A method for establishing a day-ahead dispatching model for hydropower, wind power and solar power

Abstract

A method for establishing a day-ahead scheduling model for hydropower, wind power, and solar power systems comprises the following steps: Step 1: generating scenarios that consider the spatiotemporal correlation of hydropower, wind power, and solar power systems based on the uncertainty of hydropower, wind power, and solar power prediction errors; Step 2: obtaining the effective moment of inertia of cascade hydropower and wind turbines under each scenario, and establishing effective moment of inertia constraints for the hydropower, wind power, and solar power systems; Step 3: establishing a day-ahead scheduling model for the hydropower, wind power, and solar power systems that considers the uncertainty of hydropower, wind power, and solar power systems and the flexible adjustment capabilities of cascade hydropower. The present invention aims to address the technical issues in existing power grid scheduling, such as the difficulty of cascade hydropower systems in fully utilizing their flexible adjustment capabilities, the limitation of wind power transmission, and the reduction of the effective moment of inertia of the hydropower, wind power, and solar power systems, resulting from the uncertainty of hydropower, wind power, and solar power systems. The present invention provides a day-ahead scheduling model for hydropower, wind power, and solar power systems that considers effective moment of inertia and uncertainty, and a method for establishing the model.

Description

Technical Field

[0001] The present invention relates to the field of new energy and electronic information technology, in particular to the field of power grid operation control technology based on hydropower, wind power and solar power, and specifically to a hydropower, wind power and solar power day-ahead scheduling model considering effective rotational inertia and uncertainty and a method for establishing the model. Background Art

[0002] To achieve clean, low-carbon energy transformation and sustainable development, new energy sources such as wind power and photovoltaics have experienced rapid development. However, due to the randomness and uncertainty of wind and solar energy in their physical distribution, as well as the instability of their large-scale integration into the power system, the power system's equivalent inertia level and the flexible adjustment of supply and demand have been significantly impacted. Large-scale, technologically mature cascade hydropower is a realistic and reliable option for large-scale centralized consumption of new energy in my country, both now and in the future. However, hydropower inflow is significantly affected by the spatiotemporal distribution of water resources and also exhibits a degree of uncertainty. Therefore, for the day-ahead scheduling of hydropower, wind, and solar power hybrid operations, it is necessary to fully consider the uncertainty of wind and solar power output and cascade hydropower inflow to leverage the flexible adjustment capabilities of cascade hydropower. Furthermore, it is necessary to quantify the fluctuations in effective moment of inertia caused by wind and water uncertainties to improve the frequency stability of the power grid after disturbances.

[0003] Patent application publication number CN110717688A discloses a method for short-term joint optimization scheduling of hydropower, wind power, and solar power, taking into account the uncertainty of renewable energy output. This method uses predicted wind and solar power output as a proxy for future combined output and constructs a stochastic expectation peak-shaving model for multi-scenario joint scheduling of hydropower, wind power, and solar power. Patent application publication number CN112467807A discloses a method and system for day-ahead optimization scheduling of a multi-energy power system. This method uses an improved generative adversarial network based on the Wasserstein distance to generate wind and solar power output scenarios and virtual net loads, and establishes a day-ahead optimization scheduling model for the multi-energy power system. This method primarily considers the uncertainty of wind and solar power output but fails to account for the complex impact of uncertainty in cascade hydropower inflow. This makes it difficult for the compiled scheduling plan to meet the actual operating requirements of the power grid and power stations, resulting in a certain degree of deviation during execution. This prevents cascade hydropower from fully utilizing its flexible adjustment capabilities and limits the export of wind and solar power. Moreover, the above method does not take into account the problem of reduced effective rotational inertia level and frequency support capacity of the power system caused by the uncertainty of wind and solar power, which makes it easy for safety and stability problems to occur when the power grid is subject to active power disturbances, greatly restricting the absorption of wind and solar power. Summary of the Invention

[0004] The purpose of the present invention is to solve the technical problems that the uncertainty of water, wind and solar power in the existing power grid scheduling makes it difficult for cascade hydropower to fully exert its flexible adjustment capabilities, limits the transmission of wind and solar power, and reduces the effective rotational inertia of the water, wind and solar power systems. A water, wind and solar power day-ahead scheduling model that takes into account the effective rotational inertia and uncertainty and a method for establishing the model are provided.

[0005] In order to solve the above technical problems, the technical solution adopted by the present invention is:

[0006] A method for establishing a day-ahead scheduling model for hydropower, wind power, and solar power includes the following steps:

[0007] Step 1: To address the uncertainty of water, wind and solar power prediction errors, generate scenarios that take into account the spatiotemporal correlation of water, wind and solar power;

[0008] Step 2: Obtain the effective moment of inertia of cascade hydropower and wind turbines in each scenario, and establish effective moment of inertia constraints for the hydropower, wind, and solar power systems;

[0009] Step 3: Establish a day-ahead dispatch model for hydropower, wind-solar power systems that takes into account the uncertainty of hydropower, wind-solar power systems and the flexible regulation capability of cascaded hydropower.

[0010] In step 3, the constraints of the day-ahead dispatch model of the water, wind and solar power systems established include the effective moment of inertia constraint of the water, wind and solar power systems, water balance constraint, water level constraint, initial and final water level constraint, reservoir capacity-water level constraint, tailwater level-outflow constraint, unit output constraint, unit power generation flow constraint, unit vibration zone constraint, unit start-up and shutdown duration constraint, unit output ramp constraint, unit power generation head constraint, head loss constraint, unit dynamic characteristic relationship, section transmission constraint, and power generation plan constraint;

[0011] The step 1 specifically includes the following steps:

[0012] Step 1-1: For the target scheduling time, use the kernel distribution function to fit the probability distribution function of the water, wind and solar power prediction error, perform Latin hypercube sampling on the probability distribution function of the water, wind and solar power prediction error, and generate the water, wind and solar power scenario;

[0013] Step 1-2: Use the Pearson correlation coefficient to obtain the time correlation coefficient matrix of historical data and the time correlation coefficient matrix of water-wind-light scene;

[0014] Step 1-3: Use the connection function to obtain the spatial correlation coefficient matrix of historical data and the spatial correlation coefficient matrix of water-landscape scenes;

[0015] Steps 1-4: Based on the differential evolution algorithm, the water-scenery scenes are sorted with the goal of minimizing the difference between the temporal correlation coefficient matrix of the historical data and the spatial correlation coefficient matrix of the historical data and the water-scenery scenes, and the sorted water-scenery scenes are obtained;

[0016] Steps 1-5: Obtain randomness and correlation indices of the water-landscape scenes before and after sorting, and conduct scene evaluation;

[0017] Step 1-6: Use the K-means clustering algorithm to reduce the sorted scenes and obtain typical scenes, that is, scenes that take into account the temporal and spatial correlation of water, scenery and light.

[0018] The step 2 specifically includes the following steps:

[0019] Step 2-1: Utilize the energy absorbed or released by the rotors of the cascade hydropower units and the wind turbines when the power system frequency changes to obtain the effective rotational inertia of the cascade hydropower units and the wind turbines;

[0020] Step 2-2: Based on the condition that the minimum effective moment of inertia that the hydropower, wind-solar system can provide should be greater than the effective moment of inertia required by the power grid, establish the effective moment of inertia constraint of the hydropower, wind-solar system.

[0021] In step 1-1, when obtaining the water-wind-solar forecast error, the specific steps are as follows:

[0022] Arrange the water, wind, and solar power forecast values ​​in ascending order and divide them into several forecast boxes according to the equal frequency principle. Each box has a corresponding forecast error. For the forecast value at the target time, calculate the kernel distribution function of the corresponding forecast box and obtain the probability distribution function of the water, wind, and solar power forecast error:

[0023]

[0024] Among them, F k,t (x) is the probability distribution function of the prediction error of power station k at time t, M is the number of data in the prediction box, M is a positive integer, h k,t is the bandwidth of the kernel distribution function of power station k at time t, K(·) is the kernel function, x k,t is the prediction error variable of power station k at time t; is the size of the mth prediction error in the prediction box of power station k at time t, m = 1, 2, ..., M, k = 1, 2, ..., K1; K1 is the total number of power stations included in the hydro-wind-solar system, which is a positive integer; t = 1, 2, ..., T1, T1 is the total number of scheduling times;

[0025] The probability interval [0,1] is evenly divided into S intervals, and a point is randomly selected in each interval as the value of the probability distribution function of the water-wind-solar prediction error, and the corresponding prediction error value is obtained:

[0026]

[0027] Among them, x s,k,tis the prediction error of power station k at time t under scenario s, s=1,2,…,S, S is the number of generated scenarios, is the inverse function of the probability distribution function of the prediction error of power station k at time t, Y s,k,t is the quantile drawn by power station k at time t in interval s.

[0028] In step 1-2, the Pearson correlation coefficient is used to obtain the time correlation coefficient matrix of the historical data and the time correlation coefficient matrix of the water-wind-light scene, where:

[0029] Time correlation coefficient matrix τ of historical data k Elements in The calculation is as follows:

[0030]

[0031] In formula (3), is the time correlation coefficient of the historical data of power station k at time i and time j; N1 is the total number of historical data; is the n1th value of the historical prediction error of power station k at time i; is the n1th value of the historical forecast error of power station k at time j; is the average value of the historical forecast error of power station k at time i; is the average value of the historical forecast error of power station k at time j;

[0032] Time correlation coefficient matrix τ' of water scene k Elements in The calculation is as follows:

[0033]

[0034] in, is the time correlation coefficient of the water-wind-solar scene of power station k at time i and time j; s,k,i is the prediction error of power station k at time i under scenario s; x s,k,j is the prediction error of power station k at time j under scenario s; is the average value of the prediction error scenario of power station k at time i; is the average value of the prediction error scenario of power station k at time j; S is the number of generated scenarios.

[0035] In steps 1-3, the connection function is used to obtain the spatial correlation coefficient matrix of historical data and the spatial correlation coefficient matrix of water-landscape scenes, where:

[0036] Select t-Copula as the connection function and use Matlab to fit to obtain formula (5):

[0037]

[0038] in, is the joint probability distribution function of the prediction error at time t; x 1,t is the prediction error variable of power station 1 at time t; x 2,t is the prediction error variable of power station 2 at time t; is the prediction error variable of power station n2 at time t; The degrees of freedom are v t , the correlation coefficient matrix is ​​ρ t The n-dimensional t distribution function; The degrees of freedom are v t The inverse function of the t distribution; F 1,t (x 1,t ) is the probability distribution function of the prediction error of power station 1 at time t; F 2,t (x 2,t ) is the probability distribution function of the prediction error of power station 2 at time t; is the probability distribution function of the prediction error of power station n2 at time t;

[0039] Selection matrix ρ t As the spatial correlation coefficient matrix of historical data;

[0040] Spatial correlation coefficient matrix ρ' of water scene t Elements in The calculation formula is as follows:

[0041]

[0042] in, is the spatial correlation coefficient of the water-wind-light scenes of power station k and power station l at time t; h s,a,t is the value of the random variable of the ath dimension of the multidimensional t distribution at time t under scenario s; h s,b,t is the value of the random variable of the bth dimension of the multidimensional t distribution at time t under scenario s; is the average value of the random variable of the ath dimension of the multidimensional t distribution at time t in all scenarios; is the average value of the random variable of the bth dimension of the multidimensional t distribution at time t in all scenarios; S is the number of generated scenarios;

[0043] From formula (7) and formula (8), we can get h s,a,t and h s,b,t as follows:

[0044]

[0045]

[0046] Among them, Fa,t (·) is the probability distribution function of power station a at time t; x s,a,t is the prediction error of power station a at time t under scenario s; F b,t (·) is the probability distribution function of power station b at time t; x s,b,t is the prediction error of power station b at time t under scenario s, It is the inverse function of the t distribution with v degrees of freedom.

