Hybrid pumping and storage optimal scheduling method and system giving consideration to peak regulation and risk avoidance

Abstract

The invention relates to the technical field of multi-objective optimization operation of a hybrid pumping and storage-containing water-wind-solar complementary system, and discloses a hybrid pumping and storage optimization scheduling method and system considering peak regulation and risk avoidance. The method comprises the following steps: acquiring historical data of a cascade basin water-wind-light power station and a power grid load, generating scheduling initial data through prediction, and generating scheduling scene data capable of representing the uncertainty of new energy based on the scheduling initial data and a prediction error; constructing an optimization model, solving the optimization model, and based on the solution of the optimization model, generating scheduling instructions of the drainage basin cascade hydropower station and the hybrid pumped storage power station; through the optimization model, the risk of power abandoning and load losing at the hydropower side is minimized, collaborative optimization meeting the peak regulation requirement of the power grid side is achieved, the flexible regulation advantage of hybrid pumping and storage is brought into full play, and stable operation and reliability of the system are promoted; and through linear processing, the solving speed and precision are improved, the real-time calculation efficiency of the system is ensured, and making of a scheduling decision is more efficient.

Description

Technical Field

[0001] The present invention relates to the technical field of multi-objective optimization operation of a water-wind-solar complementary system including a hybrid pumped storage, and specifically to a hybrid pumped storage optimization scheduling method and system that takes into account both peak load regulation and risk avoidance. Background Art

[0002] As the proportion of new energy in my country's energy structure increases year by year, the challenges of its consumption brought about by seasonal, random and intermittent fluctuations are becoming increasingly prominent. Relying on the existing large-scale cascade hydropower bases, building a water, wind and solar multi-energy complementary dispatching system provides an effective way to solve the problem of new energy consumption. However, constrained by factors such as water supply, shipping, and flood control, the flexible adjustment capacity of conventional hydropower is limited. With the continuous increase in the penetration rate of new energy, the phenomenon of wind and solar abandonment is difficult to avoid. The formation of a water, wind and solar multi-energy complementary system with energy storage has higher flexibility and helps to alleviate the phenomenon of wind and solar abandonment. Pumped storage, as a mature form of energy storage, can play a flexible adjustment role in both electricity and hydraulic power, and is suitable for water, wind and solar multi-energy complementary systems.

[0003] Pumped storage power stations can be divided into pure pumped storage power stations and hybrid pumped storage power stations according to their types. As a relatively mature technology at present, pure pumped storage power stations have been widely used in the dispatching of complementary systems. However, due to the limitations of geological and head conditions, pure pumped storage power stations are difficult to site and build, and it is difficult to meet the demand for energy storage scale for high-proportion access to new energy. The integration and transformation of conventional cascade hydropower into hybrid pumped storage power stations can shorten the construction period, reduce the difficulty of site selection, and effectively promote the development of pumped storage. Therefore, cascade hydropower with hybrid pumped storage will be a common form of cascade hydropower in the future basin.

[0004] In the day-ahead dispatch, it is necessary to formulate a suitable power generation plan for the water-wind-solar complementary system with hybrid pumped storage to ensure that hydropower has sufficient regulation capacity to compensate for the fluctuation of renewable energy. Affected by the forecast error of renewable energy, there may be a deviation between the power generation plan and the actual output, which is easy to cause the risk of power abandonment and load loss, and bring potential risks to the complementary operation of the hydropower side. Therefore, in the complementary system with hybrid pumped storage, the risk of power abandonment and load loss should be reduced. In addition, in the actual dispatch, in order to enable the receiving power grid to adopt the dispatch plan submitted by the complementary system, the power generation plan should also meet the peak load regulation needs of the power grid as much as possible. Cascade hydropower with hybrid pumped storage has the dual characteristics of source and load, and has better peak load regulation performance. However, due to the anti-peak load regulation characteristics of renewable energy, in the peak load regulation operation mode, the power generation plan of the complementary system is often similar to the load trend, but contrary to the renewable energy power generation trend. When the system regulation capacity is insufficient, achieving the peak load regulation demand may increase the risk of power abandonment and load loss of the complementary system. Therefore, how to balance the risk of power abandonment and load loss and the peak load regulation demand during the operation of the complementary system needs to be studied urgently.

[0005] In the context of high-proportion access to new energy, how to design a hybrid pumped storage optimization scheduling method and system that takes into account both peak load regulation and risk avoidance, so as to give full play to the advantages of hybrid pumped storage and achieve coordinated optimization of minimizing the risk of power abandonment and load loss on the hydropower side and meeting the peak load regulation needs on the grid side, has become an important issue. Summary of the invention

[0006] In view of the above problems, the purpose of the present invention is to provide a hybrid pumped storage optimization scheduling method and system that takes into account both peak load regulation and risk avoidance, which overcomes the problems of hybrid pumped storage power stations combined with cascade hydropower under the fluctuation of highly uncertain new energy (wind power and photovoltaic power), such as difficulty in peak load regulation on the grid side, high risk of power abandonment and load loss on the hydropower side, high water abandonment, and insufficient flexibility in upward and downward regulation. The technical solution is as follows: The hybrid pumped storage optimization dispatching method taking into account both peak load regulation and risk avoidance includes the following steps: Step 1: Data processing: Obtain historical data of cascade basin hydropower, wind and solar power stations and grid loads, generate initial dispatch data through prediction, and generate dispatch scenario data based on the initial dispatch data; Step 2: Construct a hybrid pumped storage optimization model that takes into account both peak load regulation and risk avoidance: determine the first optimization goal by minimizing the peak-to-valley difference of the remaining load to meet the peak load regulation demand on the grid side; determine the second optimization goal by minimizing the sum of the amount of abandoned power and the amount of lost load to meet the needs of reducing the risk of complementary operation and smoothing the fluctuation of new energy, and set corresponding constraints; Step 3: Model solving and dispatch instruction generation: Linearize the hybrid pumped storage optimization model that takes into account both peak load regulation and risk avoidance, and convert the original high-dimensional, multi-constrained, multi-variable complex mixed integer nonlinear programming (MINLP) model into a mixed integer linear programming (MILP) model. After importing the dispatch scenario data, call the solver to generate a Pareto solution set in combination with the constraint method, and determine the solution of the optimization model in the solution set based on the fuzzy entropy weight method to generate dispatch instructions for the cascade hydropower stations and hybrid pumped storage power stations in the basin.

