Short-term water-wind-solar-energy-storage combined dispatching method considering variable-speed pumped storage unit

Abstract

The invention relates to the field of multi-energy complementary operation, in particular to a short-term water, wind and light storage combined dispatching method considering a variable-speed pumped storage unit. According to the technical scheme, the method comprises the steps that firstly, through combined application of a virtual power plant, Latin hypercube sampling and a synchronous back substitution reduction algorithm, uncertain factors in distributed wind and light power plant scheduling are quantized, and a series of typical deterministic scene combinations are generated; then, based on a variable-speed pumped storage unit strategy, a day-ahead output plan of the unit is flexibly adjusted within an allowable output amplitude variation range, and the peak-valley difference of a power grid load is reduced; and finally, by taking the minimum starting number of the units as a target, guiding management of a pumped storage multi-station centralized control center, realizing unit combination optimization among stations, and reserving necessary standby capacity units for a power grid. The method can accurately and effectively guide the management of a large-scale multi-plant-station pumped storage power station, solves the short-term peak regulation problem of water-wind-light-storage combined dispatching, and has relatively high practicability.

Description

Technical Field

[0001] The present invention relates to the field of multi-energy complementary operation, and in particular to a short-term water-wind-solar-storage joint scheduling method considering variable-speed pumped storage units. Background Art

[0002] Pumped-storage power stations (hereinafter referred to as pumped-storage power stations) play an important role in peak-to-valley regulation and promoting clean energy consumption with their continuously growing installed capacity, flexible two-way regulation performance and green and environmentally friendly operation mode, and have become a key means to solve the problem of time-space mismatch between load demand and new energy power generation. However, with the accelerated formation of a new power system with new energy as the main body, its inherent randomness and uncertainty have brought more severe challenges to the short-term joint dispatch of renewable energy. First, the inherent volatility and intermittency of distributed wind and solar power stations are aggravated by their dispersion, miniaturization and large number, making it difficult to achieve accurate prediction and control during the grid connection process, which has a huge impact on the stability of the power grid. Secondly, traditional fixed-speed units are limited by the rated pumping power, and it is difficult to flexibly track the changes in grid load from time to time, which greatly limits the regulation capacity of pumped-storage power stations. Thirdly, when multiple plants and stations of pumped-storage power stations are managed in a unified manner, the unit combination problem is often ignored, resulting in unnecessary unit startup during the dispatch process, resulting in waste of power equipment and capacity, and posing a potential threat to the operation safety of power stations and power grids.

[0003] In summary, these problems have brought great difficulties to the short-term dispatch optimization management and multi-energy joint modeling among multiple plants in large-scale pumped storage power stations. At present, most of the relevant research results and literature reports in China are aimed at fixed-speed units in small-scale, single pumped storage power stations, and there is a lack of research on the optimal dispatch of multi-plant pumped storage control centers with variable speed units. Taking the short-term water-wind-solar-storage joint dispatch in the new power system as the research background, the pumped storage multi-plant control center (P1~P6) in a southeastern province, the hydropower stations H1 and H2 in a large river basin in the southwest that transmits electricity to the province, and the surrounding distributed wind power and photovoltaic power stations are taken as research objects. A short-term water-wind-solar-storage joint dispatch method considering variable speed pumped storage units is proposed. The method of the present invention can minimize the peak-to-valley difference of the power grid load while ensuring that the number of units started during the dispatch process is minimized, and it has strong practicality and wide promotion value. Summary of the invention

[0004] The technical problem to be solved by the present invention is to provide a short-term water-wind-solar-storage joint scheduling method taking into account variable-speed pumped storage units, which fully meets physical constraints or set conditions such as power grid operation constraints, reservoir operating conditions, power station operating conditions, and unit operation requirements, and then calculates the unit combination, startup status, and unit output to achieve the goal of minimizing the peak-to-valley difference.

