A method for day-ahead optimal dispatch of cascade hydro-sightseeing complementary power generation system based on PSA

Abstract

The application discloses a day-ahead optimal scheduling method for a cascade water and wind-solar complementary power generation system based on PSA, relates to the technical field of multi-energy complementary power generation, and first establishes a target function with the minimization of source-load deviation as the target function, then utilizes wind power and photovoltaic predicted output data sets, utilizes a Monte Carlo generation scene+K-means method for prediction, and obtains corresponding wind power and photovoltaic output scenes. According to constraint conditions of each power station, a wind-solar-water combined power generation system model is constructed. By combining the obtained hydropower output planning scheme and the wind-solar prediction curve, the optimal solution of the wind-solar-water combined system is solved by utilizing the PSA algorithm. The application considers the influence of wind power and photovoltaic output randomness on the wind-solar-water system, formulates a day-ahead output planning scheme for the wind-solar-water complementary power generation system, and ensures that the system can stably operate.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of multi-energy complementary power generation, in particular to a day-ahead optimal scheduling technology for cascade water, wind and light complementary power generation system based on PID-based search algorithm (PSA). BACKGROUND

[0002] As the current highest penetration rate of clean energy, wind power and photovoltaic power have the advantages of environmental protection and renewable energy, but their randomness, intermittency and volatility have a serious impact on the power grid, which brings great challenges to the safe operation of the power system and the dispatching of the power grid. In order to realize clean power grid, wind power and photovoltaic power generation station need to take measures to suppress fluctuations before entering the grid to reduce the impact on the power grid. Under this background, cascade hydropower stations become the key solution to multi-energy complementary dispatching on the power supply side due to their strong regulation capacity, effectively suppressing the volatility of wind and light output, improving the consumption efficiency of the power grid, and expected to be an effective way to absorb a large amount of new energy in the long term. Therefore, in the current global background of promoting clean energy development, it is particularly important to accurately analyze the operating characteristics of the water, wind and light complementary power generation system. This comprehensive energy system provides strong support for the stable operation and sustainable development of the power system.

[0003] Currently, the research on multi-energy complementary dispatching mainly focuses on the water, wind and light multi-energy complementary dispatching of single hydropower stations, and there is little research on the multi-energy complementary dispatching of large-scale wind and light grid-connected cascade hydropower stations, which does not meet the actual needs of actively promoting the construction of large-scale water, wind and light complementary bases in China. Cascade hydropower stations can obtain better power generation benefits than single hydropower stations, and have the characteristics of flexibility and stability. The use of cascade hydropower to make up for the intermittency of wind and light power generation can effectively suppress the volatility of wind and light power station output. Therefore, a multi-energy complementary power generation system is constructed, but the complex coupling relationship between different energies needs to be considered. At the same time, the inherent uncertainty and randomness of renewable energy also increase the difficulty of finding the best solution. This coupling and randomness make the feasible solution space of multi-energy complementary dispatching problem extremely complex, so a more in-depth and comprehensive model and solution method is needed to solve it. Traditional optimal scheduling methods have the shortcomings of low efficiency, slow optimization speed and difficulty in guaranteeing convergence to the global optimal solution, etc. The present application adopts a new meta-heuristic algorithm PSA, which has faster convergence speed and higher accuracy. SUMMARY

[0004] The present application provides a day-ahead optimal scheduling method for cascade water, wind and light complementary power generation system based on PSA, which aims to solve the problems of low efficiency, slow optimization speed and difficulty in converging to the best optimization result in the model of multi-energy complementary power generation system.