[0007] The hybrid pumped storage optimization dispatching system that takes into account both peak load regulation and risk avoidance includes: Data acquisition module: used to obtain historical data of wind power and photovoltaic output in cascade basins, historical flow data of reservoirs of cascade hydropower stations in cascade basins, and historical data of receiving-end power grid load; Data processing module: used for data prediction, and splitting, scenario generation, reduction and combination based on the predicted data to obtain scheduling scenario data; Model building module: used to establish a hybrid pumped storage optimization model that takes into account both peak load regulation and risk avoidance, and to perform linearization processing to convert the original mixed integer nonlinear programming model into a mixed integer linear programming model; Solution module: Generates the Pareto solution set by combining the solver with the constraint method, and determines the solution of the optimization model in the solution set based on the fuzzy entropy weight method; Control and dispatch module: Based on the optimal solution of the parameter status of the cascade hydropower station and the optimal solution of the parameter status of the hybrid pumped-storage power station obtained by the solution module, the dispatch instructions of the cascade hydropower stations and hybrid pumped-storage power stations in the basin are generated.

[0008] The beneficial effects of the present invention are: The hybrid pumped storage optimization scheduling method and system that takes into account both peak load regulation and risk avoidance provided by the present invention effectively improves the regulation capacity of the hybrid pumped storage power station, enabling it to more efficiently cooperate with cascade hydropower to smooth out the fluctuations of new energy, effectively reducing the system's power abandonment and load loss phenomena, and improving the absorption rate of new energy. At the same time, it enables the cascade hydro-wind-solar complementary system containing hybrid pumped storage to smooth load fluctuations, reduce the peak-to-valley difference of the system's residual load, enhance the stability of the power grid, and reduce the phenomenon of water abandonment. In addition, the present invention reduces the amount of water abandonment and improves the utilization rate of water energy by optimizing the switching of pumping and power generation conditions in the hybrid pumped storage power station. BRIEF DESCRIPTION OF THE DRAWINGS

[0009] Figure 1 This is an overall flow chart of the hybrid pumped storage optimization scheduling method that takes into account both peak load regulation and risk avoidance of the present invention.

[0010] Figure 2 This is a structural schematic diagram of the hybrid pumped storage optimization scheduling system that takes into account both peak load regulation and risk avoidance of the present invention.

[0011] FIG3( a ) is a diagram showing the spring day-ahead scheduling result of the present invention.

[0012] FIG3( b ) is a diagram showing the summer day-ahead scheduling result of the present invention.

[0013] FIG3( c ) is a diagram showing the autumn day-ahead scheduling result of the present invention.

[0014] FIG3( d ) is a diagram showing the winter day-ahead scheduling result of the present invention.

[0015] The symbols in the figure are explained as follows: hybrid pumped storage optimization scheduling system-10, data acquisition module-101, data processing module-102, data prediction unit-1021, scenario generation unit-1022, model construction module-103, model generation unit-1031, linearization processing unit-1032, solution module-104, model solution unit-1041, optimal solution selection unit-1042, control scheduling module-105. DETAILED DESCRIPTION

[0016] The present invention is further described in detail below with reference to the accompanying drawings and specific embodiments.

[0017] Example 1, reference Figure 1 , which is an embodiment of the present invention, provides a hybrid pumped storage optimization scheduling method that takes into account both peak load regulation and risk avoidance, including: S1: Data processing: Obtain historical data of cascade basin hydropower, wind and solar power stations and power grid loads, generate initial scheduling data through prediction, and generate scheduling scenario data based on the initial scheduling data.

[0018] S1.1: Obtain historical data of hydropower, wind and solar power stations and grid loads in cascade basins, including historical data of wind power and photovoltaic output in cascade basins, historical flow data of reservoirs of cascade hydropower stations in cascade basins, and historical data of grid loads at receiving ends; S1.2: Generate initial dispatch data: predict the day-ahead forecast data of wind power and photovoltaic power output in the cascade basin, the day-ahead forecast data of the flow of each cascade hydropower station reservoir in the cascade basin, and the day-ahead forecast data of the load of the receiving power grid as the initial dispatch data; S1.3: Based on the wind power and photovoltaic output forecast data in the obtained initial dispatch data, the wind power and photovoltaic output forecast errors are assumed to obey the normal distribution and be independent. and Respectively represent The forecast error distribution of wind power and photovoltaic output in different time periods; Indicates that the variables in the formula are normally distributed; For the The wind power output forecast error for each period, For the The photovoltaic output prediction error of the time period; then, The wind power and photovoltaic combined forecast error distribution in the time period can be expressed as . Subsequently, the Latin hypercube sam-pling (LHS) stratified random sampling method is used in combination with its prediction error to simulate and obtain a scenario set with a specific number of wind power and photovoltaic output forecast data, and the iterative self-organizing data analysis techniques algorithm (ISODATA) algorithm is used to cluster the generated scenario set of wind power and photovoltaic output forecast data to generate a representative scenario set of wind power and photovoltaic output forecast data that characterizes the uncertainty of new energy and the probability of each scenario; and the scenario set is combined with the day-ahead forecast data of the flow of each cascade hydropower station reservoir in the cascade basin in the initial scheduling data and the day-ahead forecast data of the load of the receiving power grid to generate scheduling scenario data; S1.3.1: Generate dispatch scenario data, including: scenario sets and scenario probabilities of wind power and photovoltaic output forecast data in cascade river basins, receiving-end grid load forecast data, and predicted flow data of cascade river basin hydropower station reservoirs.

[0019] S2: Construct a hybrid pumped storage optimization model that takes into account both peak load regulation and risk avoidance: set the first optimization goal to meet the peak load regulation demand on the grid side; set the second optimization goal to meet the needs of reducing complementary operation risks and smoothing the fluctuations of new energy, and set corresponding constraints.

[0020] S2.1: Set the minimum peak-to-valley difference of the residual load as the first optimization goal, and the formula is expressed as: (1); (2); Where: Represents the peak-to-valley difference of the residual load, represents a certain period of time during the scheduling period. Represents the total time of the entire scheduling period, , Representative The remaining load of the period, Represents the original load of the power grid, Representative Power generation plan for the time period.