[0047] In steps 1-4, the water-wind-light scenes are sorted based on the differential evolution algorithm with the goal of minimizing the difference between the temporal correlation coefficient matrix of the historical data and the water-wind-light scenes, as well as the difference between the spatial correlation coefficient matrix of the historical data and the water-wind-light scenes, to obtain the sorted water-wind-light scenes. The specific steps are as follows:

[0048] (1) Establish the objective function of the differential evolution algorithm

[0049]

[0050] Among them, K1 is the total number of power stations included in the hydro-wind-solar system, T1 is the total number of dispatching times, is the F-norm of the difference between the time correlation coefficient matrix of the historical data of power station k and the time correlation coefficient matrix of the scenario of power station k; is the F-norm of the difference between the spatial correlation coefficient matrix of the historical data at time t and the spatial correlation coefficient matrix of the scene at time t;

[0051] (2) Initialize the individual

[0052] A population of size (N, D) is randomly generated, where N is the number of individuals in the population, and D is the number of components of the individual. Each individual in the population contains 4 components: the first two components are scene numbers, representing the scene where the prediction error is located; the third component is the power station number, representing the type of power station; the fourth component is the time number, representing the moment where the prediction error is located. For any individual, the type and moment of the power station are first determined according to the third and fourth components, and the data on the two scenes are swapped according to the first two components, and then the objective function is calculated while the rest of the data remains unchanged; if the objective function becomes smaller, the swap is maintained and the scene order is changed; if the objective function does not become smaller, the swap is restored and the scene order remains unchanged; when all individuals in the population have completed comparing the objective function and choosing whether to swap, the population enters the next generation through differential mutation and crossover operations, and the i-th individual No. i As shown in the following formula:

[0053] No i ={No i,1 ,No i,2 ,No i,3 ,Noi,4} (10)

[0054] Among them, No i,1 Indicates the value of the first component of the i-th individual; No i,2 Indicates the value of the second component of the i-th individual; No i,3 Indicates the value of the third component of the i-th individual; No i,4 represents the value of the jth component of the i-th individual; i = 1, 2, ... N;

[0055] The value of the jth component of the i-th individual is calculated as follows:

[0056] No i,j =round(rand(0,1)·(U j -L j )+L j ) (11)

[0057] Among them, No i,j is the value of the jth component of the i-th individual; round(·) is the rounding function; rand(0,1) represents a random number in the interval [0,1]; L j is the lower bound of the jth component of the i-th individual; U j is the upper bound of the jth component of the i-th individual; j = 1, 2, 3, 4;

[0058] (3) Differential mutation operation

[0059] The differential mutation operation first uses a random method to select a series of different individual vectors from the population in the previous generation to generate a differential vector. Then, the differential vector is scaled and superimposed with the basis vector to generate a mutant individual. The jth mutation component of the i-th individual in the G-th generation is generated by the following formula:

[0060]

[0061] in, is the value of the jth variation component of the i-th individual in the G-th generation; is the value of the jth component of the r1th individual in the G-1th generation; is the value of the jth component of the r2th individual in the G-1th generation; is the value of the jth component of the r3th individual in the G-1th generation; F is the scaling factor, which is a random number in the interval [0,1], and r1, r2, and r3 are mutually different integers randomly selected in [0,N];

[0062] In the mutation operation, the DE algorithm determines whether the new individuals generated by the differential mutation operation are beyond the boundary. If they are beyond the boundary, the following method is used to process the mutated individuals:

[0063]

[0064] (4) Crossover operation

[0065] The crossover method commonly used in DE algorithm is as follows:

[0066]

[0067] in, is the value of the jth component of the i-th individual in the G-th generation; is the value of the jth component of the i-th individual in the G-1th generation; CR is the crossover probability;

[0068] (5) After all individuals in the new generation population have completed the comparison of the objective function and the selection of whether to swap, steps (3) to (4) are repeated until the population evolves to the set number of iterations. The final scene is the sorted water scene.

[0069] In steps 1-5, the randomness and correlation indices of water, wind and light before and after sorting are calculated to conduct scenario evaluation;

[0070] To measure the effectiveness of the spatiotemporal correlation ranking, the mean absolute error was chosen as the randomness indicator and the difference score as the correlation indicator;

[0071] The calculation formula for the mean absolute error of power station k is:

[0072]

[0073] in, is the historical actual value of the prediction error of power station k at time t, x s,k,t is the prediction error of power station k at time t under scenario s;

[0074] The time difference fraction of station k is:

[0075]

[0076] The spatial difference score of station k is:

[0077]

[0078] in, The weight coefficient of power station k between time t and time w is represented by the absolute value of the time correlation coefficient of the historical data of power station k at time t and time w; is the historical actual value of the prediction error of power station k at time t; x s,k,w is the prediction error of power station k at time w under scenario s; represents the weight coefficient between power station k and power station l at time t, which is represented by the absolute value of the spatial correlation coefficient of the historical data of power station k and power station l at time t; is the historical actual value of the prediction error of power station l at time t; x s,l,t is the prediction error of power station l at time t under scenario s; γ represents the difference fractional order; It refers to the difference between the historical forecast error of power station k at time t and time w; It refers to the difference between the prediction error of power station k at time t and time w under scenario s; It refers to the difference in historical prediction errors between plant k and plant l at time t; It refers to the difference in prediction error between plant k and plant l at time t under scenario s.

[0079] In steps 1-6, the K-means clustering algorithm is used to reduce the sorted scenes to obtain typical scenes, that is, scenes that take into account the spatiotemporal correlation between water, scenery and light;

[0080] Adding the sorted scenarios to the predicted values ​​yields the hydropower inflow scenarios and wind and solar output:

[0081]

[0082]

[0083]

[0084] in, is the output of PV station g at time t under scenario s, g = 1, 2, ..., G, G is the number of PV stations in the hydro-wind-solar system, and G is a positive integer; is the output of wind farm w at time t under scenario s, w=1,2,...,W, W is the number of wind farms in the hydro-wind-solar system, and W is a positive integer; R s,i,t is the interval water inflow of hydropower station i at time t under scenario s, i=1,2,...,I, I is the number of cascade hydropower stations in the water-wind-solar system, and I is a positive integer. is the predicted output of photovoltaic power station g at time t, is the predicted output of wind farm w at time t, is the predicted interval water inflow of cascade hydropower station i at time t;

[0085] The K-means clustering algorithm is used to reduce the hydropower inflow scenario and wind and solar power output scenario to obtain the spatiotemporal correlation scenario of water, wind and solar power.

[0086] In step 2-1,

[0087] Cascade hydropower effective moment of inertia E h The calculation formula is as follows:

[0088]

[0089] Among them, H is the inherent inertia time constant of the hydropower unit itself; A is the capacity of the hydropower unit; ω is the angular velocity of the hydropower unit; ω n is the rated angular velocity of the hydropower unit; ω min is the minimum angular velocity allowed by the power system; ω max is the maximum angular velocity allowed by the power system;

[0090] Effective moment of inertia of fan E W With wind speed V W The relationship is as follows:

[0091]

[0092] Among them, J w is the moment of inertia of the fan rotor; opt is the optimal tip speed ratio; R is the fan radius; ω r_lim is the allowable speed limit of the fan rotor; v max is the maximum speed of wind turbine connected to the grid; ω s V is the speed of the fan entering the constant speed zone; Wm V is the lower limit wind speed for the wind turbine to enter the maximum power point tracking area; Ws V is the lower limit wind speed for the fan to enter the constant speed zone; Wn is the rated wind speed of the fan; V Woff is the fan cut-out wind speed, V w refers to wind speed;

[0093] Utilize wind power output P WT Calculate the wind speed V w , the relationship between wind power output and wind speed is as follows:

[0094]

[0095] Among them, P WTn is the rated output power of the fan; V Win is the wind turbine cut-in speed;

[0096] Final effective moment of inertia of the fan E W The calculation is as follows:

[0097]

[0098] in,

[0099] Among them, P wt is the per-unit value of wind power output, and R refers to the radius of the wind turbine.

[0100] Step 2-2: The effective moment of inertia constraint of the established water-wind-solar system is:

[0101]

[0102]

[0103] Among them, u i,n,t is the start-stop state variable of unit n of hydropower station i at time t, 1 represents the start state, and 0 represents the stop state; is the effective moment of inertia of unit n of hydropower station i at time t when the frequency rises; is the effective moment of inertia of unit n of hydropower station i at time t when the frequency decreases; is the effective moment of inertia of the mth wind turbine in wind farm w at time t in scenario s when the frequency rises; is the effective moment of inertia of the mth wind turbine in wind farm w at time t in scenario s when the frequency drops; is the effective moment of inertia required by the grid at time t when the frequency rises; is the effective moment of inertia required by the grid at time t when the frequency drops; I is the number of cascade hydropower stations in the water-wind-solar system, N i is the number of units in hydropower station i, W is the number of wind farms in the hydro-wind-solar system, M w is the number of wind turbines in wind farm w.

[0104] The step 3 specifically includes the following steps:

[0105] Step 3-1: Obtain the flexible adjustment requirements of the hydro-wind-solar hybrid system based on the difference between the predicted wind-solar output and the output under scenario s;

[0106] Step 3-2: Obtain the flexible adjustment capability of the hydropower-wind-solar hybrid system based on the ramping capability and output limit of the hydropower units;

[0107] Step 3-3: Establish an objective function based on flexible adjustment requirements and flexible adjustment capabilities;

[0108] Step 3-4: Establish the constraints of the water, wind and solar system.

[0109] In step 3-1, the flexible adjustment demand of the hydro-wind-solar hybrid system is obtained based on the difference between the predicted wind-solar output and the output under scenario s. The calculation formula is as follows:

[0110]

[0111]

[0112]

[0113]

[0114] in, Flexible upward adjustment requirements at time t under scenario s; The flexible downward adjustment requirement at time t under scenario s; The flexible upward adjustment demand of the photovoltaic power station at time t under scenario s; is the flexible upward adjustment demand of the wind farm at time t under scenario s, The flexible downward adjustment demand of the PV power station at time t under scenario s; is the flexible downward adjustment demand of the wind farm at time t under scenario s.

[0115] In step 3-2, the flexible adjustment capability of the hydropower-wind-solar hybrid system is obtained based on the ramping capability and output limit of the hydropower unit. The calculation formula is as follows:

[0116]

[0117] in, is the flexible upward adjustment capability of cascade hydropower at time t; is the flexible downward adjustment capability of cascade hydropower at time t; is the gradeability of hydropower station i; is the output of hydropower station i at time t; is the upper limit of the output of hydropower station i at time t; is the lower limit of the output of hydropower station i at time t.

[0118] In step 3-3, the objective function established is as follows:

[0119]

[0120]

[0121]

[0122]

[0123] in, The expected flexibility increase deficiency refers to the expected difference between the flexible increase demand and the flexible increase capacity at time t due to insufficient increase capacity. The expected flexibility downsizing shortage refers to the expected difference between the flexible downsizing demand and the flexible downsizing capacity at time t due to insufficient downsizing capacity. is the amount of water abandoned by hydropower station i at time t under scenario s; is the expected amount of water abandoned by cascade hydropower at time t; P(s) refers to the probability of scenario s.