[0005] The present application is based on the step-by-step water and wind scenery complementary power generation system day-to-day optimization scheduling method, and the steps are as follows:

[0006] Step 1: Adopt Monte Carlo scene generation method+K-means to respectively predict the wind power prediction output under four typical day conditions Photovoltaic prediction output And the load prediction power Among them:

[0007]

[0008] In the formula: Indicates the wind power prediction output at t time; Indicates the photovoltaic prediction output at t time; Indicates the maximum wind power output at t period; Indicates the maximum photovoltaic output at t period; w indicates wind; s indicates photovoltaic;

[0009] Step 2: Establish the constraint conditions of the step-by-step water and wind scenery complementary system, which are respectively the power transmission channel constraint and the step-by-step water power constraint condition, and the constraint conditions are as follows:

[0010] 1) Power transmission channel constraint

[0011]

[0012] In the formula: Indicates the output of the i-th water power station; Indicates the m-th group of wind power prediction output; Indicates the s-th group of photovoltaic prediction output; P d Indicates the channel capacity of the d-th power transmission section; I indicates the number of water power stations; M indicates the number of wind turbine generators; N indicates the number of photovoltaic generators;

[0013] 2) Step-by-step water power constraint condition

[0014] Output constraint

[0015]

[0016] In the formula: Indicates the lower limit of the output of the i-th water power station at t time; Indicates the upper limit of the output of the i-th water power station at t time; Indicates the output of the i-th water power station at t time; wherein A indicates the output coefficient of the step-by-step water power station; Q i,t Indicates the power generation flow of the i-th water power station at t period; H i,t Indicates the water head height of the i-th water power station at t period; h indicates the water power station;

[0017] Load power balance constraint

[0018]

[0019] Step 3: Based on the wind power prediction, the photovoltaic power prediction, the load prediction power and the constraint conditions, a day-ahead optimal dispatching model of the cascade water-wind-solar complementary system is established, with the minimum source-load deviation as the target.

[0020] V i,t+1 = V i,t + [I i,t - (Q i,t + S i,t )] · Δt (Formula Five)

[0021] In the formula, V i,t+1 represents the reservoir capacity of the i-th cascade hydropower station at the t+1 time period, V i,t represents the reservoir capacity of the i-th cascade hydropower station at the t time period; I i,t represents the inflow of the i-th cascade hydropower station at the t time period; Δt represents the calculation time period; S i,t represents the abandoned water flow of the i-th cascade hydropower station at the t time period.

[0022] Hydropower reservoir capacity, water level and flow constraints

[0023]

[0024] In the formula, Y represents the lower limit of the reservoir capacity of the i-th cascade hydropower station at the t time period, represents the upper limit of the reservoir capacity of the i-th cascade hydropower station at the t time period; Y represents the lower limit of the water level of the i-th cascade hydropower station at the t time period, represents the upper limit of the water level of the i-th cascade hydropower station at the t time period; Y i,t represents the water level of the i-th cascade hydropower station at the t time period; represents the lower limit of the power generation flow of the i-th cascade hydropower station at the t time period, represents the upper limit of the power generation flow of the i-th cascade hydropower station at the t time period; represents the initial reservoir capacity of the hydropower station, represents the final reservoir capacity of the hydropower station;

[0025] Step 3: Based on the wind power prediction, the photovoltaic power prediction, the load prediction power and the constraint conditions, a day-ahead optimal dispatching model of the cascade water-wind-solar complementary system is established, with the minimum source-load deviation as the target.

[0026] The objective function is as follows:

[0027]

[0028] In the formula, F represents the source-load deviation value; represents the load prediction power at the t time period; represents the wind power prediction at the t time period; This represents the predicted photovoltaic output at time t; This represents the hydropower output of the i-th level hydropower station at time t; T represents the dispatching period. This represents the predicted wind power output of the m-th wind turbine group at time t; This represents the predicted photovoltaic output of the nth photovoltaic unit at time t;

[0029] Step 4: Optimize the day-ahead scheduling model for the cascade hydro-wind-solar hybrid system;

[0030] Set the population size to B, the number of iterations to A, and the scheduling period to C, with the current iteration number being a; using water level as the decision variable, generate a set of random data representing the water level of the cascade hydropower stations over C hours, and denote it as... in The water level represents that of a first-level hydroelectric power station. This represents the water level of the secondary hydropower station. The water level represents the level of the third-stage hydroelectric power station; determine whether x satisfies the constraints of step 2; if not, continue generating water levels that satisfy the conditions; if they do, form a set of water levels x that satisfy the conditions; finally, form an initial water level population, denoted as X = (X... 1 X 2 X 3 ),in

[0031]

[0032] The X obtained above utilizes the water level and the power generation flow rate Q i,t Relations and Calculate the power output of the hydropower station, and then use the predicted wind power output obtained in step 1. Photovoltaic power forecast and load value Substitute into Formula 7 to calculate the source load deviation value, and record it as...