[0021] S2.2: Set the sum of the amount of abandoned power and the amount of lost load to be the minimum as the second optimization objective, and the formula is expressed as follows: (3); (4); (5); Where: Represents a cascade hydropower station in the basin. Represents a collection of cascade hydropower stations in the basin, ; Represents the total amount of power loss and load loss, Representative scene, A collection of scenes representing each season. , Representative scenes The corresponding probability is Representative scenes Next The amount of power wasted during the period, Representative scenes Next The amount of load loss during the period, Representative scenes Next The difference between the power generation plan and the planned output during the period, Representative scenes Next The combined output of wind power and photovoltaic power in the period, Representative scenes Next The power provided by wind power and photovoltaic power to the hybrid pumped storage hydropower in the period, Representative scenes Next Cascade hydropower station Output during the period, Representative scenes All cascade hydropower stations are under The period is the power provided by the hybrid pumped storage system. Representative scenes Hybrid pumped storage Output during the period.

[0022] S2.3: Set corresponding constraints, including: hydraulic and power constraints, power balance and ramp constraints, pumping and power generation mutual exclusion constraints, and grid power transmission constraints.

[0023] S2.3.1: Set hydraulic and electrical constraints. The specific formulas include: Unlike traditional cascade hydropower, hybrid pumped storage shares the reservoirs of its upper and lower hydropower stations, so its power generation head depends on the water levels of the reservoirs of its upper and lower hydropower stations. In addition, since the electric energy of hybrid pumped storage comes from wind power, photovoltaic power generation and traditional hydropower, the specific pumping power model is different from the power generation model. The details are as follows: (6); (7); (8); (9); (10); Where: and represent the water-to-electricity conversion efficiency and the electricity-to-water conversion efficiency, respectively. represents the density of water, represents the acceleration due to gravity, Representative scenes Next Cascade hydropower station The net head height of the time period, Representative scenes Hybrid Pumped Storage The net head height of the time period, Representative scenes Next Cascade hydropower station The power generation flow during the period, Representative scenes Hybrid Pumped Storage The power generation flow during the period, Representative scenes Hybrid Pumped Storage Pumping flow rate during the period, and Representing scenes Next Cascade hydropower station The reservoir water level and the reservoir tailwater level during the period, and Representing scenes Hybrid Pumped Storage The reservoir water level of the upper hydropower station and the reservoir water level of the lower hydropower station during the period; Representative scenes Hybrid Pumped Storage Total pumping power during the period; (11); (12); (13); (14); (15); Where: represents a set of hydropower stations, which includes conventional hydropower stations, upper-level hydropower stations corresponding to hybrid pumped storage, and lower-level hydropower stations corresponding to hybrid pumped storage. Representing a single hydroelectric power station, and They represent the reservoir capacities of the upper and lower hydropower stations corresponding to the hybrid pumped storage system. Representative scenes Next Cascade hydropower station The reservoir capacity during the period, Representative scenes Next Hydropower Station No. Reservoir inflow during the period, Representative scenes Next Hydropower Station No. The natural water flow of the reservoir during the period, Representative scenes Next Hydropower Station No. The outflow of the reservoir during the period, Representative scenes Next Hydropower Station No. The power generation flow during the period, Representative scenes Next Hydropower Station No. The amount of water abandoned during the period; and Representing scenes Next The natural water inflow from the reservoirs of the upstream and downstream hydropower stations corresponding to the period-time hybrid pumped storage; and Representing scenes Next The outflow of the reservoirs of the upper and lower hydropower stations corresponding to the hybrid pumped storage in the time period; Scenario Next Cascade hydropower station Natural water inflow from reservoirs during certain periods of time; Scenario Next Cascade hydropower station Reservoir outflow during the period; (16); Where: and Representing scenes Next The reservoir water level at the beginning of the dispatch period and the reservoir water level at the end of the dispatch period for each cascade hydropower station. and Representative The reservoir water level at the beginning of the dispatch period and the reservoir water level at the end of the dispatch period for each cascade hydropower station. and Representative The permissible fluctuation range of reservoir water level at the beginning and end of the dispatch period of each cascade hydropower station; (17); (18); (19); Where: and They represent the nonlinear fitting coefficients of the water level-storage capacity and tailwater level-discharge flow of the hydropower station reservoir, and Representing scenes Next Cascade hydropower station The average reservoir storage capacity and average reservoir discharge during the period, and Respectively represent Dead water level and normal operating water level of each cascade hydropower station reservoir; (20); (twenty one); Where: and Respectively represent Maximum power generation flow of cascade hydropower stations and hybrid pumped storage; (twenty two); Where: and Respectively represent The lower and upper limits of reservoir outflow flows for each cascade hydropower station.

[0024] S2.3.2: Set power balance and ramp constraints. The specific formulas include: (twenty three); Where: Representative scenes Next The electricity transmitted from wind power and photovoltaic power to hybrid pumped storage during the period; Representative scenes Next The amount of electricity transmitted to the grid by wind power and photovoltaic power during the period; Representative scenes Next The curtailment of wind power and photovoltaic power during the period; (twenty four); (25); (26); Where: Representative scenes The lower cascade hydropower station group is in The total power transmitted to the grid during the period, Representative scenes The lower cascade hydropower station group is in Power abandonment during the period, and Respectively represent The lower and upper limits of the output of each cascade hydropower station, Representative The climbing capacity of each cascade hydropower station; (27); (28); (29); Where: Representative scenes The hybrid pumped storage system purchases electricity from the electricity market. Representative scenes Hybrid pumped storage The amount of electrical energy transmitted to the grid during the period, Representative scenes Hybrid pumped storage Power abandonment during the period, and They represent the lower and upper limits of the hybrid pumped storage output respectively.

[0025] S2.3.3: Set the mutual exclusion constraint between pumping and power generation. The specific formula includes: (30); (31); Where: Representative scenes The upper level of the hybrid pumped storage hydropower station Output during the period.

[0026] S2.3.4: Set the power transmission constraints of the power grid. The specific formula is expressed as: (32); Where: It is the upper limit of the channel capacity for power transmission and grid connection.

