Unmanned aerial vehicle unit scheduling optimization method and system for pumped storage power station

Abstract

The invention belongs to the field of intelligent power grids, and discloses an unmanned aerial vehicle unit scheduling optimization method and system for a pumped storage power station. According to the method, the historical operation condition data of the pumped storage power station is processed through the machine learning model to predict the real-time working state of the pumped storage power station in the future, and the scheduling of the unmanned aerial vehicle group is optimized in combination with the corresponding power grid constraint condition; and calling a mathematical optimization toolkit solver to solve the constraint conditions set by the real-time working state of the pumped storage power station and the planning model constructed by the corresponding power grid constraint conditions to realize dynamic scheduling optimization. According to the method, the dynamic change of the pumped storage power station is considered, and the objective function model constructed after the real-time working state of the pumped storage power station is obtained can better obtain the optimal unit scheduling optimization method according to the constraint condition, so that a solver is adopted to quickly and reliably solve, and the intelligent scheduling degree of the whole power grid can be improved.

Description

Technical Field

[0001] The present invention belongs to the technical field of smart grids, and in particular relates to a method and system for optimizing the dispatching of unmanned aerial vehicle groups in a pumped storage power station. Background Art

[0002] As a highly advocated clean energy, hydropower has now penetrated into the distribution networks of cities and towns as a distributed power source, causing serious impacts on the quality of the distribution network, such as the voltage and frequency stability of the distribution network. In order to cope with the impact of hydropower changes on the power grid, the distribution network needs to increase investment, resulting in a decrease in the benefits of hydropower access. This also greatly restricts the further development of hydropower in urban and rural distribution networks. Connecting pumped-storage power stations to the distribution network can increase the proportion of clean energy in the power system, mobilize users' enthusiasm for investing in the construction of pumped-storage power stations, and achieve the purpose of protecting the environment and reducing the cost of electricity resources. Improving the distribution network's ability to absorb pumped-storage power stations and suppressing the impact of hydropower on the distribution network have become difficult problems facing the further development of hydropower, a clean energy. The distribution network's absorption of hydropower includes two aspects:

[0003] One is to increase the power consumption, and the other is to improve the ability of the distribution network to maintain voltage and frequency stability. Through research, a method and system for optimizing the dispatching of unmanned aerial vehicle groups in pumped-storage power stations was found, thereby improving the distribution network's ability to absorb hydropower. Energy storage equipment can participate in grid dispatching as both a controllable load and a controllable power source, but the applicant believes that in the power grid, thermal power has very mature real-time control technology and mature equipment, while energy storage equipment does not have the conditions required for flexible control. Therefore, the pumped storage device is used to process the historical operating data of the pumped storage power station through a machine learning model to predict the real-time working status of the pumped storage power station in the future, and combined with the corresponding power grid constraints, the dispatching of unmanned aerial vehicle groups is optimized, and the mathematical optimization toolkit solver is called to solve the constraints set for the real-time working status of the pumped storage power station and the planning model constructed by the corresponding power grid constraints to achieve dynamic dispatch optimization.

[0004] The present invention takes into account the dynamic changes of the pumped-storage power station. The objective function model constructed after obtaining the real-time working status of the pumped-storage power station can better obtain the best unit scheduling optimization method according to the constraint conditions, thereby using the solver to quickly and reliably solve the problem, which can improve the intelligent scheduling level of the entire power grid. Summary of the invention

[0005] In order to solve the above technical problems, the present invention proposes a method and system for optimizing the dispatching of unmanned aerial vehicle groups in a pumped storage power station.

[0006] In a first aspect of the present invention, a method for optimizing the dispatching of a drone group of a pumped storage power station is provided, wherein the method comprises the following steps:

[0007] S1. Establish an objective function model with the goal of minimizing the power loss of all units in the power grid connected to the pumped storage power station;

[0008] S2. Process the historical operation data of the pumped-storage power station through a machine learning model to predict the real-time working status of the pumped-storage power station, and set constraints based on the real-time working status;

[0009] S3, constructing a dynamic programming model according to the objective function model and the real-time working status;

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

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

[0012] Furthermore, the first historical operating status data includes water storage fluctuation parameters of the pumped-storage power station during operation.

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

[0014] Furthermore, the second real-time operating status data is processed according to the trained machine learning model to obtain the number of real-time scheduling cycle time periods corresponding to the second real-time working state to construct a constraint condition, the second real-time operating status data is a parameter with the same parameter characteristics as the first historical operating status data but a different characteristic vector value, and the second real-time working state is a parameter with the same parameter characteristics as the first historical working state but a different characteristic vector value.

[0015] Furthermore, the dynamic programming model is constructed according to the objective function model and the real-time working state, and the dynamic programming model is constructed with the constraint conditions 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 dispatching of unmanned aerial vehicle groups in a pumped storage power station is also provided, wherein the system comprises:

[0018] Objective function building module: Establishing an objective function model with the goal of minimizing the power loss of all units in the power grid connected to the pumped storage power station;

[0019] Model prediction and constraint setting module: This module processes the historical operation data of the pumped-storage power station through a machine learning model to predict the real-time working status of the pumped-storage power station in the future, and sets constraints based on the real-time working status;

[0020] Dynamic programming model module: constructing a dynamic programming model according to the objective function model and the real-time working status;

[0021] Optimization solution module: Use the solver optimization algorithm tool to solve the dynamic programming model and obtain the optimal scheduling solution of the pumped storage power station.