[0033]

[0034] Find the minimum value F among the B source load deviation values. b1 The corresponding water level x; determine the current iteration number a, if a = 1, then set the current source load deviation value F. b1 Set as target value F b ′1, water level x is set to x′, error e is set to x′-x, if a>1, then the current source load deviation value is compared with the previous target value, and the one that is less is set as the latest target value F. b ′1′, water level is set to x″, error e is x″-x;

[0035] The system deviation is minimized and the control process is optimized by dynamically adjusting the error signal e by PID, the water level population X is updated by formula eight, and whether the updated water level x meets the constraint condition of step 2 is judged, if not, the optimal water level is continuously searched, if yes, the obtained water level x is output. The update strategy is:

[0036] x (t+1) = x (t) + η·Δu (t) + (1-η)·o (t) (Formula eight) whether the current iteration number a is greater than or equal to A, if a < A, continue to optimize, if a > A, terminate optimization, and output the final calculation result source load deviation F b ′1′ and water level x, so as to better track the load curve and achieve the purpose of stabilizing the fluctuation of light and wind;

[0037] The PSA algorithm of the present application is a new type of meta-heuristic algorithm, which makes up for the shortcomings of traditional algorithms such as poor convergence and slow convergence speed; the PSA algorithm has faster convergence speed and higher precision, so that the water, wind and light complementary system obtains a more optimized scheduling method, and the complementary nature and seasonal characteristics of water, wind and light energy are fully utilized to ensure the safe, stable and economic operation of the power system. BRIEF DESCRIPTION OF DRAWINGS

[0038] Figure 1 is a water, wind and light complementary power generation structure diagram, Figure 2 is a model solution flowchart, Figure 3 a is a photovoltaic output prediction curve, Figure 3 b is a wind power output prediction curve, Figure 3 c is a load output prediction curve, Figure 4 is a typical day optimization result column chart, Figure 5 is a comparison result of PSO and PSA convergence curves. DETAILED DESCRIPTION

[0039] In order to make the purpose, technical scheme and advantages of the present application clearer, the present application is further described in detail below in combination with the drawings and examples. The present application provides a PSA-based cascade water, wind and light complementary power generation system day-ahead optimization scheduling method. The operation mode of water, wind and light complementation refers to the joint operation of water power, wind power and photovoltaic three kinds of clean energy, in which wind and light output is mainly used for power supply guarantee, and the output fluctuation of wind and light is adjusted by cascade water power to realize stable energy supply. The water, wind and light complementary optimization model under this mode considers the constraint conditions under various factors, and provides an important reference for the sustainable development and safe operation of the power system. The specific invention steps are as follows:

[0040] Step 1: Adopt Monte Carlo scene generation method + K-means to predict wind power prediction output under four typical days photovoltaic prediction output and load forecasting power wherein:

[0041]

[0042] wherein: Pw(t) represents the wind power prediction output at time t; Ppv(t) represents the photovoltaic prediction output at time t; Pwmax(t) represents the maximum wind power output at time t; Ppvmax(t) represents the maximum photovoltaic power output at time t; w represents wind; s represents photovoltaic;

[0043] The formula of the wind-solar power station model is:

[0044] The wind power output is mainly affected by the wind speed, and its expression is:

[0045]

[0046] wherein: P N Pw represents the wind power rating; v in vci represents the cut-in wind speed; v out vcut represents the cut-out wind speed; v t v(t) represents the wind speed at time t;

[0047] Photovoltaic prediction output

[0048] The photovoltaic power output is mainly affected by the solar irradiance, which is related to the position of the sun, the geographical position of the photovoltaic power generation equipment, and the weather conditions, and its expression is:

[0049]

[0050] wherein: f(x t , a, b) represents the Beta distribution; B(a, b) represents the Beta function; a and b represent the shape parameters; P c Ppv represents the photovoltaic power rating; B g B0 represents the light intensity reference value; G e G0 represents the rated light intensity; x i x(t) represents the light intensity at time t;