[0027] Furthermore, the purpose of constructing and solving the optimization model is to achieve the management of the peak-shaving demand on the grid side and the abandoned power and load loss on the hydropower side in the complex cascade water, wind, solar and pumped storage power system. The peak-shaving demand on the grid side is met by the first optimization goal (minimizing the peak-to-valley difference of the residual load), and the second optimization goal (minimizing the sum of abandoned power and lost load) is met to reduce the complementary operation risk and smooth the fluctuation of new energy. Combined with various constraints (such as hydraulic and power constraints, power balance and ramp constraints, pumping and power generation mutual exclusion constraints, and grid power transmission constraints), it is ensured that the operation mode of each cascade hydropower station and hybrid pumped storage power station in the basin can reasonably utilize the water energy dispatch output and control the water level under the premise of meeting physical and operational constraints. This step can not only effectively meet the peak-shaving demand on the grid side, but also reduce the abandoned power and load loss phenomenon on the hydropower side, and promote the stable operation and reliability of the system. At the same time, the scenario set generated based on the prediction error can effectively characterize the uncertainty of new energy, give full play to the flexible adjustment advantages of hybrid pumped storage, maximize the use of renewable energy such as wind energy and photovoltaics, and achieve environmental benefits. In addition, converting complex nonlinear models into linearized models through linearization processing helps to improve the speed and accuracy of the solution, ensures the efficiency of the system's real-time calculation, and makes scheduling decisions more efficient.

[0028] S3: Model solving and dispatch instruction generation: Linearize the optimization model, convert the original high-dimensional, multi-constrained, multi-variable complex mixed integer nonlinear programming model into a mixed integer linear programming optimization model, and after importing the dispatch scenario data, call the solver to generate the Pareto solution set in combination with the constraint method, and select the solution of the optimization model from the Pareto solution set based on the fuzzy entropy weight method to generate the dispatch instructions for the cascade hydropower stations and hybrid pumped storage power stations in the basin. The specific process is as follows: S3.1: The fuzzy entropy weight method is used to select the solution of the optimization model from the Pareto solution set. The specific steps are as follows: (33); Where: represents the membership value of the minimization objective, Represents the Pareto solution set An optimal solution to the goal, represent The maximum value of all solution sets under the target, represent The minimum value among all solutions under the target; (34); Where: Representative solution The standardized membership values ​​of the solutions, where the maximum value is considered the best solution; Representative table Pareto solution set The weight of the target, Represents the number of targets, represents the number of Pareto solutions, represents the number of Pareto solutions; for Target A solution, for Target A solution.

[0029] The above steps provide the optimal solution for the parameter status of the cascade hydropower station and the optimal solution for the parameter status of the hybrid pumped-storage power station.

[0030] S3.2: The generation of dispatching instructions for cascade hydropower stations and hybrid pumped-storage power stations in the river basin includes: generating specific start-stop states of conventional hydropower units and hybrid pumped-storage units, reservoir water level fluctuation conditions of the hydropower station and power dispatching instructions based on the optimal solution of the parameter states of the cascade hydropower stations and the optimal solution of the parameter states of the hybrid pumped-storage power station; S3.2.1: The optimal solution of the parameter status of the hydropower station includes: output power, water level change status, power generation flow, and abandoned water flow; S3.2.2: The optimal solution of the parameter status of the hybrid pumped-storage power station includes: the pumping status, power generation status, pumping power, power generation power, pumping flow and power generation flow of the pumped-storage power station.

[0031] Furthermore, based on the solution of the optimization model, dispatching instructions for cascade hydropower stations and hybrid pumped-storage power stations in the basin are generated, aiming to achieve accurate and efficient power dispatching. By solving the optimal solution of cascade hydropower stations and hybrid pumped-storage power stations in the basin, the start and stop states of specific conventional hydropower units and hybrid pumped-storage units, the fluctuation of reservoir water levels in hydropower stations and power dispatching instructions are generated, which can reasonably control the operation mode of cascade hydropower units and hybrid pumped-storage units in the basin in different time periods to achieve supply and demand balance.

[0032] Embodiment 2 is an embodiment of the present invention, which provides a hybrid pumped storage optimization dispatching system that takes into account both peak load regulation and risk avoidance, including: Reference Figure 2 , which is a structural schematic diagram of a hybrid pumped storage optimization scheduling system that takes into account both peak load regulation and risk avoidance provided by an embodiment of the present invention. As shown in the figure, the hybrid pumped storage optimization scheduling system 10 that takes into account both peak load regulation and risk avoidance includes a data acquisition module 101, a data processing module 102, a model building module 103, a solution module 104, and a control scheduling module 105. The specific units of each module are as follows: The data acquisition module 101 is used to acquire the historical data of wind power and photovoltaic power output in the cascade basin, the historical data of flow of reservoirs of each cascade hydropower station in the cascade basin, and the historical data of the load of the receiving end power grid; The data processing module 102 is used for data prediction, and performs splitting, scene generation, reduction and combination based on the predicted data to obtain scheduling scene data.

[0033] Optionally, the data processing module 102 includes a data prediction unit 1021 and a scene generation unit 1022, wherein: The data prediction unit 1021 is used to generate initial dispatch data, based on the historical data of wind power and photovoltaic output in the cascade basin, the historical data of flow of each cascade hydropower station reservoir in the cascade basin, and the historical data of the load of the receiving power grid acquired by the data acquisition module 101, to predict the day-ahead forecast data of wind power and photovoltaic output in the cascade basin, the day-ahead forecast data of flow of each cascade hydropower station reservoir in the cascade basin, and the day-ahead forecast data of the load of the receiving power grid within a future set dispatch period (24 hours) as the initial dispatch data; The scenario generation unit 1022 is used to generate scheduling scenario data. It is based on the day-ahead forecast data of wind power and photovoltaic output in the initial scheduling data obtained by the data prediction unit 1021, and uses the Latin hypercube stratified random sampling method combined with its prediction error to simulate and obtain a scenario set with a specific number of wind power and photovoltaic output forecast data. It uses an iterative self-organizing data analysis algorithm to cluster the scenario set of wind power and photovoltaic output forecast data to generate a representative scenario set of wind power and photovoltaic output forecast data that characterizes the uncertainty of new energy and the probability of each scenario; and combines it with the day-ahead forecast data of the flow of the reservoirs of each cascade hydropower station in the cascade basin in the initial scheduling data and the day-ahead forecast data of the load of the receiving power grid to generate scheduling scenario data.