[0124] In step 3-4, the constraints of the water-wind-solar system are established as follows;

[0125] (1) Water balance constraints

[0126]

[0127] Among them, V s,i,t+1 is the storage capacity of hydropower station i at time t+1 under scenario s; V s,i,t is the storage capacity of hydropower station i at time t under scenario s; I s,i,t is the inflow flow of hydropower station i at time t under scenario s; Q s,i,t is the outflow of hydropower station i at time t under scenario s; Δt is the scheduling time interval; τ i-1 is the water flow lag time between hydropower station i-1 and hydropower station i; is the hydropower station i-1 at time t-τ under scenario s i-1 flow rate; R s,i,t is the interval flow between hydropower station i-1 and hydropower station i at time t under scenario s; is the power generation flow of hydropower station i at time t under scenario s;

[0128] (2) Water level constraint

[0129]

[0130] in, is the reservoir water level of hydropower station i at time t under scenario s; Z i,t is the lower limit of the water level of hydropower station i at time t; is the upper limit of the water level of hydropower station i at time t;

[0131] (3) Initial and final water level constraints

[0132]

[0133]

[0134] in, is the initial water level in the scheduling period; is the final water level in the scheduling period; ΔZ is the allowable deviation, Refers to the initial water level of cascade hydropower station i in the scheduling period under scenario s, It refers to the water level at the end of the dispatching period of cascade hydropower station i under scenario s;

[0135] (4) Reservoir capacity-water level constraints

[0136]

[0137] in, is the storage capacity-water level function of hydropower station i;

[0138] (5) Tailwater level-outflow constraint

[0139]

[0140] in, is the outflow-tailwater level function of hydropower station i; is the tailwater level of power station i at time t under scenario s;

[0141] (6) Unit output constraints

[0142]

[0143] Among them, P i,n,t is the output of unit n of hydropower station i at time t under scenario s, is the upper limit of the output of unit n of hydropower station i, P i,n is the lower limit of the output of unit n of hydropower station i. The output of hydropower station i is expressed as:

[0144]

[0145] Among them, N i represents the number of units of hydropower station i;

[0146] (7) Unit power generation flow constraints

[0147]

[0148] in, is the power generation flow of unit n of hydropower station i at time t under scenario s; is the upper limit of power generation flow of unit n of hydropower station i; is the lower limit of the power generation flow of unit n of hydropower station i. The power generation flow of hydropower station i is expressed as:

[0149]

[0150] (8) Unit vibration zone constraints

[0151]

[0152] in, is the output upper limit of vibration zone k of unit n of hydropower station i; is the lower limit of the output of unit n in vibration zone k of hydropower station i;

[0153] (9) Unit start-up and shutdown duration constraints:

[0154]

[0155] in, is the startup operation variable of unit n of hydropower station i at time t, 1 indicates startup operation; is the shutdown operation variable of unit n of hydropower station i at time t, 1 indicates shutdown operation; is the minimum startup duration of unit n of hydropower station i; is the minimum shutdown duration of unit n of hydropower station i; is the maximum number of startup times of unit n of hydropower station i within the scheduling period;

[0156] (10) Unit output ramp constraints:

[0157] -ΔP i,n ≤P i,n,t+1 -P i,n,t ≤ΔP i,n (49)

[0158] Among them, P i,n,t+1 is the output of unit n of hydropower station i at time t+1 under scenario s, ΔP i,n is the ramping capability of unit n of hydropower station i;

[0159] (11) Unit generating head constraints:

[0160]

[0161] in, is the reservoir water level of hydropower station i at time t-1 under scenario s, H s,i,n,t is the generating head of unit n of hydropower station i at time t under scenario s; is the head loss of unit n of hydropower station i at time t under scenario s;

[0162] (12) Head loss constraint:

[0163]

[0164] Among them, a i is the head loss coefficient of power station i; b i is the head loss constant of power station i;

[0165] (13) Relationship of unit power characteristics:

[0166]

[0167] in, is the head-flow-output function of unit n of hydropower station i;

[0168] (14) Section transmission constraints;

[0169]

[0170] Among them, I o is the number of hydropower stations included in section o; W o is the number of wind farms included in section o; G o is the number of photovoltaic power stations included in section o; is the transmission capacity of section o, is the output of wind farm w at time t under scenario s;

[0171] (15) Power generation plan constraints:

[0172]

[0173] Among them, ε is a smaller deviation threshold, It represents the power generation plan issued by the power grid, G is the number of photovoltaic power stations in the water-wind-solar system, I is the number of cascade hydropower stations in the water-wind-solar system, and W is the number of wind farms in the water-wind-solar coefficient.

[0174] The present invention also includes a day-ahead scheduling model for water, wind, and solar power that takes into account effective moment of inertia and uncertainty. The objective function of the model is:

[0175]

[0176]

[0177]

[0178]

[0179] in, The expected flexibility increase deficiency refers to the expected difference between the flexible increase demand and the flexible increase capacity at time t due to insufficient increase capacity. The expected flexibility downsizing shortage refers to the expected difference between the flexible downsizing demand and the flexible downsizing capacity at time t due to insufficient downsizing capacity. is the amount of water abandoned by hydropower station i at time t under scenario s; is the expected amount of water abandoned by cascade hydropower at time t; P(s) refers to the probability of scenario s.

[0180] The constraints of the above model include the effective moment of inertia constraint of the water-wind-solar system, water balance constraint, water level constraint, initial and final water level constraint, reservoir capacity-water level constraint, tailwater level-outflow constraint, unit output constraint, unit power flow constraint, unit vibration zone constraint, unit start-up and shutdown duration constraint, unit output ramp constraint, unit power head constraint, head loss constraint, unit dynamic characteristic relationship, section transmission constraint, and power generation plan constraint; the details are as follows;

[0181] 1) The effective moment of inertia constraint of the water-wind-solar system is:

[0182]

[0183]

[0184] 2) Water balance constraints are:

[0185]

[0186] Among them, V s,i,t+1 is the storage capacity of hydropower station i at time t+1 under scenario s; V s,i,t is the storage capacity of hydropower station i at time t under scenario s; I s,i,t is the inflow flow of hydropower station i at time t under scenario s; Q s,i,t is the outflow of hydropower station i at time t under scenario s; Δt is the scheduling time interval; τ i-1 is the water flow lag time between hydropower station i-1 and hydropower station i; is the hydropower station i-1 at time t-τ under scenario s i-1 flow rate; R s,i,t is the interval flow between hydropower station i-1 and hydropower station i at time t under scenario s; is the power generation flow of hydropower station i at time t under scenario s;

[0187] 3) The water level constraint is:

[0188]

[0189] in, is the reservoir water level of hydropower station i at time t under scenario s; Z i,t is the lower limit of the water level of hydropower station i at time t; is the upper limit of the water level of hydropower station i at time t;

[0190] 4) The initial and final water level constraints are:

[0191]

[0192]

[0193] in, is the initial water level in the scheduling period; is the final water level in the scheduling period; ΔZ is the allowable deviation, Refers to the initial water level of cascade hydropower station i in the scheduling period under scenario s, It refers to the water level at the end of the dispatching period of cascade hydropower station i under scenario s;

[0194] 5) Reservoir capacity-water level constraint is:

[0195]

[0196] in, is the storage capacity-water level function of hydropower station i;

[0197] 6) Tailwater level-outflow constraint is:

[0198]

[0199] in, is the outflow-tailwater level function of hydropower station i; is the tailwater level of power station i at time t under scenario s;

[0200] 7) Unit output constraints are:

[0201]

[0202] Among them, P i,n,t is the output of unit n of hydropower station i at time t under scenario s, is the upper limit of the output of unit n of hydropower station i, P i,n is the lower limit of the output of unit n of hydropower station i. The output of hydropower station i is expressed as:

[0203]

[0204] Among them, N i represents the number of units of hydropower station i;

[0205] 8) The power generation flow constraint of the unit is:

[0206]

[0207] in, is the power generation flow of unit n of hydropower station i at time t under scenario s; is the upper limit of power generation flow of unit n of hydropower station i; is the lower limit of the power generation flow of unit n of hydropower station i. The power generation flow of hydropower station i is expressed as:

[0208]

[0209] 9) The vibration zone constraints of the unit are:

[0210]

[0211] in, is the output upper limit of vibration zone k of unit n of hydropower station i; is the lower limit of the output of unit n in vibration zone k of hydropower station i;

[0212] 10) The unit start-up and shutdown duration constraints are:

[0213]

[0214] in, is the startup operation variable of unit n of hydropower station i at time t, 1 indicates startup operation; is the shutdown operation variable of unit n of hydropower station i at time t, 1 indicates shutdown operation; is the minimum startup duration of unit n of hydropower station i; is the minimum shutdown duration of unit n of hydropower station i; is the maximum number of startup times of unit n of hydropower station i within the scheduling period;

[0215] 11) The unit output ramp constraint is:

[0216] -ΔP i,n ≤P i,n,t+1 -P i,n,t ≤ΔP i,n (49)

[0217] Among them, P i,n,t+1 is the output of unit n of hydropower station i at time t+1 under scenario s, ΔP i,n is the ramping capability of unit n of hydropower station i;

[0218] 12) The unit's generating head constraint is:

[0219]

[0220] in, is the reservoir water level of hydropower station i at time t-1 under scenario s, H s,i,n,t is the generating head of unit n of hydropower station i at time t under scenario s; is the head loss of unit n of hydropower station i at time t under scenario s;

[0221] 13) The head loss constraint is:

[0222]

[0223] Among them, a i is the head loss coefficient of power station i; b i is the head loss constant of power station i;

[0224] 14) The relationship between the unit's dynamic characteristics is:

[0225]

[0226] in, is the head-flow-output function of unit n of hydropower station i;

[0227] 15) Section transmission constraints are:

[0228]

[0229] Among them, I o is the number of hydropower stations included in section o; W o is the number of wind farms included in section o; G o is the number of photovoltaic power stations included in section o; is the transmission capacity of section o, is the output of wind farm w at time t under scenario s;

[0230] 16) The power generation plan constraints are:

[0231]

[0232] Among them, ε is a smaller deviation threshold, It represents the power generation plan issued by the power grid, G is the number of photovoltaic power stations in the water-wind-solar system, I is the number of cascade hydropower stations in the water-wind-solar system, and W is the number of wind farms in the water-wind-solar coefficient.

[0233] The present invention also includes an electronic device, including a memory and a processor. The processor and the memory are connected via a communication bus. The memory is used to store a computer program. When the computer program is executed by the processor, the above-mentioned water, wind and solar day-ahead scheduling model considering effective rotational inertia and uncertainty is implemented.

[0234] Compared with the prior art, the present invention has the following technical effects:

[0235] 1) The water-wind-solar scenario generated by the present invention better reflects the randomness and correlation characteristics of water-wind-solar prediction errors. By combining Latin hypercube sampling and differential evolution algorithms to generate water-wind-solar scenarios, and comparing randomness, temporal correlation, and spatial correlation indices, it is demonstrated that the water-wind-solar scenario generated by the present invention can better characterize the randomness, temporal correlation, and spatial correlation characteristics of water-wind-solar prediction errors, providing a more realistic data foundation for the subsequent formulation of scheduling plans.

