obtained from the real-time working state and the objective function model.

[0045] Furthermore, the constraint conditions are as follows:

[0046]

[0047] In the formula, P rc，t P represents the average power loss of the thermal power unit r in the power grid during time period t. kr P represents the single-cycle switching loss power of the thermal power unit r in the power grid. wc，t Let P be the average power loss of the pumped storage power station w in the power grid during time period t. kw M represents the single-cycle switching loss power of the pumped storage power station w in the power grid. r，t M represents the operating state of the power grid thermal power unit r during the real-time dispatch cycle period t. w，t The operating status of the pumped storage power station w in the real-time dispatch cycle period t is 1 when it is in operation and 0 when it is off.

[0048] Furthermore, the machine learning model is an improved SVM model, and its decision calculation formula is as follows:

[0049] (ω T X+b)≥V tp

[0050] Where ω and b are the normal vector and intercept of the hyperplane, respectively, X is the first historical running data or the second real-time running data, and V... tpq The parameter represents the water storage fluctuation during the qth time period, where the total scheduling time t is divided into time periods p.

[0051] A system for optimizing the scheduling of unmanned aerial vehicles (UAVs) in a pumped storage power station is also provided, characterized in that the system includes:

[0052] Objective function construction module: To establish an objective function model with the objective of minimizing the power loss of all generating units in the power grid, including pumped storage power stations;

[0053] Model prediction and constraint setting module: It uses machine learning models to process historical operating data of pumped storage power stations to predict the real-time operating status that pumped storage power stations will face in the future, and sets constraints based on the real-time operating status.

[0054] Dynamic programming model module: Constructs a dynamic programming model based on the objective function model and the real-time operating status;

[0055] Optimization and solution module: The dynamic programming model is solved using the solver optimization algorithm tool to obtain the optimal scheduling solution for the pumped storage power station.

[0056] In this embodiment, the mathematical optimization toolkit CPLEX solver specifically uses the SBB and CPLEX solvers in the GAMS software to linearize the planning model.

[0057] Furthermore, a scheduling optimization system for unmanned aerial vehicles (UAVs) in a pumped storage power station is characterized by:

[0058] The objective function model is as follows:

[0059]

[0060] In the formula, min P is the minimum power loss function for all units, and T p N represents the total number of time periods in the scheduling cycle. n P represents the number of thermal power units in the power grid. rc，t P represents the average power loss of the thermal power unit r in the power grid during time period t. wc，t Let ε be the average power loss of the pumped storage power station w in the power grid during time period t. rc，t Let ε be the power loss during start-up and shutdown of the thermal power unit r in the power grid during time period t. wc，t P represents the power loss during start-up and shutdown of a pumped storage power station w in time period t. or，t P represents the transmission loss power of the thermal power unit r in the power grid during time period t. ow，t The transmission loss power of the pumped storage power station w in the power grid during time period t.

[0061] This invention employs a machine learning model to process historical operational data of pumped storage power stations to predict their future real-time operating status. Combined with relevant grid constraints, it optimizes the scheduling of unmanned aerial vehicle (UAV) groups. The invention also uses a mathematical optimization toolkit solver to solve a planning model constructed based on the constraints set for the real-time operating status of the pumped storage power station and the corresponding grid constraints, thereby achieving dynamic scheduling optimization.

[0062] This invention considers the dynamic changes of pumped storage power stations. After obtaining the real-time operating status of the pumped storage power stations, the objective function model constructed can better obtain the optimal unit scheduling optimization method according to the constraints. The time period classification model is set for the pumped storage power station situation, and the constructed dynamic programming model can be solved quickly and reliably using a solver, which can improve the intelligent scheduling level of the entire power grid.

[0063] The combination of multiple embodiments of the present invention can achieve all the above effects, but it is not required that each embodiment of the present invention achieve all the above advantages and effects, because each embodiment of the present invention can constitute a separate technical solution and make one or more contributions to the prior art.

[0064] For any module structures not specifically defined in this invention, the existing technical specifications shall prevail. The existing technical specifications mentioned in the foregoing background and specific embodiments sections are considered part of this invention and are used to understand the meaning of certain technical features or parameters. The scope of protection of this invention is determined by the actual contents of the claims.