[0034] The model building module 103 is used to establish a hybrid pumped storage optimization model that takes into account both peak load regulation and risk avoidance, and to perform linearization processing to convert the original mixed integer nonlinear programming model into a mixed integer linear programming model.

[0035] Optionally, the model building module 103 includes a model generating unit 1031 and a linearization processing unit 1032, wherein: The model generation unit 1031 is used to establish a hybrid pumped storage optimization model that takes into account both peak load regulation and risk avoidance, wherein the hybrid pumped storage optimization model that takes into account both peak load regulation and risk avoidance includes a first optimization objective, a second optimization objective, and constraints, wherein the first optimization objective is to minimize the peak-to-valley difference of the remaining load, and the second optimization objective is to minimize the sum of the amount of abandoned electricity and the amount of lost load; the constraints include hydraulic and electric power constraints, power balance and ramp constraints, pumping and power generation mutual exclusion constraints, and power transmission constraints of the power grid; The linearization processing unit 1032 is used to perform linearization processing on the nonlinear part of the hybrid pumped storage optimization model that takes into account both peak load regulation and risk avoidance, and convert the original mixed integer nonlinear programming model into a mixed integer linear programming model.

[0036] The solution module 104 generates a Pareto solution set by using a solver combined with a constraint method, and determines a solution of the optimization model in the solution set based on a fuzzy entropy weight method.

[0037] Optionally, the solution module 104 includes a model solution unit 1041 and an optimal solution selection unit 1042, wherein: The model solving unit 1041 is used to call the solver and solve in combination with the constraint method to generate a uniformly distributed Pareto solution set; The optimal solution selection unit 1042 is used to select the optimal solution in the Pareto solution set generated by the model solving unit 1041 through the fuzzy entropy weight method to obtain the optimal solution of the parameter state of the cascade hydropower station and the optimal solution of the parameter state of the hybrid pumped storage power station.

[0038] The control and dispatching module 105 generates dispatching instructions for the cascade hydropower stations and the hybrid pumped-storage power stations in the basin based on the optimal solutions of the parameter states of the cascade hydropower stations and the hybrid pumped-storage power stations obtained by the solution module 104 .

[0039] Example 3, with reference to Figures 3 (a), 3 (b), 3 (c) and 3 (d), is an embodiment of the present invention. Based on the hybrid pumped storage optimization scheduling method and system that takes into account both peak load regulation and risk avoidance proposed in the present invention, it is scientifically demonstrated through simulation experiments and comprehensive effect evaluation. In order to verify the feasibility of the present invention, the day-ahead scheduling operation effect in different seasons was studied. In addition, the optimization and improvement effect of the method proposed in the present invention patent on guiding the operation of the system before and after the installation of the hybrid pumped storage power station was also studied to verify the beneficial effects of the method proposed in the present invention patent and its applicability to the introduction of a hybrid pumped storage power station.

[0040] The cascade hydropower-wind-solar complementary system in a certain river basin is taken as the research object, which includes cascade hydropower stations, wind power stations and photovoltaic power stations. The cascade hydropower stations in the research object are taken as references to construct application examples. GZ power station, MMY power station and DQ power station currently have 4 turbine units. By installing pumped storage units between GZ and MMY power stations, a cascade hybrid pumped storage power station is formed. The specific power station and unit parameters are shown in Table 1.

[0041] The basic water conservancy parameters include: the relationship between water level and reservoir capacity, the relationship between tailwater level and discharge flow, the relationship between power, head and flow, as well as the starting water level setting of each hydropower station reservoir during the four-season scheduling cycle, the fluctuation range of the final water level, the inflow of the top-level hydropower station reservoir, and the flow between reservoirs of each level of connected hydropower stations.

[0042] System parameters include: 24 hours as the dispatching cycle and 1 hour as the time scale. In addition, the hybrid pumped storage power station contains 4 200MW vertical shaft single-stage reversible mixed flow turbine generator sets with a total installed capacity of 800MW.

[0043] Table 1 Basic parameters of cascade hydropower stations .

[0044] The historical data of wind power and photovoltaic power output in the basin in the past three years, the historical flow data of each cascade hydropower station reservoir, and the historical data of the receiving power grid load were collected. The data are all in hourly time steps, with a total of 26,280 time steps.

[0045] Based on the historical data collected by the system, the day-ahead forecast data of wind power and photovoltaic output in the cascade river basin, the day-ahead forecast data of the flow of each cascade hydropower station reservoir in the cascade river basin, and the day-ahead forecast data of the load of the receiving power grid are predicted as the initial dispatching data. The wind power and photovoltaic output forecast data in the obtained initial dispatching data are used as the basis, and the Latin hypercube stratified random sampling method is used in combination with its prediction error to simulate and obtain a scenario set with a specific number of wind power and photovoltaic output forecast data. The iterative self-organizing data analysis algorithm is used to cluster the generated wind power and photovoltaic output scenarios to generate a representative scenario set of wind power and photovoltaic output forecast data that characterizes the uncertainty of new energy for typical days in each season and the probability of each scenario; and the day-ahead forecast data of the flow of each cascade hydropower station reservoir in the cascade river basin and the day-ahead forecast data of the load of the receiving power grid in the initial dispatching data are combined to generate the dispatching scenario data.

[0046] Subsequently, a hybrid pumped-storage optimization model that takes into account both peak load regulation and risk avoidance is established. The peak load regulation demand on the grid side is met through the first optimization objective (minimizing the peak-to-valley difference of the residual load), and the second optimization objective (minimizing the sum of the amount of abandoned power and the amount of lost load) is used to meet the needs of reducing the risk of complementary operation and smoothing the fluctuation of renewable energy, combined with a variety of constraints (such as hydraulic and power constraints, power balance and ramp constraints, mutual exclusion constraints between pumping and power generation, and grid power transmission constraints).