[0022] The present invention processes the historical operating data of the pumped-storage power station through a machine learning model to predict the real-time working status that the pumped-storage power station will face in the future, and optimizes the scheduling of the unmanned aerial vehicle group in combination with the corresponding power grid constraints. The mathematical optimization toolkit solver is called to solve the constraints set for the real-time working status of the pumped-storage power station and the planning model constructed by the corresponding power grid constraints to achieve dynamic scheduling optimization.

[0023] The present invention takes into account the dynamic changes of the pumped-storage power station. The objective function model constructed after obtaining the real-time working status of the pumped-storage power station can better obtain the best unit scheduling optimization method according to the constraint conditions, and the time period classification model is set according to the situation of the pumped-storage power station. The constructed dynamic programming model adopts a solver to solve quickly and reliably, which can improve the intelligent scheduling level of the entire power grid. BRIEF DESCRIPTION OF THE DRAWINGS

[0024] Figure 1 It is a flow chart of a method for optimizing the dispatching of a UAV group of a pumped storage power station according to the present invention;

[0025] Figure 2 It is a structural schematic diagram of a UAV group dispatching optimization system for a pumped storage power station of the present invention;

[0026] Figure 3 It is a schematic diagram of the distribution network structure in the present invention;

[0027] Figure 4 It is the output power of the distribution network of the present invention in one day. DETAILED DESCRIPTION

[0028] The invention is further described below in conjunction with the accompanying drawings and specific implementation methods.

[0029] In a first aspect of the present invention, in order to solve the above-mentioned technical problems, the present invention provides a method and system for optimizing the dispatching of unmanned aerial vehicle groups in a pumped storage power station.

[0030] In a first aspect of the present invention, a method for optimizing the dispatching of a drone group of a pumped storage power station is provided, wherein the method comprises the following steps:

[0031] S1. Establish an objective function model with the goal of minimizing the power loss of all units in the power grid connected to the pumped storage power station;

[0032] S2. Process the historical operation data of the pumped-storage power station through a machine learning model to predict the real-time working status of the pumped-storage power station, and set constraints based on the real-time working status;

[0033] S3, constructing a dynamic programming model according to the objective function model and the real-time working status;

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

[0035] Furthermore, the objective function model is:

[0036]

[0037] Where min P is the minimum power loss function of all units, T p is the total number of time periods in the scheduling cycle, N n is the number of thermal power generating units in the power grid, P rc，t is the average power loss of thermal power unit r in the power grid during period t, P wc，t is the average power loss of the pumped storage power station w in the power grid during period t, ε rc，t is the power loss of thermal power unit r during the start-up and shutdown period t, ε wc，t is the power loss of the pumped storage power station w in the power grid during the start-up and shutdown period t, P or，t is the transmission loss power of thermal power unit r in the power grid during period t, P ow，t is 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 working state of the pumped-storage power station in the future, the machine learning model is trained using the first historical operating status data and the first historical working state, and the machine learning model is a neural network model or a SVM model.

[0039] Furthermore, the first historical operation data includes a water storage fluctuation parameter of the pumped storage power station during operation, and the calculation formula is:

[0040]

[0041] Where V is the water storage fluctuation parameter, h is the working range height of the pumped storage power station, and hu is the upper limit height of water storage of pumped storage power station, h d It is the lower limit of water storage height of pumped storage power station.

[0042] Furthermore, the first historical operation status data also includes the voltage peak value and pumping power of the pumped-storage power station, and the first historical working state is the corresponding historical scheduling cycle time period number.

[0043] Furthermore, the second real-time operating status data is processed according to the trained machine learning model to obtain the number of real-time scheduling cycle time periods corresponding to the second real-time working state to construct a constraint condition, the second real-time operating status data is a parameter with the same parameter characteristics as the first historical operating status data but a different characteristic vector value, and the second real-time working state is a parameter with the same parameter characteristics as the first historical working state but a different characteristic vector value.

[0044] Furthermore, the dynamic programming model is constructed according to the objective function model and the real-time working state, and the dynamic programming model is constructed with the constraint conditions obtained from the real-time working state and the objective function model.