[0005] The technical solution of the present invention is:

[0006] A short-term hydro-wind-solar-storage joint dispatch method considering variable-speed pumped storage units first quantifies the uncertainties in the dispatch of distributed wind and solar power plants through the joint application of virtual power plants, Latin hypercube sampling and synchronous back-reduction algorithms, and generates a series of typical deterministic scenario combinations. Then, based on the variable-speed pumped storage unit strategy, the day-ahead output plan of the unit is flexibly adjusted within the allowable output range to reduce the peak-to-valley difference of the power grid load. Finally, with the minimum number of unit starts as the goal, the management of the centralized control center of the pumped storage multi-station is guided, and the optimization of the unit combination between stations is realized, and the necessary spare capacity units are reserved for the power grid. The specific steps include:

[0007] (1) Through the virtual power plant strategy, distributed wind and solar power stations with different generating capacities in the same region that are connected to the same main power grid are integrated to obtain a virtual wind power plant and photovoltaic power plant respectively, which are then uniformly controlled and dispatched;

[0008] (2) The uncertainty of the virtual wind and solar power plant output in step (1) is simulated and quantified by the Latin hypercube sampling algorithm to generate a large number of deterministic scenarios that meet the characteristics of wind and solar power output;

[0009] (3) Using the synchronous back-substitution reduction algorithm, the massive deterministic scenarios in step (2) are feature mined to reduce the representative wind and solar typical output processes under different scenario probabilities;

[0010] (4) The typical wind-solar scenarios with different probabilities in step (3) are freely combined to obtain the wind-solar joint scenario with coupled output correlation as the model input to quantify the uncertainty of the wind-solar joint output;

[0011] (5) Construct the objective function of the joint scheduling model.

[0012] In order to improve the management level of the centralized control center of the pumped storage multi-station and reserve more spare unit capacity for the power grid, the objective function of the short-term hydro-wind-solar-storage joint scheduling model, in addition to the very common peak-to-valley difference of the grid residual load and the maximum power generation revenue in engineering, proposes the minimum number of pumped storage units in combination with engineering practice to form a multi-objective model. When the peak-shaving resources are sufficient, the number of unnecessary units in operation can be reduced, the unit loss can be reduced, and the spare capacity can be increased.

[0013] (6) Establish operating constraints for hydropower stations and pumped-storage power stations.

[0014] Due to the rated pumping power, the pumping conditions of traditional fixed-speed pumped storage units can only be controlled by shutdown or full-load operation, which greatly limits the regulation capacity of pumped storage power stations and the peak regulation target results are poor. Based on this, by comprehensively considering the coordinated operation of multiple factors such as head, guide vane opening, unit speed, output, etc. during the dispatching process, the present invention also proposes a variable speed unit strategy that considers the power step range, and controls the pumping speed range of the variable speed unit within ±10%, so as to ensure safe operation while flexibly tracking the load fluctuations of the power grid from time to time, leaving a stable output process for the base load power supply.

[0015] (7) Solving multiple objectives and selecting the optimal solution.

[0016] For the multi-objective problem of the short-term hydro-wind-solar-storage joint scheduling model, that is, the model in step (5) and the constraints in step (6), the non-dominated sorting genetic algorithm II is used to solve it, and the Pareto solution set under different weights is obtained. Then, the optimal solution is selected through the membership function to obtain the final scheduling solution.

[0017] The beneficial effects of the present invention are as follows: In view of the increasing flexibility requirements of the power grid in the new power system, the present invention proposes a short-term water, wind, light and storage joint scheduling method that takes into account variable-speed pumped storage units, constructs a multi-objective joint scheduling model, and optimizes the regulation performance and unit combination problem of the pumped storage power station. The method combines distributed wind and solar power plants in a virtualized manner, quantifies uncertainty through probability simulation and feature clustering, obtains a deterministic typical input scenario, and combines the actual operation of the pumped storage power station to convert the variable speed unit strategy and the principle of minimum number of unit starts into a fixed mathematical expression as the key constraint and objective function of the joint scheduling model, which significantly reduces the peak-to-valley difference of the power grid load, and also improves the utilization rate of multi-station pumped storage unit resources, reserving sufficient standby scheduling capacity for the power system. Compared with the prior art, the present invention can accurately and effectively guide the management of large-scale multi-station pumped storage power stations, solve the short-term peak-shaving problem of water, wind, light and storage joint scheduling, and has strong practicality. BRIEF DESCRIPTION OF THE DRAWINGS

[0018] Figure 1 It is the overall solution framework diagram.