[0236] 2) This invention takes into account the impact of the uncertainty of cascade hydropower inflows and fully exploits the flexible regulation capabilities of cascade hydropower. Throughout the entire scheduling cycle, a cascade hydropower generation plan is formulated to meet all water inflow scenarios. This scheduling plan meets the actual operating requirements of the power grid and hydropower stations, fully leveraging the flexible regulation capabilities of cascade hydropower.

[0237] 3) This invention proposes effective moment of inertia constraints to ensure the effective moment of inertia level of the hydropower, wind and solar power systems. The energy that may be absorbed or released by the wind turbine and turbine rotors is used to calculate the rising and falling effective moments of inertia of the hydropower, wind and solar power systems under various scenarios. Effective moment of inertia constraints are then established to rationally schedule the start and stop of hydropower units, ensuring the effective moment of inertia level of the system and improving the frequency support capability of the power grid.

[0238] 4) The present invention comprehensively considers the uncertainty of water, wind and solar power and the problem of the decrease in effective rotational inertia caused by it, accurately characterizes the uncertainty of water, wind and solar power, quantitatively calculates the effective rotational inertia, and proposes a method for establishing a water, wind and solar power day-ahead scheduling model considering the effective rotational inertia and uncertainty. On the basis of meeting the grid scheduling instructions, it effectively taps the flexible adjustment capability of the water, wind and solar power complementary system, and at the same time improves the frequency stability capability of the grid after disturbance. BRIEF DESCRIPTION OF THE DRAWINGS

[0239] The present invention will be further described below with reference to the accompanying drawings and embodiments:

[0240] Figure 1 It is the overall flow chart of the present invention;

[0241] Figure 2 is a spatial difference score map before and after sorting of water scene in an embodiment of the present invention;

[0242] Figure 3 This is the output scene diagram of the photovoltaic power station 1 finally obtained in the embodiment of the present invention;

[0243] Figure 4 This is the wind farm 1 output scene diagram finally obtained in the embodiment of the present invention;

[0244] Figure 5 This is the final water inflow scene diagram of the cascade hydropower station 4 obtained in the embodiment of the present invention;

[0245] Figure 6 This is a diagram of the effective moment of inertia provided by the water-wind-solar system in an embodiment of the present invention;

[0246] Figure 7 This is a comparison chart of the regulation capacity and regulation demand of the water, wind and solar system in an embodiment of the present invention;

[0247] Figure 8 This is a diagram showing the operation of unit 1 of a cascade hydropower station according to an embodiment of the present invention;

[0248] Figure 9 This is a diagram showing the operation of unit 2 of a cascade hydropower station according to an embodiment of the present invention;

[0249] Figure 10 This is a diagram showing the operation of three units in a cascade hydropower station according to an embodiment of the present invention;

[0250] Figure 11 This is a diagram showing the operation of four units in a cascade hydropower station according to an embodiment of the present invention;

[0251] Figure 12 This is a comparison chart of the output of the water, wind, and solar power systems and the power generation plan of the power grid in an embodiment of the present invention;

[0252] Figure 13 It is a structural schematic diagram of the electronic device provided by the present invention. DETAILED DESCRIPTION

[0253] A method for establishing a day-ahead scheduling model for hydropower, wind power, and solar power. The overall process of this method is as follows: Figure 1 shown.

[0254] Taking June 18, 2020 as an example, 5 photovoltaic power stations, 3 wind farms and 4 cascade hydropower stations in the experimental area were dispatched. The specific data of each power station is shown in Table 1:

[0255] Table 1 Power station capacity and quantity

[0256] power station Capacity / MW Quantity / unit Hydropower Station 1 600 3 Hydropower Station 2 695 4 Hydropower Station 3 600 3 Hydropower Station 4 1250 5 Photovoltaic power station 1 240 80 Photovoltaic power station 2 300 100 Photovoltaic power station 3 420 140 Photovoltaic power station 4 360 120 Photovoltaic power station 5 300 100 Wind Farm 1 400 80 Wind Farm 2 300 60 Wind Farm 3 500 100

[0257] The method comprises the following steps:

[0258] Step 1: To address the uncertainty of water, wind and solar power prediction errors, generate scenarios that take into account the spatiotemporal correlation of water, wind and solar power;

[0259] Step 1-1: For the target scheduling time, use the kernel distribution function to fit the probability distribution function of the water, wind and solar power prediction error, perform Latin hypercube sampling on the probability distribution function of the water, wind and solar power prediction error, and generate the water, wind and solar power scenario;

[0260] When obtaining the water-wind-solar prediction error, the specific steps are as follows:

[0261] Arrange the water, wind, and solar power forecast values ​​in ascending order and divide them into several forecast boxes according to the equal frequency principle. Each box has a corresponding forecast error. For the forecast value at the target time, calculate the kernel distribution function of the corresponding forecast box and obtain the probability distribution function of the water, wind, and solar power forecast error:

[0262]

[0263] Among them, F k,t(x) is the probability distribution function of the prediction error of power station k at time t, M is the number of data in the prediction box, M is a positive integer, h k,t is the bandwidth of the kernel distribution function of power station k at time t, K(·) is the kernel function, x k,t is the prediction error variable of power station k at time t; is the size of the mth prediction error in the prediction box of power station k at time t, m = 1, 2, ..., M, k = 1, 2, ..., K1; K1 is the total number of power stations included in the hydro-wind-solar system, which is a positive integer; t = 1, 2, ..., T1, T1 is the total number of scheduling times;

[0264] The probability interval [0,1] is evenly divided into S intervals, and a point is randomly selected in each interval as the value of the probability distribution function of the water-wind-solar prediction error, and the corresponding prediction error value is obtained:

[0265]

[0266] Among them, x s,k,t is the prediction error of power station k at time t under scenario s, s=1,2,…,S, S is the number of generated scenarios, is the inverse function of the probability distribution function of the prediction error of power station k at time t, Y s,k,t is the quantile drawn by power station k at time t in interval s.

[0267] Step 1-2: Use the Pearson correlation coefficient to obtain the time correlation coefficient matrix of historical data and the time correlation coefficient matrix of water-wind-light scene, where:

[0268] Time correlation coefficient matrix τ of historical data k Elements in The calculation is as follows:

[0269]

[0270] In formula (3), is the time correlation coefficient of the historical data of power station k at time i and time j; N1 is the total number of historical data; y n1,k,i is the n1th value of the historical prediction error of power station k at time i; is the n1th value of the historical forecast error of power station k at time j; is the average value of the historical forecast error of power station k at time i; is the average value of the historical forecast error of power station k at time j;

[0271] Time correlation coefficient matrix τ' of water scene k Elements in The calculation is as follows:

[0272]

[0273] in, is the time correlation coefficient of the water-wind-solar scene of power station k at time i and time j; s,k,i is the prediction error of power station k at time i under scenario s; x s,k,j is the prediction error of power station k at time j under scenario s; is the average value of the prediction error scenario of power station k at time i; is the average value of the prediction error scenario of power station k at time j; S is the number of generated scenarios;

[0274] Step 1-3: Use the connection function to obtain the spatial correlation coefficient matrix of historical data and the spatial correlation coefficient matrix of water-landscape scene, where:

[0275] Select t-Copula as the connection function and use Matlab to fit to obtain formula (5):

[0276]

[0277] in, is the joint probability distribution function of the prediction error at time t; x 1,t is the prediction error variable of power station 1 at time t; x 2,t is the prediction error variable of power station 2 at time t; is the prediction error variable of power station n2 at time t; The degrees of freedom are v t , the correlation coefficient matrix is ​​ρ t The n-dimensional t distribution function; The degrees of freedom are v t The inverse function of the t distribution; F 1,t (x 1,t ) is the probability distribution function of the prediction error of power station 1 at time t; F 2,t (x 2,t ) is the probability distribution function of the prediction error of power station 2 at time t; is the probability distribution function of the prediction error of power station n2 at time t;

[0278] Selection matrix ρ t As the spatial correlation coefficient matrix of historical data;

[0279] Then the spatial correlation coefficient matrix ρ' of the water scene is t Elements in The calculation formula is as follows:

[0280]

[0281] Among them, h s,a,t is the value of the random variable of the ath dimension of the multidimensional t distribution at time t under scenario s; h s,b,t is the value of the random variable of the bth dimension of the multidimensional t distribution at time t under scenario s; is the average value of the random variable of the ath dimension of the multidimensional t distribution at time t in all scenarios; is the average value of the random variable of the bth dimension of the multidimensional t distribution at time t in all scenarios; S is the number of generated scenarios;

[0282] From formula (7) and formula (8), we can get h s,a,t and h s,b,t as follows:

[0283]

[0284]

[0285] Among them, F a,t (·) is the probability distribution function of power station a at time t; x s,a,t is the prediction error of power station a at time t under scenario s; F b,t (·) is the probability distribution function of power station b at time t; x s,b,t is the prediction error of power station b at time t under scenario s, It is the inverse function of the t distribution with v degrees of freedom;

[0286] Steps 1-4: Based on the differential evolution algorithm, the water-wind-light scenes are sorted with the goal of minimizing the difference between the temporal correlation coefficient matrix of historical data and the spatial correlation coefficient matrix of historical data and the water-wind-light scenes. The sorted water-wind-light scenes are obtained. The specific steps are as follows:

[0287] (1) Establish the objective function of the differential evolution algorithm

[0288]

[0289] Among them, K1 is the total number of power stations included in the hydro-wind-solar system, T1 is the total number of dispatching times, is the F-norm of the difference between the time correlation coefficient matrix of the historical data of power station k and the time correlation coefficient matrix of the scenario of power station k; is the F-norm of the difference between the spatial correlation coefficient matrix of the historical data at time t and the spatial correlation coefficient matrix of the scene at time t;

[0290] (2) Initialize the individual

[0291] A population of size (N, D) is randomly generated, where N is the number of individuals in the population, and D is the number of components of the individual. Each individual in the population contains 4 components: the first two components are scene numbers, representing the scene where the prediction error is located; the third component is the power station number, representing the type of power station; the fourth component is the time number, representing the moment where the prediction error is located. For any individual, the type and moment of the power station are first determined according to the third and fourth components, and the data on the two scenes are swapped according to the first two components, and then the objective function is calculated while the rest of the data remains unchanged; if the objective function becomes smaller, the swap is maintained and the scene order is changed; if the objective function does not become smaller, the swap is restored and the scene order remains unchanged; when all individuals in the population have completed comparing the objective function and choosing whether to swap, the population enters the next generation through differential mutation and crossover operations, and the i-th individual No. i As shown in the following formula:

[0292] No i ={No i,1 ,No i,2 ,No i,3 ,No i,4} (10)

[0293] Among them, No i,1 Indicates the value of the first component of the i-th individual; No i,2 Indicates the value of the second component of the i-th individual; No i,3 Indicates the value of the third component of the i-th individual; No i,4 represents the value of the jth component of the i-th individual; i = 1, 2, ... N;

[0294] The value of the jth component of the i-th individual is calculated as follows:

[0295] No i,j =round(rand(0,1)·(U j -L j )+L j ) (11)

[0296] Among them, No i,j is the value of the jth component of the i-th individual; round(·) is the rounding function; rand(0,1) represents a random number in the interval [0,1]; L j is the lower bound of the jth component of the i-th individual; U j is the upper bound of the jth component of the i-th individual; j = 1, 2, 3, 4;

[0297] (3) Differential mutation operation

[0298] The differential mutation operation first uses a random method to select a series of different individual vectors from the population in the previous generation to generate a differential vector. Then, the differential vector is scaled and superimposed with the basis vector to generate a mutant individual. The jth mutation component of the i-th individual in the G-th generation is generated by the following formula:

[0299]

[0300] in, is the value of the jth variation component of the i-th individual in the G-th generation; is the value of the jth component of the r1th individual in the G-1th generation; is the value of the jth component of the r2th individual in the G-1th generation; is the value of the jth component of the r3th individual in the G-1th generation; F is the scaling factor, which is a random number in the interval [0,1], and r1, r2, and r3 are mutually different integers randomly selected in [0,N];

[0301] In the mutation operation, the DE algorithm determines whether the new individuals generated by the differential mutation operation are beyond the boundary. If they are beyond the boundary, the following method is used to process the mutated individuals:

[0302]

[0303] (4) Crossover operation

[0304] The crossover method commonly used in DE algorithm is as follows:

[0305]

[0306] in, is the value of the jth component of the i-th individual in the G-th generation; is the value of the jth component of the i-th individual in the G-1th generation; CR is the crossover probability;

[0307] (5) After all individuals in the new generation population have completed the comparison of the objective function and the selection of whether to swap, steps (3) to (4) are repeated until the population evolves to the set number of iterations. The final scene is the sorted water-landscape scene.