Claims

1. A method for optimizing the scheduling of unmanned aerial vehicle (UAV) units in a pumped storage power station, characterized in that, The method includes the following steps: S1. Establish an objective function model with the goal of minimizing the power loss of all generating units in the power grid, including pumped storage power stations. S2. Use machine learning models to process historical operating data of pumped storage power stations to predict the real-time operating status of pumped storage power stations, and set constraints based on the real-time operating status. Before predicting the real-time operating status of the pumped storage power station, the machine learning model is trained using first-historical operating data and first-historical operating status. The machine learning model is an improved SVM model, and its decision calculation formula is as follows: ； Where ω and b are the normal vector and intercept of the hyperplane, respectively, and X is the first historical running data. The water storage fluctuation parameter is defined as the number of time periods t in the real-time scheduling cycle, with p as the time period. The first historical operational data includes parameters on the water storage fluctuations of the pumped storage power station during its operation, calculated using the following formula: ； In the formula, For parameters related to water storage fluctuations, This refers to the working range height of the pumped storage power station. This refers to the upper limit of the water storage height for pumped storage power stations. This refers to the lower limit of the water storage height for pumped storage power stations. The constraints are as follows: ； In the formula, Let r be the average power loss of a thermal power unit in the power grid during the number of time periods t in the real-time dispatch cycle. For power grid thermal power units Power loss during a single power-on / off cycle Pumped storage power station for power grid Average power loss during real-time scheduling period t Pumped storage power station for power grid Power loss during a single power-on / off cycle This refers to the operating status of the thermal power unit r in the real-time dispatch cycle period t. Pumped storage power station for power grid During the real-time scheduling period t, the power-on status value is 1, and the power-off status value is 0. S3. Construct a dynamic programming model based on the objective function model and the real-time working status; The objective function model is as follows: ； In the formula, This is the minimum power loss function for all units. The total number of time periods in the scheduling cycle. The number of thermal power units in the power grid. Let r be the average power loss of a thermal power unit in the power grid during the number of time periods t in the real-time dispatch cycle. Pumped storage power station for power grid Average power loss during real-time scheduling period t Let r be the power loss from start-up and shutdown of a thermal power unit in the power grid during the number of time periods t in the real-time dispatch cycle. Pumped storage power station for power grid Power loss during start-up and shutdown within a real-time scheduling cycle period t. Let r be the transmission loss power of the thermal power unit r in the real-time dispatch cycle number t. Pumped storage power station for power grid Transmission loss power during real-time scheduling period t; S4. Use the solver optimization algorithm tool to solve the dynamic programming model and obtain the optimal scheduling solution for the pumped storage power station.

2. The method for optimizing the scheduling of unmanned aerial vehicle (UAV) units in a pumped storage power station as described in claim 1, characterized in that: The first historical operating data also includes the peak voltage and pumping power of the pumped storage power station, and the first historical operating status is the corresponding historical scheduling cycle period.

3. The method for optimizing the scheduling of unmanned aerial vehicle (UAV) units in a pumped storage power station as described in claim 2, characterized in that: Based on the machine learning model obtained through training, the second real-time operation data is processed to obtain the number of real-time scheduling period periods corresponding to the second real-time working state, and constraints are constructed. The second real-time operation data is a parameter with the same parameter features as the first historical operation data but different feature vector values. The second real-time working state is a parameter with the same parameter features as the first historical working state but different feature vector values.

4. A scheduling and optimization system for unmanned aerial vehicles (UAVs) in a pumped storage power station, the system implementing the method described in any one of claims 1-3, characterized in that the system... include: Objective function construction module: To establish an objective function model with the objective of minimizing the power loss of all generating units in the power grid, including pumped storage power stations; Model prediction and constraint setting module: It uses machine learning models to process historical operating data of pumped storage power stations to predict the real-time operating status that pumped storage power stations will face in the future, and sets constraints based on the real-time operating status. Dynamic programming model module: Constructs a dynamic programming model based on the objective function model and the real-time operating status; The objective function model is as follows: ； In the formula, This is the minimum power loss function for all units. The total number of time periods in the scheduling cycle. The number of thermal power units in the power grid. Let r be the average power loss of a thermal power unit in the power grid during the number of time periods t in the real-time dispatch cycle. Pumped storage power station for power grid Average power loss during real-time scheduling period t Let r be the power loss from start-up and shutdown of a thermal power unit in the power grid during the number of time periods t in the real-time dispatch cycle. Pumped storage power station for power grid Power loss during start-up and shutdown within a real-time scheduling cycle period t. Let r be the transmission loss power of the thermal power unit r in the real-time dispatch cycle number t. Pumped storage power station for power grid Transmission loss power during real-time scheduling period t; Optimization and solution module: The dynamic programming model is solved using the solver optimization algorithm tool to obtain the optimal scheduling solution for the pumped storage power station.

Images