[0019] Figure 2 This is the result diagram of uncertainty processing of virtual wind power plant.

[0020] Figure 3 It is the result graph of multi-objective Pareto solution set and membership function value.

[0021] Figure 4 This is the load distribution result diagram of the variable speed units in the P1 power station during the dry season.

[0022] Figure 5 It is a schematic diagram of the output process and residual load of each power source in different scenarios during the dry season.

[0023] Figure 6 It is a schematic diagram of the output process and residual load of each power source in different scenarios during the flood season. DETAILED DESCRIPTION

[0024] The specific implementation of the present invention is further described below in conjunction with the accompanying drawings and technical solutions.

[0025] like Figure 1 As shown, a short-term water-wind-solar-storage joint scheduling method considering variable-speed pumped storage units of the present invention mainly includes short-term joint scheduling modeling of multi-station pumped storage power stations, solution process and practical application:

[0026] 1. Short-term joint dispatch modeling of multi-station pumped storage power stations

[0027] (1) Objective function

[0028] Objective function 1: Minimize the peak-to-valley difference of the grid residual load

[0029]

[0030] Where: f1 is the peak-to-valley difference of the residual load of the power grid, MW; T is the total number of optimized time periods; C t ' is the remaining load of the power grid during period t, MW; C t is the initial load of the power grid, MW; n, m, k, g are the numbers of hydropower stations, wind power stations, photovoltaic power stations and pumped storage power stations respectively; They are the output of hydropower station, wind power station, photovoltaic power station and pumped storage power station in time period t, MW.

[0031] Objective function 2: Maximizing power generation revenue on the power supply side

[0032]

[0033] Where: f2 is the power generation income on the power supply side; R t is the on-grid electricity price during period t, $ / MWh.

[0034] Objective function 3: Minimize the number of pumped storage units started

[0035]

[0036] Where: f3 is the number of pumped storage units in operation during the dispatching day; u g,i,t is the operating status of the i-th unit of the g-th pumped-storage power station in period t, 1 represents power generation, 0 represents shutdown, and -1 represents pumping.

[0037] (2) Constraints

[0038] 2.1 Hydropower Station Constraints

[0039] (a) Water balance constraints

[0040]

[0041] Where: V n,t+1 、V n,t are the final storage capacity and initial storage capacity of hydropower station n in period t, respectively, 3 ; is the inflow and outflow of hydropower station n in time period t, m 3 / s; is the interval flow of hydropower station n in time period t, m 3 / s; is the number of water flow delay periods of the u-th hydropower station, h; The power generation flow and abandoned water flow of hydropower station n in time period t, m 3 / s, Δt is the calculation period, h.

[0042] (b) Water level constraints on the dam

[0043]

[0044] Where: Z n,t is the water level above the dam of hydropower station n in time period t, Z n,t and is the minimum and maximum water level on the dam of hydropower station n in time period t, m; z n is the initial water level of hydropower station n, m; ξ H It is the control range of water level on the dam of hydropower station within the dispatching day, %.

[0045] (c) Outbound flow constraints

[0046]

[0047] Where: is the allowable outflow of hydropower station n in time period t, is the minimum and maximum outflow allowed at hydropower station n during time period t, m 3 / s; is the tailwater level of hydropower station n in time period t, m; f n ZQ (·) is the tailwater level discharge function relationship of hydropower station n

[0048] (d) Output constraints

[0049]

[0050] Where: is the output of hydropower station n in time period t, is the guaranteed output and installed capacity of hydropower station n in time period t, MW; ε n,t is the water consumption rate of hydropower station n in time period t, m 3 / KWh.