[0308] Steps 1-5: Calculate the randomness and correlation indexes of the water-landscape scenes before and after sorting, and conduct scene evaluation;

[0309] To measure the effectiveness of the spatiotemporal correlation ranking, the mean absolute error was chosen as the randomness indicator and the difference score as the correlation indicator;

[0310] The calculation formula for the mean absolute error of power station k is:

[0311]

[0312] in, is the historical actual value of the prediction error of power station k at time t, x s,k,t is the prediction error of power station k at time t under scenario s;

[0313] The time difference fraction of station k is:

[0314]

[0315] The spatial difference score of station k is:

[0316]

[0317] in, The weight coefficient of power station k between time t and time w is represented by the absolute value of the time correlation coefficient of the historical data of power station k at time t and time w; is the historical actual value of the prediction error of power station k at time t; x s,k,w is the prediction error of power station k at time w under scenario s; represents the weight coefficient between power station k and power station l at time t, which is represented by the absolute value of the spatial correlation coefficient of the historical data of power station k and power station l at time t; is the historical actual value of the prediction error of power station l at time t; x s,l,t is the prediction error of power station l at time t under scenario s; γ represents the difference fractional order; It refers to the difference between the historical forecast error of power station k at time t and time w; It refers to the difference between the prediction error of power station k at time t and time w under scenario s; It refers to the difference in historical prediction errors between plant k and plant l at time t; It refers to the difference in prediction error between plant k and plant l at time t under scenario s;

[0318] The spatial difference scores of the wind power prediction errors before and after sorting are as follows: Figure 2 As shown in the figure, it can be seen that the spatial difference scores of most moments after sorting are smaller than those before sorting, indicating that the sorted scenes can better describe the spatial correlation of the scenery.

[0319] Table 3 shows the absolute error and temporal variance scores for the hydropower, wind, and solar power forecast error evaluation. As can be seen, with the exception of hydropower station 1, the average absolute error of the ranked hydropower stations is reduced by approximately 10%; the average absolute error of all ranked photovoltaic power stations is reduced by approximately 10%; and the average absolute error of all ranked wind farms is reduced by approximately 20%. With the exception of hydropower stations 3, 4, and wind farm 3, the temporal variance scores of the ranked stations are reduced by approximately 15% to 25%. This demonstrates that the scenarios ranked by spatiotemporal correlation better capture the randomness and temporal correlation of hydropower, wind, and solar power.

[0320] Table 3 Comparison of randomness and time correlation of scenes before and after sorting

[0321]

[0322] Steps 1-6: Use the K-means clustering algorithm to reduce the sorted scenes to obtain typical scenes, that is, scenes that take into account the temporal and spatial correlation of water, scenery and light.

[0323] Adding the sorted scenarios to the predicted values ​​yields the hydropower inflow scenarios and wind and solar output:

[0324]

[0325]

[0326]

[0327] in, is the output of PV station g at time t under scenario s, g = 1, 2, ..., G, G is the number of PV stations in the hydro-wind-solar system, and G is a positive integer; is the output of wind farm w at time t under scenario s, w=1,2,...,W, W is the number of wind farms in the hydro-wind-solar system, and W is a positive integer; R is the interval water inflow of hydropower station i at time t under scenario s, i=1,2,...,I, I

[0328] s,i,t

[0329] is the number of cascade hydropower stations in the water-wind-solar system, I is a positive integer, is the predicted output of photovoltaic power station g at time t, is the predicted output of wind farm w at time t, is the predicted interval water inflow of cascade hydropower station i at time t;

[0330] The K-means clustering algorithm is used to reduce the hydropower inflow scenario and wind and solar power output scenario to obtain the spatiotemporal correlation scenario of water, wind and solar power.

[0331] Figure 3The final output scene diagram of photovoltaic power station 1 is obtained. Figure 4 The final wind farm 1 output scene diagram is: Figure 5 The final water inflow scene diagram of the cascade hydropower station 4 is obtained. It can be seen from the figure that the scene fluctuation trend is consistent with the actual fluctuation trend and the actual value can be well covered in the generated scene set, which proves that the scene generated by this method can effectively represent the actual situation.

[0332] Step 2: Obtain the effective moment of inertia of cascade hydropower and wind turbines under various scenarios, and propose effective moment of inertia constraints for the hydropower-wind-solar hybrid system;

[0333] Step 2-1: Utilize the energy absorbed or released by the rotors of the cascade hydropower units and the wind turbines when the power system frequency changes to obtain the effective rotational inertia of the cascade hydropower units and the wind turbines;

[0334] Cascade hydropower effective moment of inertia E h The calculation formula is as follows:

[0335]

[0336] Among them, H is the inherent inertia time constant of the hydropower unit itself; A is the capacity of the hydropower unit; ω is the angular velocity of the hydropower unit; ω n is the rated angular velocity of the hydropower unit; ω min is the minimum angular velocity allowed by the power system; ω max is the maximum angular velocity allowed by the power system;

[0337] Effective moment of inertia of fan E W With wind speed V W The relationship is as follows:

[0338]

[0339] Among them, J w is the moment of inertia of the fan rotor; opt is the optimal tip speed ratio; R is the fan radius; ω r_lim is the permissible speed limit of the fan rotor; when the frequency rises, it is 0.7pu, and when the frequency falls, it is 1.3pu; ω max is the maximum speed of wind turbine connected to the grid; ω s V is the speed of the fan entering the constant speed zone; Wm V is the lower limit wind speed for the wind turbine to enter the maximum power point tracking area; Ws V is the lower limit wind speed for the fan to enter the constant speed zone; Wn is the rated wind speed of the fan; V Woff is the fan cut-out wind speed, V w refers to wind speed;

[0340] Utilize wind power output P WT Calculate the wind speed V w , the relationship between wind power output and wind speed is as follows:

[0341]

[0342] Among them, P WTn is the rated output power of the fan; V Win is the wind turbine cut-in speed;

[0343] Final effective moment of inertia of the fan E W The calculation is as follows:

[0344]

[0345] in,

[0346] Among them, P wt is the per-unit value of wind power output, and R refers to the radius of the wind turbine.

[0347] Step 2-2: Based on the condition that the minimum effective moment of inertia provided by the hydropower, wind-solar system should be greater than the effective moment of inertia required by the power grid, establish the effective moment of inertia constraint of the hydropower, wind-solar system;

[0348] The effective moment of inertia constraint of the established water-wind-solar system is:

[0349]

[0350]

[0351] Among them, u i,n,t is the start-stop state variable of unit n of hydropower station i at time t, 1 represents the start state, and 0 represents the stop state; is the effective moment of inertia of unit n of hydropower station i at time t when the frequency rises; is the effective moment of inertia of unit n of hydropower station i at time t when the frequency decreases; is the effective moment of inertia of the mth wind turbine in wind farm w at time t in scenario s when the frequency rises; is the effective moment of inertia of the mth wind turbine in wind farm w at time t in scenario s when the frequency drops; is the effective moment of inertia required by the grid at time t when the frequency rises; is the effective moment of inertia required by the grid at time t when the frequency drops; I is the number of cascade hydropower stations in the water-wind-solar system, N i is the number of units in hydropower station i, W is the number of wind farms in the hydro-wind-solar system, M w is the number of wind turbines in wind farm w.

[0352] Step 3: Establish a day-ahead dispatch model for hydropower, wind-solar power systems that takes into account the uncertainty of hydropower, wind-solar power systems and the flexible regulation capability of cascaded hydropower.

[0353] In step 3, the constraints of the day-ahead dispatch model of the water, wind and solar system established include the effective moment of inertia constraint of the water, wind and solar system, water balance constraint, water level constraint, initial and final water level constraint, reservoir capacity-water level constraint, tail water level-outflow constraint, unit output constraint, unit power generation flow constraint, unit vibration zone constraint, unit start-up and shutdown duration constraint, unit output ramp constraint, unit power generation head constraint, head loss constraint, unit dynamic characteristic relationship, section transmission constraint, and power generation plan constraint.

[0354] When establishing the day-ahead dispatch model of the water, wind and solar system, the details are as follows:

[0355] Step 3-1: Calculate the flexible adjustment requirements of the hydro-wind-solar hybrid system based on the difference between the predicted wind and solar output and the output under scenario s. The calculation formula is as follows:

[0356]

[0357]

[0358]

[0359]

[0360] in, Flexible upward adjustment requirements at time t under scenario s; The flexible downward adjustment requirement at time t under scenario s; The flexible upward adjustment demand of the photovoltaic power station at time t under scenario s; is the flexible upward adjustment demand of the wind farm at time t under scenario s, The flexible downward adjustment demand of the PV power station at time t under scenario s; is the flexible downward adjustment demand of the wind farm at time t under scenario s;

[0361] Step 3-2: Calculate the flexible adjustment capability of the hydropower-wind-solar hybrid system based on the ramping capability and output limit of the hydropower unit. The calculation formula is as follows:

[0362]

[0363] in, is the flexible upward adjustment capability of cascade hydropower at time t; is the flexible downward adjustment capability of cascade hydropower at time t; is the gradeability of hydropower station i; is the output of hydropower station i at time t; is the upper limit of the output of hydropower station i at time t; is the lower limit of the output of hydropower station i at time t;

[0364] Step 3-3: Establish an objective function based on the flexible adjustment requirements and flexible adjustment capabilities. The objective function calculation formula is as follows:

[0365]

[0366]

[0367]

[0368]

[0369] in, The expected flexibility increase deficiency refers to the expected difference between the flexible increase demand and the flexible increase capacity at time t due to insufficient increase capacity. The expected flexibility downsizing shortage refers to the expected difference between the flexible downsizing demand and the flexible downsizing capacity at time t due to insufficient downsizing capacity. is the amount of water abandoned by hydropower station i at time t under scenario s; is the expected amount of water abandoned by the cascade hydropower at time t; P(s) refers to the probability of scenario s;

[0370] Step 3-4: Establish the constraints of the water, wind and solar system, as follows;

[0371] (1) Water balance constraints

[0372]