[0047] The Pareto solution set is generated by calling the solver and combining the constraint method, and the fuzzy entropy weight method is used to select the optimal solution in the Pareto solution set. The optimal solutions for the four seasons are shown in Table 2. It can be seen from the results in the table that in all seasons, the values ​​of the first optimization target and the second optimization target of the hybrid pumped storage system are lower than those of the system without hybrid pumped storage. Since the first and second optimization targets are both minimized as the optimal, the hybrid pumped storage optimization scheduling method and system that takes into account peak load regulation and risk avoidance proposed in this patent can achieve the coordinated optimization of the risk of power abandonment and load loss on the hydropower side and the peak load regulation demand on the power grid side, and is suitable for the multi-objective regulation requirements of cascade basins after the introduction of hybrid pumped storage power stations.

[0048] Table 2 Optimization target values ​​under the four-season optimal solution .

[0049] By multiplying the scenarios in each seasonal scenario set by their corresponding probabilities, we obtained the day-ahead dispatch results for four different seasons and discussed them, as shown in Figure 3 (a), Figure 3 (b), Figure 3 (c) and Figure 3 (d). In terms of the overall trend, the power generation plan of the hybrid pumped storage system is consistent with the load trend of the receiving power grid. The hybrid pumped storage system will pump water to store energy at the load valley and generate electricity at the load peak. Taking spring as an example, according to Figure 3 (a), at the two load valleys near time periods 4 and 14 in spring, the hybrid pumped storage system raises the residual load valley by pumping water to fill the valley; and at the two peaks near time periods 11 and 20, the hybrid pumped storage system reduces the peak of the hybrid pumped storage system by generating peak load. The characteristics of summer, autumn and winter are similar to those of spring, so they will not be repeated. In the summer shown in Figure 3 (b) and the autumn shown in Figure 3 (c), the residual load is relatively flat due to the abundant water inflow; while in the spring shown in Figure 3 (a) and the winter shown in Figure 3 (d), the residual load change trend is similar to the load change due to the small water inflow. From the optimization results of the four seasons in Figure 3, it can be seen that the peak-shaving performance of the hybrid pumped storage system in all four seasons is better than that of the non-hybrid pumped storage system. In addition, while meeting the peak-shaving demand, the hybrid pumped storage system reduces the deviation between the actual output and the power generation plan with the help of the flexible adjustment capability of the hybrid pumped storage, effectively reducing the risk of power abandonment and load loss. This proves the effectiveness of the hybrid pumped storage optimization scheduling method proposed in this patent that takes into account both peak-shaving and risk avoidance.

[0050] Table 3 Data table of various indicators under the optimal solution of four seasons .

[0051] According to the results in Table 3, under the hybrid pumped storage optimization scheduling method and system that takes into account both peak load regulation and risk avoidance proposed in the present invention, the comprehensive performance of the hybrid pumped storage system in all scenarios is good, and its power abandonment and load loss are lower than those of the system without hybrid pumped storage, especially in reducing power abandonment. It shows that the hybrid pumped storage optimization scheduling method that takes into account both peak load regulation and risk avoidance can effectively utilize the energy storage characteristics of the hybrid pumped storage and offset the impact of the uncertainty of new energy forecasts. In addition, the solution running time of the present invention is short, the solution efficiency is high, and the optimal day-ahead scheduling plan can be provided in a short time, which is suitable for real-time application in actual power systems. By simultaneously meeting the peak load regulation demand on the power grid side, the risk of power abandonment and load loss on the hydropower side is reduced, and the flexible adjustment advantage of the hybrid pumped storage is fully utilized to ensure the safety, stability and reliability of the power system.

Claims

1. A hybrid pumped storage optimization dispatching method that takes into account both peak load regulation and risk avoidance, characterized in that: The following steps are involved: Step 1: Data processing: Obtain historical data of cascade basin hydropower, wind and solar power stations and grid loads, generate initial dispatch data through prediction, and generate dispatch scenario data based on the initial dispatch data; Step 2: Construct a hybrid pumped storage optimization model that takes into account both peak load regulation and risk avoidance: determine the first optimization goal by minimizing the peak-to-valley difference of the remaining load to meet the peak load regulation demand on the grid side; determine the second optimization goal by minimizing the sum of the amount of abandoned power and the amount of lost load to meet the needs of reducing the risk of complementary operation and smoothing the fluctuations of wind power and photovoltaic power, and set corresponding constraints; Step 3: Model solving and dispatch instruction generation: Linearize the hybrid pumped-storage optimization model that takes into account both peak load regulation and risk avoidance, and convert the original high-dimensional, multi-constraint, multi-variable complex mixed integer nonlinear programming model into a mixed integer linear programming model. After importing the dispatch scenario data, call the solver to generate a Pareto solution set in combination with the constraint method, and determine the solution of the optimization model in the Pareto solution set based on the fuzzy entropy weight method to generate dispatch instructions for the basin cascade hydropower stations and hybrid pumped-storage power stations.

2. The hybrid pumped storage optimization scheduling method taking into account both peak load regulation and risk avoidance according to claim 1 is characterized in that: The step 1 is specifically as follows: Step 1.1: Obtain historical data of hydropower, wind and solar power stations and grid loads in cascade basins, including historical data of wind power and photovoltaic output in cascade basins, historical flow data of reservoirs of cascade hydropower stations in cascade basins, and historical data of grid loads at receiving ends; Step 1.2: Forecast the day-ahead forecast data of wind power and photovoltaic power output in the cascade basin, the day-ahead forecast data of the flow of each cascade hydropower station reservoir in the cascade basin, and the day-ahead forecast data of the load of the receiving power grid as the initial dispatching data; Step 1.3: Take the wind power and photovoltaic output forecast data in the obtained initial dispatch data as the basis, assume that the wind power and photovoltaic output forecast errors obey the normal distribution and are independent, and and Respectively represent The forecast error distribution of wind power and photovoltaic output in different time periods; Indicates that the variables in the formula are normally distributed; For the The wind power output forecast error of the time period, For the PV output prediction error for each time period; The forecast error of wind power and photovoltaic combined output in period t is expressed as: ； Subsequently, the Latin hypercube stratified random sampling method is used in combination with its prediction error to simulate and obtain a scenario set with a specific number of wind power and photovoltaic output forecast data, and the iterative self-organizing data analysis algorithm is used to cluster the generated scenario set of wind power and photovoltaic output forecast data to generate a representative scenario set of wind power and photovoltaic output forecast data that characterizes the uncertainty of new energy and the probability of each scenario; and it is combined with the day-ahead forecast data of the flow of each cascade hydropower station reservoir in the cascade basin in the initial scheduling data and the day-ahead forecast data of the load of the receiving power grid to generate scheduling scenario data.