[0051] Step 2: Establish the constraint conditions of the cascade water-wind-solar complementary system, which are the power transmission channel constraint and the cascade water power constraint condition, and the constraint conditions are:

[0052] 1) Power transmission channel constraint

[0053]

[0054] wherein: Pw(i) represents the output of the i-th water power station; Pw(m) represents the m-th group of wind power prediction output; P d P

[0055] 2) Step hydropower constraints

[0056] Output constraints

[0057]

[0058] In the formula: P P P i,t Q i,t h

[0059] Load power balance constraints

[0060]

[0061] Step hydropower water balance constraints

[0062] V i,t+1 = V i,t + [I i,t - (Q i,t + S i,t )] · Δt (Equation Seven)

[0063] In the formula: V i,t+1 P i,t P i,t Q i,t S

[0064] Hydropower reservoir capacity, water level, and flow constraints

[0065]

[0066] In the formula: P P h represents the upper limit of the water level of the i-th cascade hydropower station at t time; Y represents the lower limit of the water level of the i-th cascade hydropower station at t time; and i,t represents the water level of the i-th cascade hydropower station at t time. represents the lower limit of the power generation flow of the i-th cascade hydropower station at t time period, represents the upper limit of the power generation flow of the i-th cascade hydropower station at t time period. represents the initial reservoir capacity of the hydropower station. represents the final reservoir capacity of the hydropower station.

[0067] Step 3: Based on the wind power prediction output, the photovoltaic prediction output, the load prediction power and the constraint conditions described in steps 1 and 2, a day-ahead optimization scheduling model of the cascade water-wind-solar complementary system is established, with the minimum source-load deviation as the target.

[0068] The objective function is as follows:

[0069]

[0070] In the formula, F represents the source-load deviation value. represents the load prediction power at t time; represents the wind power prediction output at t time; represents the photovoltaic prediction output at t time; represents the hydropower output of the i-th cascade hydropower station at t time; and T represents the scheduling period. represents the wind power prediction output of the m-th group of wind turbine generators at t time. represents the photovoltaic prediction output of the n-th group of photovoltaic generators at t time.

[0071] Step 4: Optimization of the day-ahead optimization scheduling model of the cascade water-wind-solar complementary system

[0072] The population size B, the iteration number A and the scheduling period C are set, and the current iteration number a is set. Taking the water level as the decision variable, a set of random data representing the C-hour water level of the cascade hydropower station is generated, denoted as X wherein represents the water level of the first cascade hydropower station, represents the water level of the second cascade hydropower station, represents the water level of the third cascade hydropower station. Whether x satisfies the constraint condition of step 2 is judged. If not, continue to generate water levels that satisfy the condition, and if yes, a set of water levels x that satisfy the condition is formed. Finally, an initial water level population is formed, denoted as X=(X 1 X 2 X 3 ), wherein

[0073]

[0074] The X obtained above is used to calculate the power generation flow Q i,t relationship and The output of the hydropower station is calculated, and the wind power predicted output obtained in step 1 is added The photovoltaic predicted output is added And the load value The source-load deviation value is obtained by substituting formula seven, and is recorded as F

[0075]

[0076] The minimum value F of the B source-load deviation values is found b1 The corresponding water level x is found; the current iteration number a is judged, if a = 1, the current source-load deviation value F is set as the target value F b1 ′1, the water level x is set as x′, the error e is set as x′-x, if a > 1, the current source-load deviation value is compared with the last target value, and the smaller one is set as the latest target value F b ′1′, the water level is set as x″, and the error e is x″-x; b

[0077] The output of the PID adjustment is calculated by using the target value error of the complementary power generation system, and the expression is

[0078] Δu(t) = K p ·r2·[e k (t)-e k-1 (t)]+K i ·r3·e k (t)+K d ·r4·[e k (t)-2e k-1 (t)+e k-2 (t)] (Formula Ten)

[0079] In the formula, r2, r3 and r4 represent the vector of random numbers from 0 to 1 in n rows and 1 column; K p , K i and K d respectively represent the adjustment coefficients of proportion, integration and differentiation. Adding a condition factor can make the water-wind-solar complementary system obtain a better scheduling scheme, and the expression is:

[0080] o(t) = (cos(1-t / T)+λr5·L)·e k (t) (Formula Eleven)

[0081] In the formula, r5 represents the vector of random numbers from 0 to 1 in n rows and d columns; and λ is the calculation adjustment coefficient:

[0082] λ = [ln(T-t+2) / ln(T)] 2 (Formula Twelve)

[0083] ​λ decreases with the increase of t, which is beneficial to the PSA algorithm to search the optimal scheduling scheme sufficiently.