[0051] (e) Climbing capability constraints

[0052]

[0053] Where: is the maximum allowable output amplitude of hydropower station n, MW.

[0054] (f) Power generation flow constraints

[0055]

[0056] Where: q n,t is the allowable power generation flow of hydropower station n in time period t, are the minimum and maximum power generation flows allowed by hydropower station n in time period t, m 3 / s.

[0057] (g) Head constraint

[0058]

[0059] Where: is the net water head of hydropower station n in time period t, m; is the head loss of hydropower station n in period t, m; is the functional relationship between the head and water consumption rate of hydropower station n.

[0060] 2.2 Constraints of Pumped Storage Power Stations

[0061] (a) Power generation flow constraints

[0062]

[0063] Where: are the upper and lower storage capacities of the pumped storage power station g in time period t, m 3 ; is the inflow flow of the pumped storage power station g in time period t, m 3 / s.

[0064] (b) Water level constraints

[0065]

[0066]

[0067] Where: is the water level of the upper and lower reservoirs of the pumped storage power station g in time period t, m; and are the minimum and maximum values ​​of the upper reservoir water level of the pumped storage power station g in time period t, respectively; and are the minimum and maximum values ​​of the lower reservoir water level of the pumped storage power station g in time period t, respectively; are the initial upper and lower reservoir water levels of the pumped storage power station g on the actual dispatching day, m; and are the initial water level and the end-of-day water level of the upper and lower reservoirs respectively; They are the control range of water level in the upper and lower reservoirs of the pumped-storage power station g during the dispatching day, %.

[0068] (c) Unit operating condition constraints

[0069] In actual operation, large pumped-storage power plants must follow a series of strict operating status restrictions. First, the units must experience a specified downtime when switching between power generation and pumping modes. Secondly, different units of the same pumped-storage power plant shall not perform pumping and power generation operations at the same time during the same time period. Furthermore, different pumped-storage power plants in the same area shall not carry out pumping and power generation activities at the same time during the same time period. The purpose of these restrictions is to ensure the efficient and orderly dispatching of multiple pumped-storage power plants, prevent conflicts in the unit dispatching process, and ensure the stable operation of the power system.

[0070]

[0071] Where: Ω i are the remaining units in the pumped storage power station g except unit i; Ω g The remaining pumped-storage power stations except pumped-storage power station g.

[0072] (d) Unit flow constraints

[0073]

[0074] Where: is the power generation and pumping flow of the pumped storage power station g in time period t, m 3 / s.

[0075] (e) Variable speed output constraints

[0076]

[0077] Where: is the power generation and pumping power of the i-th unit of the pumped-storage power station g in time period t, MW; δ is the pumping power amplitude of the variable speed unit, which is 10%.

[0078] (f) Net head constraint

[0079]

[0080] Where: is the water head of the pumped storage power station g in time period t, m.

[0081] (g) Unit characteristic curve

[0082]

[0083] Where: is the power generation and pumping characteristic curve of the i-th unit of the pumped-storage power station g.

[0084] 2. Solution process of short-term joint dispatch model of multi-station pumped storage power stations

[0085] (1) The distributed wind and solar power stations connected to the same main power grid are integrated through the virtual power plant strategy to obtain the virtual wind farm S and photovoltaic power plant W.

[0086] (2) Based on step (1), the uncertainty of the virtual wind and solar power plant is simulated and quantified through Latin hypercube sampling to generate a large number of deterministic output scenarios that meet the operating characteristics.

[0087] (3) Based on step (2), feature mining is performed on massive scenes through a synchronous back-substitution reduction algorithm to extract a fixed number of representative landscape typical scenes.

[0088] (4) Based on step (3), the representative wind-solar scenarios with different probabilities are freely combined to obtain the wind-solar joint output with coupling correlation as the input scenario.

[0089] (5) Construct a joint scheduling model.

[0090] (6) Solving multiple objectives and selecting the optimal solution.