3. The hybrid pumped storage optimization scheduling method taking into account both peak load regulation and risk avoidance according to claim 2 is characterized in that: In step 2, the first optimization objective is obtained by minimizing the peak-to-valley difference of the residual load, and the formula is expressed as: ； ； Where: Represents the peak-to-valley difference of the residual load, subscript represents a certain period of time during the scheduling period. , Represents the total time of the entire scheduling period; Representative The remaining load of the period, Represents the original load of the power grid, Representative Power generation plan for each time period; The second optimization objective is obtained by minimizing the sum of the amount of abandoned power and the amount of lost load, and the formula is expressed as: ； ； ； Where: Represents a cascade hydropower station in the basin. Represents a collection of cascade hydropower stations in the basin, ; Represents the total amount of power loss and load loss, Representative scene, A collection of scenes representing each season. , Representative scenes The corresponding probability is Representative scenes Next The amount of power wasted during the period, Representative scenes Next The amount of load loss during the period, Representative scenes Next The difference between the power generation plan and the planned output during the period, Representative scenes Next Wind power and photovoltaic power generation are combined during this period. Representative scenes Next The power provided by wind power and photovoltaic power to the hybrid pumped storage hydropower in the period, Representative scenes Next Cascade hydropower station Output during the period, Representative scenes All cascade hydropower stations are under The period is the power provided by the hybrid pumped storage. Representative scenes Hybrid pumped storage Output during the period.

4. The hybrid pumped storage optimization scheduling method taking into account both peak load regulation and risk avoidance according to claim 3 is characterized in that: In step 2, the constraints include hydraulic and electric power constraints, power balance and ramp constraints, pumping and power generation mutual exclusion constraints, and power transmission constraints of the power grid; The hydraulic and electrical constraints are expressed as follows: ； ； ； ； ； Where: and represent the water-to-electricity conversion efficiency and the electricity-to-water conversion efficiency, respectively. represents the density of water, represents the acceleration due to gravity, Representative scenes Next Cascade hydropower station The net head height of the time period, Representative scenes Hybrid Pumped Storage The net head height of the time period, Representative scenes Next Cascade hydropower station The power generation flow during the period, Representative scenes Hybrid Pumped Storage The power generation flow during the period, Representative scenes Hybrid Pumped Storage Pumping flow rate during the period, and Representing scenes Next Cascade hydropower station The reservoir water level and the reservoir tailwater level during the period, and Representing scenes Hybrid Pumped Storage The reservoir water level of the upper hydropower station and the reservoir water level of the lower hydropower station during the period; Representative scenes Hybrid Pumped Storage Total pumping power during the period; ； ； ； ； ； Where: represents a set of hydropower stations, which includes conventional hydropower stations, upper-level hydropower stations corresponding to hybrid pumped storage, and lower-level hydropower stations corresponding to hybrid pumped storage. Represents a single hydroelectric power station within it; and They represent the reservoir capacities of the upper and lower hydropower stations corresponding to the hybrid pumped storage system. Representative scenes Next Cascade hydropower station The reservoir capacity during the period, Representative scenes Next Hydropower Station No. Reservoir inflow during the period, Representative scenes Next Hydropower Station No. The natural water flow of the reservoir during the period, Representative scenes Next Hydropower Station No. The outflow of the reservoir during the period, Representative scenes Next Hydropower Station No. The power generation flow during the period, Representative scenes Next Hydropower Station No. The amount of water abandoned during the period; and Representing scenes Next The natural water inflow from the reservoirs of the upstream and downstream hydropower stations corresponding to the period-time hybrid pumped storage; and Representing scenes Next The outflow of the reservoirs of the upper and lower hydropower stations corresponding to the hybrid pumped storage in the time period; Scenario Next Cascade hydropower station Natural water inflow from reservoirs during certain periods of time; Scenario Next Cascade hydropower station Reservoir outflow during the period; Δ t Indicates that from t The time difference between the first period and the next period, i.e., the time difference between the t+1 period; ； Where: and Representing scenes Next The reservoir water level at the beginning of the dispatch period and the reservoir water level at the end of the dispatch period for each cascade hydropower station. and Respectively represent The reservoir water level at the beginning of the dispatch period and the reservoir water level at the end of the dispatch period for each cascade hydropower station. and Respectively represent The permissible fluctuation range of reservoir water level at the beginning and end of the dispatch period of each cascade hydropower station; ； ； ； Where: and They represent the nonlinear fitting coefficients of the water level-storage capacity and tailwater level-discharge flow of the hydropower station reservoir, and Representing scenes Next Cascade hydropower station The average reservoir storage capacity and average reservoir discharge during the period, and Respectively represent Dead water level and normal operating water level of each cascade hydropower station reservoir; Z n,i,t Representative scenes Next Cascade hydropower station Reservoir water level during the period; ； ； Where: and Respectively represent Maximum power generation flow of cascade hydropower stations and hybrid pumped storage; ； Where: and Respectively represent The lower and upper limits of the outflow flow from the reservoirs of each cascade hydropower station; The power balance and climbing constraint formula is expressed as: ； Where: Representative scenes Next The electricity transmitted from wind power and photovoltaic power to hybrid pumped storage during the period; Representative scenes Next The electricity transmitted to the grid by wind power and photovoltaic power during the period, Representative scenes Wind power and photovoltaic Power abandonment during certain time periods; ； ； ； Where: Representative scenes The lower cascade hydropower station group is in The total power transmitted to the grid during the period, Representative scenes The lower cascade hydropower station group is in Power abandonment during the period, and Respectively represent The lower and upper limits of the output of each cascade hydropower station, Representative The climbing capacity of each cascade hydropower station; ； ； ； Where: Representative scenes The hybrid pumped storage system purchases electricity from the electricity market. Representative scenes Hybrid pumped storage The amount of electrical energy transmitted to the grid during the period, Representative scenes Hybrid pumped storage Power abandonment during the period, and They represent the lower and upper limits of the hybrid pumped storage output respectively; The formula for the mutual exclusion constraint between pumping and power generation is expressed as: ； ； Where: Representative scenes The upper level of the hybrid pumped storage hydropower station Output during the time period; The formula of the power transmission constraint of the power grid is expressed as: ； Where: It is the upper limit of the channel capacity for power transmission and grid connection.