[0373] Among them, V s,i,t+1 is the storage capacity of hydropower station i at time t+1 under scenario s; V s,i,t is the storage capacity of hydropower station i at time t under scenario s; I s,i,t is the inflow flow of hydropower station i at time t under scenario s; Q s,i,t is the outflow of hydropower station i at time t under scenario s; Δt is the scheduling time interval; τ i-1 is the water flow lag time between hydropower station i-1 and hydropower station i; is the hydropower station i-1 at time t-τ under scenario s i-1 flow rate; R s,i,t is the interval flow between hydropower station i-1 and hydropower station i at time t under scenario s; is the power generation flow of hydropower station i at time t under scenario s;

[0374] (2) Water level constraint

[0375]

[0376] in, is the reservoir water level of hydropower station i at time t under scenario s; Z i,t is the lower limit of the water level of hydropower station i at time t; is the upper limit of the water level of hydropower station i at time t;

[0377] (3) Initial and final water level constraints

[0378]

[0379]

[0380] in, is the initial water level in the scheduling period; is the final water level in the scheduling period; ΔZ is the allowable deviation, Refers to the initial water level of cascade hydropower station i in the scheduling period under scenario s, It refers to the water level at the end of the dispatching period of cascade hydropower station i under scenario s;

[0381] (4) Reservoir capacity-water level constraints

[0382]

[0383] in, is the storage capacity-water level function of hydropower station i;

[0384] (5) Tailwater level-outflow constraint

[0385]

[0386] in, is the outflow-tailwater level function of hydropower station i; is the tailwater level of power station i at time t under scenario s;

[0387] (6) Unit output constraints

[0388]

[0389] Among them, P i,n,t is the output of unit n of hydropower station i at time t under scenario s, is the upper limit of the output of unit n of hydropower station i, P i,n is the lower limit of the output of unit n of hydropower station i. The output of hydropower station i is expressed as:

[0390]

[0391] Among them, Ni represents the number of units of hydropower station i;

[0392] (7) Unit power generation flow constraints

[0393]

[0394] in, is the power generation flow of unit n of hydropower station i at time t under scenario s; is the upper limit of power generation flow of unit n of hydropower station i; is the lower limit of the power generation flow of unit n of hydropower station i. The power generation flow of hydropower station i is expressed as:

[0395]

[0396] (8) Unit vibration zone constraints

[0397]

[0398] in, is the output upper limit of vibration zone k of unit n of hydropower station i; is the lower limit of the output of unit n in vibration zone k of hydropower station i;

[0399] (9) Unit start-up and shutdown duration constraints:

[0400]

[0401] in, is the startup operation variable of unit n of hydropower station i at time t, 1 indicates startup operation; is the shutdown operation variable of unit n of hydropower station i at time t, 1 indicates shutdown operation; is the minimum startup duration of unit n of hydropower station i; is the minimum shutdown duration of unit n of hydropower station i; is the maximum number of startup times of unit n of hydropower station i within the scheduling period;

[0402] (10) Unit output ramp constraints:

[0403] -ΔP i,n ≤P i,n,t+1 -P i,n,t ≤ΔP i,n (49)

[0404] Among them, P i,n,t+1 is the output of unit n of hydropower station i at time t+1 under scenario s, ΔP i,n is the ramping capability of unit n of hydropower station i;

[0405] (11) Unit generating head constraints:

[0406]

[0407] in, is the reservoir water level of hydropower station i at time t-1 under scenario s, H s,i,n,t is the generating head of unit n of hydropower station i at time t under scenario s; is the head loss of unit n of hydropower station i at time t under scenario s;

[0408] (12) Head loss constraint:

[0409]

[0410] Among them, a i is the head loss coefficient of power station i; b i is the head loss constant of power station i;

[0411] (13) Relationship of unit power characteristics:

[0412]

[0413] in, is the head-flow-output function of unit n of hydropower station i;

[0414] (14) Section transmission constraints;

[0415]

[0416] Among them, I o is the number of hydropower stations included in section o; W o is the number of wind farms included in section o; G o is the number of photovoltaic power stations included in section o; is the transmission capacity of section o, is the output of wind farm w at time t under scenario s;

[0417] (15) Power generation plan constraints:

[0418]

[0419] Among them, ε is a smaller deviation threshold, It represents the power generation plan issued by the power grid, G is the number of photovoltaic power stations in the water-wind-solar system, I is the number of cascade hydropower stations in the water-wind-solar system, and W is the number of wind farms in the water-wind-solar coefficient.

[0420] The present invention also includes a day-ahead scheduling model for water, wind, and solar power taking into account effective moment of inertia and uncertainty, wherein the objective function of the model is:

[0421]

[0422]

[0423]

[0424]

[0425] in, The expected flexibility increase deficiency refers to the expected difference between the flexible increase demand and the flexible increase capacity at time t due to insufficient increase capacity. The expected flexibility downsizing shortage refers to the expected difference between the flexible downsizing demand and the flexible downsizing capacity at time t due to insufficient downsizing capacity. is the amount of water abandoned by hydropower station i at time t under scenario s; is the expected amount of water abandoned by cascade hydropower at time t; P(s) refers to the probability of scenario s.

[0426] The constraints of the model include the effective moment of inertia constraint of the water-wind-solar system, water balance constraint, water level constraint, initial and final water level constraint, reservoir capacity-water level constraint, tailwater level-outflow constraint, unit output constraint, unit power generation flow constraint, unit vibration zone constraint, unit start-up and shutdown duration constraint, unit output ramp constraint, unit power generation head constraint, head loss constraint, unit dynamic characteristic relationship, section transmission constraint, and power generation plan constraint; the details are as follows;

[0427] 1) The effective moment of inertia constraint of the water-wind-solar system is:

[0428]

[0429]

[0430] 2) Water balance constraints are:

[0431]

[0432] Among them, V s,i,t+1 is the storage capacity of hydropower station i at time t+1 under scenario s; V s,i,t is the storage capacity of hydropower station i at time t under scenario s; I s,i,t is the inflow flow of hydropower station i at time t under scenario s; Q s,i,t is the outflow of hydropower station i at time t under scenario s; Δt is the scheduling time interval; τ i-1 is the water flow lag time between hydropower station i-1 and hydropower station i; is the hydropower station i-1 at time t-τ under scenario s i-1 flow rate; R s,i,tis the interval flow between hydropower station i-1 and hydropower station i at time t under scenario s; is the power generation flow of hydropower station i at time t under scenario s;

[0433] 3) The water level constraint is:

[0434]

[0435] in, is the reservoir water level of hydropower station i at time t under scenario s; Z i,t is the lower limit of the water level of hydropower station i at time t; is the upper limit of the water level of hydropower station i at time t;

[0436] 4) The initial and final water level constraints are:

[0437]

[0438]

[0439] in, is the initial water level in the scheduling period; is the final water level in the scheduling period; ΔZ is the allowable deviation, Refers to the initial water level of cascade hydropower station i in the scheduling period under scenario s, It refers to the water level at the end of the dispatching period of cascade hydropower station i under scenario s;

[0440] 5) Reservoir capacity-water level constraint is:

[0441]

[0442] in, is the storage capacity-water level function of hydropower station i;

[0443] 6) Tailwater level-outflow constraint is:

[0444]

[0445] in, is the outflow-tailwater level function of hydropower station i; is the tailwater level of power station i at time t under scenario s;

[0446] 7) Unit output constraints are:

[0447]

[0448] Among them, P i,n,t is the output of unit n of hydropower station i at time t under scenario s, is the upper limit of the output of unit n of hydropower station i, P i,n is the lower limit of the output of unit n of hydropower station i. The output of hydropower station i is expressed as:

[0449]

[0450] Among them, N i represents the number of units of hydropower station i;

[0451] 8) The power generation flow constraint of the unit is:

[0452]

[0453] in, is the power generation flow of unit n of hydropower station i at time t under scenario s; is the upper limit of power generation flow of unit n of hydropower station i; is the lower limit of the power generation flow of unit n of hydropower station i. The power generation flow of hydropower station i is expressed as:

[0454]

[0455] 9) The vibration zone constraints of the unit are:

[0456]

[0457] in, is the output upper limit of vibration zone k of unit n of hydropower station i; is the lower limit of the output of unit n in vibration zone k of hydropower station i;

[0458] 10) The unit start-up and shutdown duration constraints are:

[0459]

[0460] in, is the startup operation variable of unit n of hydropower station i at time t, 1 indicates startup operation; is the shutdown operation variable of unit n of hydropower station i at time t, 1 indicates shutdown operation; is the minimum startup duration of unit n of hydropower station i; is the minimum shutdown duration of unit n of hydropower station i; is the maximum number of startup times of unit n of hydropower station i within the scheduling period;

[0461] 11) The unit output ramp constraint is:

[0462] -ΔP i,n ≤P i,n,t+1 -P i,n,t ≤ΔP i,n (49)

[0463] Among them, P i,n,t+1 is the output of unit n of hydropower station i at time t+1 under scenario s, ΔP i,n is the ramping capability of unit n of hydropower station i;

[0464] 12) The unit's generating head constraint is:

[0465]

[0466] in, is the reservoir water level of hydropower station i at time t-1 under scenario s, H s,i,n,t is the generating head of unit n of hydropower station i at time t under scenario s; is the head loss of unit n of hydropower station i at time t under scenario s;

[0467] 13) The head loss constraint is:

[0468]

[0469] Among them, a i is the head loss coefficient of power station i; b i is the head loss constant of power station i;

[0470] 14) The relationship between the unit's dynamic characteristics is:

[0471]

[0472] in, is the head-flow-output function of unit n of hydropower station i;

[0473] 15) Section transmission constraints are:

[0474]

[0475] Among them, I o is the number of hydropower stations included in section o; W o is the number of wind farms included in section o; G o is the number of photovoltaic power stations included in section o; is the transmission capacity of section o, is the output of wind farm w at time t under scenario s;

[0476] 16) The power generation plan constraints are:

[0477]

[0478] Among them, ε is a smaller deviation threshold, It represents the power generation plan issued by the power grid, G is the number of photovoltaic power stations in the water-wind-solar system, I is the number of cascade hydropower stations in the water-wind-solar system, and W is the number of wind farms in the water-wind-solar coefficient.

[0479] Furthermore, Figure 6 The effective rotational inertia provided by the water, wind and solar power systems is shown. As can be seen from the figure, the scheduling plan formulated by this method can meet the problem of the decrease in the effective rotational inertia of the system caused by the uncertainty of wind power in each time period, and effectively ensure the frequency stability capability of the power grid. Figure 7 The relationship between the regulation capacity and regulation demand of the water, wind and solar systems is demonstrated, indicating that the scheduling plan formulated by this method can meet the regulation demand in all time periods and has a large flexibility margin. Figure 8 Shows the operation status of the unit 1 of the hydropower station, Figure 9 Shows the operation status of the unit 2 of the hydropower station, Figure 10 Shows the operation status of unit 3 of hydropower station, Figure 11 The operating status of the units of Hydropower Station 4 is shown. To ensure the effective rotational inertia level of the water-wind-solar system, Hydropower Station 1 and Hydropower Station 3 are all powered on during the entire scheduling cycle, and Hydropower Station 2 is fully powered on most of the time except in the early scheduling period when the water inflow is small; Hydropower Station 4 is affected by the water inflow, and flexibly adjusts the start and stop of the units in the early scheduling period, and fully powers on in the middle and late scheduling periods. Figure 12 The relationship between the output of the hydropower, wind and solar power systems and the planned output of the power grid is demonstrated. The actual output process of the hydropower, wind and solar power systems deviates slightly from the given cascade power generation plan, and the power supply support function is reliable. From the output process point of view, after considering the start and stop constraints of the units, the output of each hydropower station is stable and there are no frequent fluctuations, which meets the actual operation requirements of the power station.