[0091] For the multi-objective optimization problem of the short-term hydro-wind-solar-storage joint scheduling model, the non-dominated sorting genetic algorithm II (NSGA-II) is used to solve it, and the appropriate algorithm parameters are selected: population size, number of iterations, crossover probability, mutation probability, and evolutionary method, etc., and the Pareto solution set under multiple weights is successfully obtained. Since different weights correspond to the optimal solutions under different decisions, in order to screen out the best solution suitable for the scheduling model in the present invention from the Pareto solution set, the membership function is used to further analyze the Pareto solution set. In this way, the optimal solution can be selected as the actual scheduling strategy to better meet the needs of scheduling decision makers. The expression of the membership function is shown in formula (17):

[0092]

[0093] Where: x v,b is the bth Pareto solution set of the vth objective function; u v.b (x) is the membership value of the b-th Pareto solution set of the v-th objective function; and is the maximum and minimum value in the solution set of the vth objective function; u b (x) is the normalized result of the membership value of the b-th Pareto solution set.

[0094] (7) After completing all uncertainty analysis, multi-objective function construction, and constraint processing, the newly constructed short-term joint dispatch model is optimized and solved based on the complementary problem of hydropower, wind power, and solar power storage in multiple pumped storage stations. The detailed process is as follows: Figure 1 shown.

[0095] 3. Practical application of short-term joint dispatch model for multi-station pumped storage power stations

[0096] Now, a provincial day-ahead power grid load with multiple peaks and huge peak-to-valley differences in southeastern my country is taken as an implementation case, and the method of the present invention is verified by taking 6 large PSHPs (P1, P2, P3, P4, P5, P6) of the province's pumped storage multi-station centralized control center, and 2 cascade hydropower stations (H1, H2) and virtualized wind (W) and solar power plants (S) in a large clean energy base in the southwest that transmits electricity to the province as objects. The basic parameters of each power station are shown in Table 1, the basic parameters of the pumped storage units are shown in Table 2, and the grid-connected electricity price adopts the peak-to-valley electricity price provided by the dispatching agency, see Table 3. The model of the present invention is named M3, and the conventional comparison model is set as the control group in this embodiment, named M1 and M2, where M1 is a traditional short-term joint dispatch model of water, wind, and light without considering the pumped storage power station, and M2 is a short-term joint dispatch model of water, wind, and light without considering the variable speed unit constraints and the goal of minimizing the number of unit startups.

[0097] Table 1

[0098]

[0099] Table 2

[0100]

[0101] Table 3

[0102]

[0103] The uncertainty of W and S of the virtual wind power plant is quantified by Latin hypercube sampling and synchronous back-substitution elimination algorithm, and the Figure 2The deterministic typical input scenarios with different characteristics are shown. Through free combination, 16 wind-solar joint scenarios are formed as model input. The model of the present invention and the comparative model are solved by using the non-dominated sorting genetic algorithm II and the membership function, and the Pareto solution set under different weights is selected. Figure 3 The optimal scheduling scheme for flood season and dry season is shown.

[0104] During the dry season, due to the small difference between the peak and valley of the power grid load, not all of the 31 pumped storage units in this embodiment need to be started to achieve effective peak load regulation, so there is a large room for optimization in the start and stop arrangements of the units in the centralized control center. According to the data analysis in Table 4, under the same peak load regulation target, the M3 model reduces the number of unit starts by three compared to the M2 model due to the optimization target of the minimum number of unit starts, thereby reserving the necessary spare capacity for the power system. Taking the unit combination optimization results of the P1 power station in scenario 1 as an example, Figure 4 It shows that the number of units started in M3 is 7, 2, 6, 2, 2 and 2, a total of 21 units, while the M2 model is 5, 3, 5, 4, 4 and 3, a total of 24 units. In addition, under the constraint of ±10% pumping power variation range, the unit output process of the M3 model is also clearly displayed. Compared with the fixed pumping power of the M2 model, the M3 model can track the fluctuation of the grid load more accurately. At the same time, the changes in the upper and lower reservoir water levels during the entire dispatching process are kept within the allowable deviation range of the control water level, ensuring the sustainable dispatching and operation of the pumped storage power station. Figure 5 The output process and peak regulation results of each power source in M3 are shown in scenarios 1, 5, 9, and 13, when the proportion of wind power in the dry season is high. The pumped storage power station provides a smooth residual load process for the base load power source with its flexible two-way regulation capability, ensuring the stability of the grid operation.