5. The hybrid pumped storage optimization scheduling method taking into account both peak load regulation and risk avoidance according to claim 1 is characterized in that: Step 3 is as follows: Step 3.1: Determine the solution of the optimization model in the Pareto solution set based on the fuzzy entropy weight method. The specific steps are as follows: ； Where: represents the membership value of the minimization objective, Represents the Pareto solution set An optimal solution to the goal, represent The maximum value of all solution sets under the target, represent The minimum value among all solutions under the target; Calculate the standardized membership value : ； Where: represents the standardized membership value, and its maximum value is regarded as the best solution; Represents the Pareto solution set The weight of the target, Represents the number of targets, represents the number of Pareto solutions; for The kth solution under the target, for Target A solution; Through the above steps, the optimal solution of the parameter state of the cascade hydropower station and the optimal solution of the parameter state of the hybrid pumped storage power station are obtained; Step 3.2: Generate dispatching instructions for cascade hydropower stations and hybrid pumped storage power stations in the basin: Generating dispatching instructions for cascade hydropower stations and hybrid pumped-storage power stations in the river basin, including: generating specific start-stop states of conventional hydropower units and hybrid pumped-storage units, reservoir water level fluctuation conditions of hydropower stations, and power dispatching instructions based on the optimal solution of the parameter states of the cascade hydropower stations and the optimal solution of the parameter states of the hybrid pumped-storage power stations; The optimal solution of the parameter status of the hydropower station includes: output power, water level change status, power generation flow, and abandoned water flow; The optimal solution of the parameter state of the hybrid pumped-storage power station includes the pumping state, power generation state, pumping power, power generation power, pumping flow and power generation flow of the pumped-storage power station.

6. A hybrid pumped storage optimization dispatching system that takes into account both peak load regulation and risk avoidance, characterized in that: include: Data acquisition module (101): used to acquire historical data of wind power and photovoltaic power output in the cascade basin, historical flow data of reservoirs of cascade hydropower stations in the cascade basin, and historical data of load on the receiving end power grid; Data processing module (102): used for data prediction, and splitting, scene generation, reduction and combination based on the predicted data to obtain scheduling scene data; Model building module (103): used to establish a hybrid pumped storage optimization model that takes into account both peak load regulation and risk avoidance, and to perform linearization processing to convert the original mixed integer nonlinear programming model into a mixed integer linear programming model; Solving module (104): generating a Pareto solution set by using a solver combined with a constraint method, and determining a solution of the optimization model in the solution set based on a fuzzy entropy weight method; A control and dispatching module (105) generates dispatching instructions for the cascade hydropower stations and the hybrid pumped-storage power stations in the basin based on the optimal solutions for the parameter states of the cascade hydropower stations and the optimal solutions for the parameter states of the hybrid pumped-storage power stations obtained by the solution module (104).

7. The hybrid pumped storage optimization dispatching system taking into account both peak load regulation and risk avoidance according to claim 6 is characterized in that: The data processing module (102) comprises a data prediction unit (1021) and a scene generation unit (1022); The data prediction unit (1021) is used to generate initial dispatching data, based on the historical data of wind power and photovoltaic power output in the cascade basin, the historical data of flow of reservoirs of cascade hydropower stations in the cascade basin, and the historical data of load of the receiving-end power grid acquired by the data acquisition module (101), to predict the day-ahead forecast data of wind power and photovoltaic output in the cascade basin, the day-ahead forecast data of flow of reservoirs of cascade hydropower stations in the cascade basin, and the day-ahead forecast data of load of the receiving-end power grid within a future set dispatching period, as the initial dispatching data; The scenario generation unit (1022) is used to generate scheduling scenario data. Based on the day-ahead forecast data of wind power and photovoltaic output in the initial scheduling data obtained by the data prediction unit (1021), a Latin hypercube stratified random sampling method is used in combination with its prediction error to simulate and obtain a scenario set with a specific number of wind power and photovoltaic output forecast data. The generated scenario set of wind power and photovoltaic output forecast data is clustered using an iterative self-organizing data analysis algorithm to generate a representative scenario set of wind power and photovoltaic output forecast data that characterizes the uncertainty of new energy and the probability of each scenario. The scenario set is combined with the day-ahead forecast data of the flow of the reservoirs of each cascade hydropower station in the cascade basin and the day-ahead forecast data of the load of the receiving power grid in the initial scheduling data to generate scheduling scenario data.

8. The optimization scheduling system according to claim 6, characterized in that: The model building module (103) includes a model generation unit (1031) and a linearization processing unit (1032); The model generation unit (1031) is used to establish a hybrid pumped storage optimization model that takes into account both peak load regulation and risk avoidance, wherein the hybrid pumped storage optimization model that takes into account both peak load regulation and risk avoidance includes a first optimization target, a second optimization target and constraint conditions, the first optimization target is to minimize the peak-to-valley difference of the remaining load, and the second optimization target is to minimize the sum of the amount of abandoned electricity and the amount of lost load; the constraint conditions include hydraulic and electric power constraints, power balance and ramp constraints, pumping and power generation mutual exclusion constraints, and power transmission constraints of the power grid; The linearization processing unit (1032) is used to perform linearization processing on the nonlinear part of the hybrid pumped storage optimization model that takes into account both peak load regulation and risk avoidance, and convert the original mixed integer nonlinear programming model into a mixed integer linear programming model.

9. The optimization scheduling system according to claim 6, characterized in that: The solution module (104) comprises a model solution unit (1041) and an optimal solution selection unit (1042); The model solving unit (1041) is used to call the solver and solve in combination with the constraint method to generate a uniformly distributed Pareto solution set; The optimal solution selection unit (1042) is used to select the optimal solution in the Pareto solution set generated by the model solving unit (1041) through a fuzzy entropy weight method, so as to obtain the optimal solution for the parameter state of the cascade hydropower station and the optimal solution for the parameter state of the hybrid pumped storage power station.

Images