[0084]

[0085] In the formula, u and v represent n rows and d columns of random number matrixes subject to standard normal distribution respectively, and β represents a factor of 1.5. The update of all water levels is related to Δu(t) and o(t).

[0086] The water level population X is updated by using formula eight, and it is judged whether the updated water level x satisfies the constraint condition of step 2, if not, the optimal water level is continuously searched, if yes, the obtained water level x is output. The update strategy is as follows:

[0087] x(t+1)=x(t)+η·Δu(t)+(1-η)·o(t)(Formula fourteen)

[0088] It is judged whether the current iteration number a is greater than or equal to A, if a < A, the optimization is continuously performed, if a > A, the optimization is terminated, and the final calculation result source load deviation F is output. b ′1′ and the water level x, so as to better track the load curve and achieve the purpose of smoothing the fluctuation of wind and light. The source load deviation and fluctuation of the water, wind and light complementary system on different typical days are shown in Table 1.

[0089] Table 1 Comparison of solving indexes on different typical days

[0090]

[0091] Step 5: The PSA and the traditional PSO algorithm in the application are compared by respectively optimizing the water, wind and light complementary system, and the comparison data of the optimization results are shown in Table 2.

[0092] Table 2 Comparison of optimization results of PSO and PSA

[0093]

[0094] In summary, the PSA algorithm of a new meta-heuristic algorithm is adopted in the application, and the shortcomings of poor convergence and slow convergence speed of the traditional algorithm are overcome. By comparing the optimization results of the PSA algorithm and the PSO algorithm, it can be seen that the convergence effect is more accurate, a representative scheme is obtained, the complementarity and seasonal characteristics of water, wind and light energy are fully utilized, and the safe, stable and economic operation of the power system is ensured.