[0105] During the flood season, due to the large peak-to-valley difference in the power grid load caused by the peak electricity consumption, all 31 pumped storage units need to be started to balance this huge peak-to-valley difference in load, so there is no room for optimization in the number of units started. However, according to the data in Table 5, although the number of units started is the same, M3 has significantly improved the regulation performance of the pumped storage power station by virtue of the regulation advantage of the variable speed unit. Compared with M2, the peak-to-valley difference in the power grid load has been reduced by 1516MW, and the load factor has been increased by 1.01%, effectively ensuring the stable operation of the power grid. Figure 6 The output process and peak-shaving results of each power source in M3 are shown in scenarios 1, 2, 3, and 4, where photovoltaic power accounts for a relatively high proportion during the flood season. Benefiting from the performance of the variable speed units, M3 can flexibly track the load fluctuations of the power grid from time to time, effectively meeting the flexibility requirements of the power system.

[0106] In summary, compared with the conventional short-term joint dispatch model of water, wind, solar and storage, the method proposed in this invention not only significantly improves the regulation performance of pumped storage power stations, but also optimizes the unit combination of multi-station pumped storage power stations, reduces the number of unnecessary startups, and ensures the stable and safe operation of the power system. This method has important practical application value for the unified dispatch and management of many large-scale pumped storage power stations.

[0107] Table 4

[0108]

[0109] Table 5

[0110]