[0480] In summary, this method fully considers the temporal and spatial correlations of hydropower, wind power, and solar power, generates a more realistic scenario set, and provides a data basis for the specification of scheduling plans. At the same time, it quantitatively calculates the effective rotational inertia of cascade hydropower and wind turbines under different frequency regulation requirements, and proposes effective rotational inertia constraints for the hydropower, wind power, and solar power complementary system. On the basis of meeting the effective rotational inertia requirements of the power grid, it fully utilizes the flexible regulation capabilities of cascade hydropower and improves the frequency stability capability of the power grid.

[0481] Figure 13 An example of a physical structure diagram of an electronic device is shown below. Figure 13 As shown, the electronic device includes: a processor 2, a communication interface 4, a memory 1, and a communication bus 3. The processor 2, the communication interface 4, and the memory 1 communicate with each other via the communication bus 3. The processor 2 can call the logic instructions in the memory 1 to execute the day-ahead scheduling model for water, wind, and solar power that takes into account the effective moment of inertia and uncertainty.

[0482] In addition, the logical instructions in the memory 1 can be implemented in the form of a software functional unit and can be stored in a computer-readable storage medium when sold or used as an independent product. Based on this understanding, the technical solution of the present invention can be embodied in the form of a software product, which is stored in a storage medium and includes a number of instructions for causing a computer device (which can be a personal computer, server, or network device, etc.) to execute all or part of the steps of the method described in the present invention. The storage medium includes various media that can store program code, such as a USB flash drive, a mobile hard disk, a read-only memory (ROM), a random access memory (RAM), a magnetic disk, or an optical disk.

[0483] On the other hand, the present invention also provides a computer program product, which includes a computer program. The computer program can be stored on a non-transitory computer-readable storage medium. When the computer program is executed by a processor, the computer can execute the method for establishing the water-wind-solar day-ahead scheduling model and the water-wind-solar day-ahead scheduling model considering effective rotational inertia and uncertainty.

[0484] On the other hand, the present invention also provides a non-transitory computer-readable storage medium having a computer program stored thereon. When the computer program is executed by a processor, it is implemented to execute the method for establishing the water-wind-solar day-ahead scheduling model and the water-wind-solar day-ahead scheduling model considering effective rotational inertia and uncertainty.

Claims

1. A method for establishing a day-ahead scheduling model for water, wind, and solar power, characterized in that: It includes the following steps: Step 1: To address the uncertainty of water, wind and solar power prediction errors, generate scenarios that take into account the spatiotemporal correlation of water, wind and solar power; Step 2: Obtain the effective moment of inertia of cascade hydropower and wind turbines in each scenario, and establish effective moment of inertia constraints for the hydropower, wind, and solar power systems; Step 3: Establish a day-ahead dispatch model for hydropower, wind-solar power systems that takes into account the uncertainty of hydropower, wind-solar power systems and the flexible regulation capability of cascaded hydropower. In step 3, the constraints of the day-ahead dispatch model of the water, wind and solar power systems established include the effective moment of inertia constraint of the water, wind and solar power systems, water balance constraint, water level constraint, initial and final water level constraint, reservoir capacity-water level constraint, tailwater level-outflow constraint, unit output constraint, unit power generation flow constraint, unit vibration zone constraint, unit start-up and shutdown duration constraint, unit output ramp constraint, unit power generation head constraint, head loss constraint, unit dynamic characteristic relationship, section transmission constraint, and power generation plan constraint; The step 1 specifically includes the following steps: Step 1-1: For the target scheduling time, use the kernel distribution function to fit the probability distribution function of the water, wind and solar power prediction error, perform Latin hypercube sampling on the probability distribution function of the water, wind and solar power prediction error, and generate the water, wind and solar power scenario; Step 1-2: Use the Pearson correlation coefficient to obtain the time correlation coefficient matrix of historical data and the time correlation coefficient matrix of water-wind-light scene; Step 1-3: Use the connection function to obtain the spatial correlation coefficient matrix of historical data and the spatial correlation coefficient matrix of water-landscape scenes; Steps 1-4: Based on the differential evolution algorithm, the water-scenery scenes are sorted with the goal of minimizing the difference between the temporal correlation coefficient matrix of the historical data and the spatial correlation coefficient matrix of the historical data and the water-scenery scenes, and the sorted water-scenery scenes are obtained; Steps 1-5: Obtain randomness and correlation indices of the water-landscape scenes before and after sorting, and conduct scene evaluation; Step 1-6: Use the K-means clustering algorithm to reduce the sorted scenes to obtain typical scenes, that is, scenes that take into account the temporal and spatial correlation of water, scenery and light; Step 2 specifically includes the following steps: Step 2-1: Utilize the energy absorbed or released by the rotors of the cascade hydropower units and the wind turbines when the power system frequency changes to obtain the effective rotational inertia of the cascade hydropower units and the wind turbines; Step 2-2: Based on the condition that the minimum effective moment of inertia provided by the hydropower, wind-solar system should be greater than the effective moment of inertia required by the power grid, establish the effective moment of inertia constraint of the hydropower, wind-solar system; In step 2-1, Effective moment of inertia of cascade hydropower The calculation formula is as follows: （21）； in, H is the inherent inertia time constant of the hydropower unit itself; A is the capacity of the hydropower unit; is the angular velocity of the hydropower unit; is the rated angular velocity of the hydropower unit; is the minimum angular velocity allowed by the power system; is the maximum angular velocity allowed by the power system; Fan effective moment of inertia and wind speed The relationship is as follows: （22）； in, is the moment of inertia of the fan rotor; is the optimal tip speed ratio; R is the fan radius; The allowable speed limit of the fan rotor; The maximum speed of the wind turbine when connected to the grid; The speed at which the fan enters the constant speed zone; The lower limit wind speed for the wind turbine to enter the maximum power point tracking area; The lower limit wind speed for the fan to enter the constant speed zone; is the rated wind speed of the fan; Cut out wind speed for the fan, refers to wind speed; Utilizing wind power output Calculate wind speed , the relationship between wind power output and wind speed is as follows: （23）； in, is the rated output power of the fan; is the wind turbine cut-in speed; Final effective moment of inertia of the fan The calculation is as follows: （24）； in, (25); in, is the per-unit value of wind power output, R Refers to the fan radius.

2. The method according to claim 1, characterized in that In step 1-1, when obtaining the water-wind-solar forecast error, the specific steps are as follows: Arrange the water, wind, and solar power forecast values ​​in ascending order and divide them into several forecast boxes according to the equal frequency principle. Each box has a corresponding forecast error. For the forecast value at the target time, calculate the kernel distribution function of the corresponding forecast box and obtain the probability distribution function of the water, wind, and solar power forecast error: （1）； in, For power stations k exist t The probability distribution function of the forecast error at time , M is the number of data in the prediction box, M The value is a positive integer. h k,t For power stations k exist t The bandwidth of the kernel distribution function at time , is the kernel function, For power stations k exist t The forecast error variable at time t; For power stations k exist t The first prediction box at the time The size of the prediction error, m= 1 , 2 ,…, M , k= 1 , 2 ,…,K 1; K 1 is the total number of power stations included in the hydro-wind-solar system, which is a positive integer; t= 1 , 2 ,…,T 1, T 1 is the total number of scheduling times; The probability interval Divide evenly S intervals, randomly select a point in each interval as the value of the probability distribution function of the water-wind-light prediction error, and calculate the corresponding prediction error value: （2）； in, For the scene s Lower Power Station k exist t The prediction error at time s= 1 , 2 ,…,S , S is the number of generated scenes, For power stations k exist t The inverse function of the probability distribution function of the forecast error at time , For interval s Lower Power Station k exist t The quantile drawn at the moment.

3. The method according to claim 1, characterized in that In step 1-2, the Pearson correlation coefficient is used to obtain the time correlation coefficient matrix of the historical data and the time correlation coefficient matrix of the water-wind-light scene, where: Time correlation coefficient matrix of historical data Elements in The calculation is as follows: （3）； In formula (3), For power stations k exist i Moment and j Time correlation coefficient of historical data at each moment; N 1 is the total number of historical data; It's a power station k exist i The historical forecast error at time values; It's a power station k exist j The historical forecast error at time values; It's a power station k The historical forecast error is i The average value of the moment; It's a power station k The historical forecast error is j The average value of the moment; Time correlation coefficient matrix of water-landscape scene Elements in The calculation is as follows: （4）； in, For power stations k exist i Moment and j Time correlation coefficient of water scene at each moment; It's a scene s Lower Power Station k exist i The prediction error at the moment; It's a scene s Lower Power Station k exist j The prediction error at the moment; It's a power station k The prediction error scenario is i The average value of the moment; It's a power station k The prediction error scenario is j The average value of the moment; S is the number of scenes generated.

4. The method according to claim 1, wherein In steps 1-3, the connection function is used to obtain the spatial correlation coefficient matrix of historical data and the spatial correlation coefficient matrix of water-landscape scenes, where: choose t - Copula As the connection function, and using Matlab fitting to obtain formula (5): （5）； in, for t Joint probability distribution function of moment-to-moment prediction errors; For power station 1 t The forecast error variable at time t; For Power Station 2 t The forecast error variable at time t; For power stations exist t The forecast error variable at time t; The degrees of freedom are v t , the correlation coefficient matrix is t of n dimension t distribution function; The degrees of freedom are v t of t Inverse function of distribution; For power station 1 t Probability distribution function of the forecast error at time; For Power Station 2 t Probability distribution function of the forecast error at time; For power stations exist t Probability distribution function of the forecast error at time; Selection Matrix t As the spatial correlation coefficient matrix of historical data; Spatial correlation coefficient matrix of water scene Elements in The calculation formula is as follows: （6）； in, for t Shike Power Station k With power station l The spatial correlation coefficient of the water scene; For the scene s Down t The multidimensionality of time t Distribution a The value of a random variable of dimension ; For the scene s Down t The multidimensionality of time t Distribution b The value of a random variable of dimension ; For all scenarios t The multidimensionality of time t Distribution a The mean value of a random variable of dimension ; For all scenarios t The multidimensionality of time t Distribution b The mean value of a random variable of dimension ; S is the number of generated scenes; From formula (7) and formula (8), we can obtain and as follows: （7）； （8）； in, For power stations a exist t Probability distribution function of time; For the scene s Lower Power Station a exist t The prediction error at the moment; For power stations b exist t Probability distribution function of time; For the scene s Lower Power Station b exist t The prediction error at time The degrees of freedom are of t The inverse function of the distribution.