Claims

1. A day-ahead optimal dispatch method for a cascade hydro-PSA wind power complementary generation system, characterized in that, The steps are: Step 1: The wind power prediction output P under four typical day situations is respectively predicted by using Monte Carlo scene generation method + K-means t w , the photovoltaic prediction output P t s and the load prediction power P t D , wherein: In the formula, P t w Pwind(t) represents the wind power prediction output at time t; P t s Ppv(t) represents the photovoltaic prediction output at time t; Pwind(t) represents the wind power prediction output at time t; P Ppv(t) represents the photovoltaic prediction output at time t; w represents wind power; s represents photovoltaic power. Step 2: Establish the constraint conditions of the cascade water, wind and light complementary system, which are respectively the power transmission channel constraint and the cascade water power constraint condition, and the constraint conditions are: 1) Power transmission channel constraint where P i h represents the output of the i-th hydro power station; represents the predicted output of the m-th group of wind power; represents the predicted output of the s-th group of photovoltaic power; P d represents the transmission capacity of the d-th transmission section; I represents the number of hydro power stations; M represents the number of wind power groups; and N represents the number of photovoltaic groups. 2) Cascade water power constraint condition Output constraint In the formula: represents the lower limit of the output of the i-th cascade hydropower station at time t; represents the upper limit of the output of the i-th cascade hydropower station at time t; represents the output of the i-th cascade hydropower station at time t; wherein K represents the output coefficient of the cascade hydropower station; Q i,t represents the power generation flow of the i-th cascade hydropower station in the t-th period; H i,t represents the water head height of the i-th cascade hydropower station in the t-th period; h represents the hydropower station; Load power balance constraint Cascade water power water balance constraint V i,t+1 = V i,t + [I i,t - (Q i,t + S i,t )] · Δt (Equation Five) wherein: V i,t+1 Vit+1represents the reservoir capacity of the ith hydroelectric power station at the t+1 time period, i,t Vitrepresents the reservoir capacity of the ith hydroelectric power station at the t time period; Iitrepresents the inflow of the ith hydroelectric power station at the t time period; i,t Iitrepresents the inflow of the ith hydroelectric power station at the t time period; Δt is the length of the calculation period; S i,t Iitrepresents the inflow of the ith hydroelectric power station at the t time period; Δt is the length of the calculation period; S Water power storage capacity, water level, flow constraint In the formula: represents the lower limit of the reservoir capacity of the i-th hydroelectric power station at time t, represents the upper limit of the reservoir capacity of the i-th hydroelectric power station at time t; represents the lower limit of the water level of the i-th hydroelectric power station at time t, represents the upper limit of the water level of the i-th hydroelectric power station at time t; Y i,t represents the water level of the i-th hydroelectric power station at time t; represents the lower limit of the power generation flow of the i-th hydroelectric power station at time t, represents the upper limit of the power generation flow of the i-th hydroelectric power station at time t; represents the initial reservoir capacity of the hydroelectric power station, represents the final reservoir capacity of the hydroelectric power station; is the initial reservoir capacity of the hydroelectric power station, is the final reservoir capacity of the hydroelectric power station; Step 3: Combine the wind power prediction output, photovoltaic prediction output, load prediction power and constraint conditions described in steps 1 and 2 to establish a day-ahead optimal scheduling model of the cascade water, wind and light complementary system with the minimum source and load deviation as the target; The objective function is as follows: F represents the source load deviation value; P represents the load predicted power at time t; t w P represents the wind power predicted output at time t; P represents the photovoltaic predicted output at time t; P represents the hydropower output of the i-th level hydropower station at time t; T represents the dispatching period; P represents the wind power predicted output of the m-th group of wind turbines at time t; P represents the photovoltaic predicted output of the n-th group of photovoltaic turbines at time t; Step 4: Optimization of the day-ahead optimal scheduling model of the cascade water, wind and light complementary system Set the population size B, the number of iterations A and the scheduling period C, the current iteration number a; take the water level as the decision variable, generate a set of random data representing the C-hour water level of the cascade hydropower stations, denoted as x = {x 1 ,x 2 ,x 3}, wherein represents the water level of the first-stage hydropower station, represents the water level of the second-stage hydropower station, represents the water level of the third-stage hydropower station; determine whether x satisfies the constraint condition of step 2; if not, continue to generate water levels satisfying the condition, and if yes, form a set of water levels satisfying the condition x; finally form an initial water level population, denoted as X = (X 1 X 2 X 3 ), wherein The above obtained X is used to calculate the water level and the power generation flow Q i,t The relationship and The output of the hydropower station is calculated, and the wind power prediction output P t w , the photovoltaic prediction output P t s And the load value P t D The source load deviation value is calculated by substituting formula seven, and is recorded find the minimum value F among B source load deviation values b1 corresponding water level x; determine the current iteration number a, if a = 1, set the current source load deviation value F b1 to the target value F' b1 , set the water level x to x', set the error e to x'-x, if a > 1, compare the current source load deviation value with the previous target value, set the smaller one as the latest target value F" b1 , set the water level to x", and the error e to x"-x; Minimize the system deviation by dynamically adjusting the error signal e of the PID and optimize the control process, update the water level population X by using formula eight, and judge whether the updated water level x meets the constraint condition of step 2, if not, continue to find the optimal water level, if yes, output the obtained water level x; the update strategy is: x(t+1) = x(t) + η·Δu(t) + (1-η)·o(t) (Formula eight) x(t) is the iteration output at this time based on the PID search algorithm, x(t+1) is the next iteration output; Δu(t) is the error at this time, η is an n-row 1-column matrix; o(t) is a zero output condition factor set to prevent the algorithm from falling into local optimum; Determine if the current iteration number a is greater than or equal to A, if a < A, continue to optimize, if a > A, terminate optimization, output the final calculation result source load deviation F" b1 And water level x, in order to better track the load curve and achieve the purpose of smoothing the fluctuation of wind power.

Images