Claims

1. A short-term hydro-wind-solar-storage joint dispatching method considering variable-speed pumped storage units, characterized in that: The steps include: (1) Through the virtual power plant strategy, distributed wind and solar power stations with different generating capacities in the same region that are connected to the same main power grid are integrated to obtain a virtual wind power plant and photovoltaic power plant respectively, which are then uniformly controlled and dispatched; (2) The uncertainty of the virtual wind and solar power plant output in step (1) is simulated and quantified by the Latin hypercube sampling algorithm to generate a large number of deterministic scenarios that meet the characteristics of wind and solar power output; (3) Using the synchronous back-substitution reduction algorithm, the massive deterministic scenarios in step (2) are feature mined to reduce the representative wind and solar typical output processes under different scenario probabilities; (4) The typical wind-solar scenarios with different probabilities in step (3) are freely combined to obtain the wind-solar joint scenario with coupled output correlation as the model input to quantify the uncertainty of the wind-solar joint output; (5) Construct the objective function of the joint scheduling model; Objective function 1: Minimize the peak-to-valley difference of the grid residual load Where: f1 is the peak-to-valley difference of the residual load of the power grid, MW; T is the total number of optimized time periods; C t ' is the remaining load of the power grid during period t, MW; C t is the initial load of the power grid, MW; n, m, k, g are the numbers of hydropower stations, wind power stations, photovoltaic power stations and pumped storage power stations respectively; are the outputs of hydropower station, wind power station, photovoltaic power station and pumped storage power station in time period t, MW; Objective function 2: Maximizing power generation revenue on the power supply side Where: f2 is the power generation income on the power supply side; R t is the on-grid electricity price during period t, $ / MWh; Objective function 3: Minimize the number of pumped storage units started Where: f3 is the number of pumped storage units in operation during the dispatching day; u g,i,t is the operating status of the i-th unit of the g-th pumped-storage power station in period t, 1 for power generation, 0 for shutdown, and -1 for pumping; (6) Constructing the operation constraints of hydropower stations and pumped storage power stations; (6.1) Hydropower station constraints (a) Water balance constraints Where: V n,t+1 、V n,t are the final storage capacity and initial storage capacity of hydropower station n in time period t, m 3 ; is the inflow and outflow of hydropower station n in time period t, m 3 / s; is the interval flow of hydropower station n in time period t, m 3 / s; is the number of water flow delay periods of the u-th hydropower station, h; The power generation flow and abandoned water flow of hydropower station n in time period t, m 3 / s, Δt is the calculation period, h; (b) Water level constraints on the dam Where: Z n,t is the water level above the dam of hydropower station n in time period t, Z n,t and is the minimum and maximum water level on the dam of hydropower station n in time period t, m; z n is the initial water level of hydropower station n, m; ξ H is the control range of the water level above the dam of the hydropower station within the dispatching day, %; (c) Outbound flow constraints Where: is the allowable outflow of hydropower station n in time period t, is the minimum and maximum outflow allowed at hydropower station n during time period t, m 3 / s; is the tailwater level of hydropower station n in time period t, m; is the tailwater level outflow function relationship of hydropower station n; (d) Output constraints Where: is the output of hydropower station n in time period t, is the guaranteed output and installed capacity of hydropower station n in time period t, MW; ε n,t is the water consumption rate of hydropower station n in time period t, m 3 / KWh; (e) Climbing capability constraints Where: is the maximum allowable output amplitude of hydropower station n, MW; (f) Power generation flow constraints Where: q n,t is the allowable power generation flow of hydropower station n in time period t, are the minimum and maximum power generation flows allowed by hydropower station n in time period t, m 3 / s; (g) Head constraint Where: is the net water head of hydropower station n in time period t, m; is the head loss of hydropower station n in period t, m; is the functional relationship between the head and water consumption rate of hydropower station n; (6.2) Constraints on pumped storage power stations (a) Power generation flow constraints Where: are the upper and lower storage capacities of the pumped storage power station g in time period t, m 3 ; is the inflow flow of the pumped storage power station g in time period t, m 3 / s; (b) Water level constraints Where: is the water level of the upper and lower reservoirs of the pumped storage power station g in time period t, m; and are the minimum and maximum values ​​of the upper reservoir water level of the pumped storage power station g in time period t, respectively; and are the minimum and maximum values ​​of the lower reservoir water level of the pumped storage power station g in time period t, respectively; are the initial upper and lower reservoir water levels of the pumped storage power station g on the actual dispatching day, m; and are the initial water level and the end-of-day water level of the upper and lower reservoirs, respectively; They are the control range of the water level of the upper and lower reservoirs of the pumped storage power station g during the dispatching day, %; (c) Unit operating condition constraints Where: Ω i are the remaining units in the pumped storage power station g except unit i; Ω g are the other pumped-storage power stations except pumped-storage power station g; (d) Unit flow constraints Where: is the power generation and pumping flow of the pumped storage power station g in time period t, m 3 / s; (e) Variable speed output constraints Where: is the power generation and pumping power of the i-th unit of the pumped-storage power station g in time period t, MW; δ is the pumping power amplitude of the variable speed unit; (f) Net head constraint Where: is the water head of the pumped storage power station g in time period t, m; (g) Unit characteristic curve Where: is the power generation and pumping characteristic curve of the i-th unit of the pumped storage power station g; (7) Solving multiple objectives and selecting the optimal solution; For the multi-objective problem of the short-term hydro-wind-solar-storage joint scheduling model, that is, the model in step (5) and the constraints in step (6), the non-dominated sorting genetic algorithm II is used to solve it, and the Pareto solution set under different weights is obtained. Then, the optimal solution is selected through the membership function to obtain the final scheduling solution.

2. A short-term water-wind-solar-storage joint scheduling method considering variable-speed pumped storage units according to claim 1, characterized in that: The expression of the membership function is shown in formula (17): Where: x v,b is the bth Pareto solution set of the vth objective function; u v.b (x) is the membership value of the b-th Pareto solution set of the v-th objective function; and is the maximum and minimum value in the solution set of the vth objective function; u b (x) is the normalized result of the membership value of the b-th Pareto solution set.

Images