[0045] Furthermore, the constraint condition is:

[0046]

[0047] Where P rc，t is the average power loss of thermal power unit r in the power grid during period t, P kr is the single on / off power loss of thermal power unit r in the power grid, P wc，t is the average power loss of the pumped storage power station w in the power grid during period t, P kw is the single on / off power loss of the pumped storage power station w, M r，t is the working state of the thermal power unit r in the real-time dispatch cycle time period t, M w，t is the working state of the pumped storage power station w in the real-time dispatching cycle time period t, the starting working state value is 1, and the shutdown state value is 0;

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

[0049] (ω T X+b)≥V tp

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

[0051] A system for optimizing the dispatching of unmanned aerial vehicle groups in a pumped storage power station is also provided, wherein the system comprises:

[0052] Objective function building module: Establishing an objective function model with the goal of minimizing the power loss of all units in the power grid connected to the pumped storage power station;

[0053] Model prediction and constraint setting module: This module processes the historical operation data of the pumped-storage power station through a machine learning model to predict the real-time working status of the pumped-storage power station in the future, and sets constraints based on the real-time working status;

[0054] Dynamic programming model module: constructing a dynamic programming model according to the objective function model and the real-time working status;

[0055] Optimization solution module: Use the solver optimization algorithm tool to solve the dynamic programming model and obtain the optimal scheduling solution of the pumped storage power station.

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

[0057] Furthermore, a UAV group dispatching optimization system for a pumped storage power station is characterized by:

[0058] The objective function model is:

[0059]

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

[0061] The present invention processes the historical operating data of the pumped-storage power station through a machine learning model to predict the real-time working status that the pumped-storage power station will face in the future, and optimizes the scheduling of the unmanned aerial vehicle group in combination with the corresponding power grid constraints. The mathematical optimization toolkit solver is called to solve the constraints set for the real-time working status of the pumped-storage power station and the planning model constructed by the corresponding power grid constraints to achieve dynamic scheduling optimization.

[0062] The present invention takes into account the dynamic changes of the pumped-storage power station. The objective function model constructed after obtaining the real-time working status of the pumped-storage power station can better obtain the best unit scheduling optimization method according to the constraint conditions, and the time period classification model is set according to the situation of the pumped-storage power station. The constructed dynamic programming model adopts a solver to solve quickly and reliably, which can improve the intelligent scheduling level of the entire power grid.

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

[0064] The present invention does not particularly specify the structure of some modules, which shall be subject to the contents recorded in the prior art. The prior art mentioned in the above background technology section and the specific embodiment section of the present invention can be used as part of the present invention to understand the meaning of some technical features or parameters. The scope of protection of the present invention shall be subject to the contents actually recorded in the claims.

Claims

1. A method for optimizing the dispatching of unmanned aerial vehicle groups in a pumped storage power station, characterized in that: The method comprises the following steps: S1. Establish an objective function model with the goal of minimizing the power loss of all units in the power grid connected to the pumped storage power station; S2. Process the historical operation data of the pumped-storage power station through a machine learning model to predict the real-time working status of the pumped-storage power station, and set constraints based on the real-time working status; S3, constructing a dynamic programming model according to the objective function model and the real-time working status; S4. Use a solver optimization algorithm tool to solve the dynamic programming model and obtain the optimal scheduling solution of the pumped storage power station.

2. The method for optimizing the dispatching of a UAV group of a pumped storage power station according to claim 1, characterized in that: Before predicting the real-time working 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 working state. The machine learning model is a neural network model or a SVM model.

3. A method for optimizing the dispatching of unmanned aerial vehicle groups in a pumped storage power station as claimed in claim 2, characterized in that: The first historical operation status data includes water storage fluctuation parameters of the pumped-storage power station during operation.

4. A method for optimizing the dispatching of unmanned aerial vehicle groups in a pumped storage power station as claimed in claim 3, characterized in that: The first historical operation status data also includes the voltage peak value and pumping power of the pumped-storage power station, and the first historical working state is the corresponding historical scheduling cycle time period number.

5. The method for optimizing the dispatching of unmanned aerial vehicle groups in a pumped storage power station according to claim 4, characterized in that: The second real-time operating status data is processed according to the machine learning model obtained through training to obtain the number of real-time scheduling cycle time periods corresponding to the second real-time working state to construct constraint conditions, wherein the second real-time operating status data is a parameter with the same parameter characteristics as the first historical operating status data but with a different characteristic vector value, and the second real-time working state is a parameter with the same parameter characteristics as the first historical working state but with a different characteristic vector value.

6. A method for optimizing the dispatching of unmanned aerial vehicle groups in a pumped storage power station as claimed in claim 5, characterized in that: The dynamic programming model is constructed according to the objective function model and the real-time working state, and the dynamic programming model is constructed with the constraint conditions obtained from the real-time working state and the objective function model.

7. A method for optimizing the dispatching of unmanned aerial vehicle groups in a pumped storage power station as claimed in claim 6, characterized in that: The machine learning model is an improved SVM model.

8. A system for optimizing the dispatching of unmanned aerial vehicle groups in a pumped storage power station, the system implementing the method according to any one of claims 1 to 7, characterized in that the system include: Objective function building module: Establishing an objective function model with the goal of minimizing the power loss of all units in the power grid connected to the pumped storage power station; Model prediction and constraint setting module: This module processes the historical operation data of the pumped-storage power station through a machine learning model to predict the real-time working status of the pumped-storage power station in the future, and sets constraints based on the real-time working status; Dynamic programming model module: constructing a dynamic programming model according to the objective function model and the real-time working status; Optimization solution module: Use the solver optimization algorithm tool to solve the dynamic programming model and obtain the optimal scheduling solution of the pumped storage power station.

Images