5. The method according to claim 1, wherein In steps 1-4, the water-wind-light scenes are sorted based on the differential evolution algorithm with the goal of minimizing the difference between the temporal correlation coefficient matrix of the historical data and the water-wind-light scenes, as well as the difference between the spatial correlation coefficient matrix of the historical data and the water-wind-light scenes, to obtain the sorted water-wind-light scenes. The specific steps are as follows: (1) Establish the objective function of the differential evolution algorithm: （9）； in, K 1 is the total number of power stations included in the hydro-wind-solar system, T 1 is the total number of scheduling times, For power stations k The time correlation coefficient matrix of historical data and power station k The difference of the time correlation coefficient matrix of the scene F norm; for t The spatial correlation coefficient matrix of the historical data at the time t The difference of the spatial correlation coefficient matrix of the scene at the moment F norm; (2) Initialize the individual: Randomly generate a scale of ( N , D ) population, N is the number of individuals in the population, D is the number of components of the individual, and each individual in the population contains 4 components: the first two components are scene numbers, representing the scenes where the prediction error is located; the third component is the power station number, representing the type of power station; the fourth component is the time number, representing the moment where the prediction error is located. For any individual, first determine the type and moment of the power station according to the third and fourth components, and swap the data on the two scenes according to the first two components, and then calculate the objective function while keeping the rest of the data unchanged; if the objective function becomes smaller, keep the swap and change the scene order; if the objective function does not become smaller, restore the swap and keep the scene order unchanged; when all individuals in the population have completed comparing the objective function and choosing whether to swap, the population enters the next generation through differential mutation and crossover operations, and the first individual in the population i individual As shown in the following formula: （10）； in, Indicates the i The value of the first component of the individual; Indicates the i The value of the second component of the individual; Indicates the i The value of the third component of the individual; Indicates the i Individual j The value of the component; i =1,2,… N ; No. i Individual j The value of each component is calculated as follows: （11）； in, For the i Individual j The value of the component; is the rounding function; Represents a random number in the interval [0,1]; It is i Individual j The lower bound of the components; It is i Individual j The upper bound of the components; j =1,2,3,4; (3) Differential mutation operation: The differential mutation operation first uses a random method to select a series of different individual vectors from the population to generate differential vectors, and then scales the differential vectors and superimposes them with the basis vectors to generate mutant individuals. G Daidi i Individual j The components of variation are generated by the following formula: （12）； in, For the G Daidi i Individual j The value of each variation component; For the G -1st generation Individual j The value of the component; For the G -1st generation Individual j The value of the component; For the G -1st generation Individual j The value of the component; F is the scaling factor, which is a random number in the interval [0,1]. 、 、 For [0, N ] are randomly selected distinct integers within ; In the mutation operation, DE The algorithm determines whether the new individuals generated by the differential mutation operation are beyond the boundary. If they are beyond the boundary, the following method is used to process the mutated individuals: （13）； (4) Crossover operation: DE The commonly used crossover method of the algorithm is as follows: （14）； in, For the G Daidi i Individual j The value of the component; For the G -1st generation i Individual j The value of the component; CR is the crossover probability; (5) After all individuals in the new generation population have completed the comparison of the objective function and the selection of whether to swap, steps (3) to (4) are repeated until the population evolves to the set number of iterations. The final scene is the sorted water-landscape scene.

6. The method according to claim 1, wherein In steps 1-5, the randomness and correlation indices of water, wind and light before and after sorting are calculated to conduct scenario evaluation; To measure the effectiveness of the spatiotemporal correlation ranking, the mean absolute error was chosen as the randomness indicator and the difference score as the correlation indicator; power station k The formula for calculating the mean absolute error is: （15）； in, For power stations k The prediction error is t The historical actual value at the moment, For the scene s Lower Power Station k exist t The prediction error at the moment; power station k The time difference score is: （16）； power station k The spatial difference score is: （17）； in, Indicates power station k exist t Moment and w The weight coefficient between the moments is determined by the power station k exist t Moment and w The absolute value of the time correlation coefficient of the historical data at the moment; For power stations k The prediction error is t The historical actual value at the moment; For the scene s Lower Power Station k exist w The prediction error at the moment; express t Shike Power Station k With power station l The weight coefficient between t Shike Power Station k With power station l The absolute value of the spatial correlation coefficient of the historical data is expressed; For power stations l The prediction error is t The historical actual value at the moment; For the scene s Lower Power Station l exist t The prediction error at the moment; γ represents the difference fraction order; refers to the power station k The historical forecast error is t The difference between moment and moment w; Refers to the scene s Lower Power Station k The prediction error is t Moment and w The difference in moments; refers to t Shike Power Station k With power station l differences in historical forecast errors; Refers to the scene s Down t Shike Power Station k With power station l The difference in prediction errors.

7. The method according to claim 1, characterized in that In steps 1-6, the K-means clustering algorithm is used to reduce the sorted scenes to obtain typical scenes, that is, scenes that take into account the spatiotemporal correlation between water, scenery and light; Adding the sorted scenarios to the predicted values ​​yields the hydropower inflow scenarios and wind and solar output: （18）； （19）； （20）； in, For the scene s Lower photovoltaic power station g exist t The effort of every moment, , G is the number of photovoltaic power stations in the water-wind-solar system, G Take a positive integer; For the scene s Downstream wind farm w exist t The effort of every moment, , W is the number of wind farms in the hydro-wind-solar system, W Take a positive integer; For the scene s Lower Hydropower Station i exist t The water comes in at the interval of time, , I is the number of cascade hydropower stations in the water-wind-solar system, I Take a positive integer, For photovoltaic power stations g exist t The predicted output of the moment, For wind farms w exist t The predicted output of the moment, Cascade hydropower stations i exist t The forecast interval of time is water inflow, For the scene s Lower Power Station k exist t The prediction error at the moment; use K The mean clustering algorithm is used to reduce the hydropower inflow scenario and wind and solar power output scenario to obtain the spatiotemporal correlation scenario of water, wind and solar power.

8. The method according to claim 1, characterized in that Step 2-2: The effective moment of inertia constraint of the established water-wind-solar system is: （26）； （27）； in, For hydropower stations i The unit n exist t The start / stop state variable at the moment, 1 indicates the start state, and 0 indicates the stop state; For hydropower stations when frequency rises i The unit n exist t The effective moment of inertia at the moment; For hydropower stations when frequency drops i The unit n exist t The effective moment of inertia at the moment; For the frequency increase scenario s Downstream wind farm w No. m A fan in t The effective moment of inertia at the moment; For the frequency drop scenario s Downstream wind farm w No. m A fan in t The effective moment of inertia at the moment; When the frequency rises, the power grid t The effective moment of inertia required at any given moment; When the frequency drops, the power grid t The effective moment of inertia required at any given moment; I is the number of cascade hydropower stations in the water-wind-solar system, It is a hydropower station i The number of units, W is the number of wind farms in the hydro-wind-solar system, It is a wind farm w The number of fans.

9. The method according to claim 1, characterized in that The step 3 specifically includes the following steps: Step 3-1: Based on the predicted value and scenario of wind and solar power output s The flexible adjustment requirements of the water-wind-solar complementary system are obtained by taking into account the output differences under different conditions; Step 3-2: Obtain the flexible adjustment capability of the hydropower-wind-solar hybrid system based on the ramping capability and output limit of the hydropower units; Step 3-3: Establish an objective function based on flexible adjustment requirements and flexible adjustment capabilities; Step 3-4: Establish the constraints of the water, wind and solar system.

10. The method according to claim 9, characterized in that In step 3-1, according to the predicted value of wind and solar power output and the scenario s The flexible adjustment requirements of the hydro-wind-solar hybrid system are obtained by the output difference under the following conditions. The calculation formula is as follows: （28）； （29）； （30）； （31）； in, For the scene s Down t Flexible adjustment requirements at any time; For the scene s Down t Flexible downward adjustment requirements at any time; For the scene s Down t Flexible adjustment requirements of photovoltaic power stations at all times; For the scene s Down t The flexible adjustment demand of wind farms at all times, For the scene s Down t Flexible adjustment requirements of photovoltaic power stations at all times; For the scene s Down t The flexible downward adjustment requirements of wind farms at all times; For photovoltaic power stations g exist t The predicted output of the moment, For the scene s Lower photovoltaic power station g exist t The effort of every moment, For wind farms w exist t The predicted output of the moment, For the scene s Downstream wind farm w exist t The effort of every moment, G is the number of photovoltaic power stations in the water-wind-solar system, W is the number of wind farms in the hydro-wind-solar system.

11. The method according to claim 9, characterized in that In step 3-2, the flexible adjustment capability of the hydropower-wind-solar hybrid system is obtained based on the ramping capability and output limit of the hydropower unit. The calculation formula is as follows: （32）； in, for t Flexible adjustment capability of cascade hydropower at any time; for t Flexible downward adjustment capability of cascade hydropower at all times; For hydropower stations i Climbing ability; For hydropower stations i exist t The effort of every moment; For hydropower stations i exist t Output limit at any time; For hydropower stations i exist t The lower output limit at any time.

12. The method according to claim 9, characterized in that In step 3-3, the objective function established is as follows: （33）； （34）； （35）； （36）； in, The flexibility increase is less than expected, specifically t The expected difference between the flexible adjustment demand caused by insufficient adjustment capacity and the flexible adjustment capacity at any given moment; To reduce the lack of flexibility, specifically t The expected difference between the flexible downward adjustment demand caused by insufficient downward adjustment capacity and the flexible downward adjustment capacity at any given moment; For the scene s Lower Hydropower Station i exist t The amount of water discarded at each moment; for t Expected amount of water abandoned by cascade hydropower at any given time; Refers to the scene s The probability of For the scene s Down t Flexible adjustment needs at all times, for t Flexible adjustment capability of cascade hydropower at any time, For the scene s Down t Flexible downward adjustment requirements at any time, for t The flexible downward adjustment capability of cascade hydropower at all times.

13. The method according to claim 9, characterized in that In step 3-4, the constraints of the water-wind-solar system are established as follows; (1) Water balance constraints: （37）； in, For the scene s Lower Hydropower Station i exist t Storage capacity at time +1; For the scene s Lower Hydropower Station i exist t Storage capacity at a given moment; For the scene s Lower Hydropower Station i exist t Inbound traffic at each moment; For the scene s Lower Hydropower Station i exist t Outbound traffic at the moment; is the scheduling time interval; For hydropower stations i -1 and hydropower station i When the water flow is stagnant; For the scene s Lower Hydropower Station i -1 at time Traffic volume; For the scene s Lower Hydropower Station i -1 and hydropower station i Between t Interval flow at the time; For the scene s Xiadi Hydropower Station i exist t Power generation flow at the moment; For the scene s Lower Hydropower Station i exist t The amount of water discarded at each moment; (2) Water level constraint: （38）； in, For the scene s Lower Hydropower Station i exist t Reservoir water level at the moment; For hydropower stations i exist t The lower limit of the water level at the moment; For hydropower stations i exist t The upper limit of the water level at the moment; (3) Initial and final water level constraints: （39）； （40）； in, is the initial water level in the scheduling period; is the final water level in the scheduling period; is the allowable deviation, Refers to the scene s Lower cascade hydropower station i At the initial water level of the scheduling period, Refers to the scene s Lower cascade hydropower station i The water level at the end of the dispatch period; (4) Reservoir capacity-water level constraints: （41）； in, For hydropower stations i Reservoir capacity-water level function; (5) Tailwater level-outflow flow constraint: （42）； in, For hydropower stations i Outflow-tailwater level function; For the scene s Lower Power Station i exist t Tailwater level at the time; (6) Unit output constraints: （43）； in, For the scene s Down t Shike Hydropower Station i The unit n of efforts, For hydropower stations i The unit n The output limit, For hydropower stations i The unit n The lower limit of output, For hydropower stations i The unit n exist t The start and stop state variables of the hydropower station i The output is expressed as: （44）； in, Indicates hydropower station i The number of units; (7) Unit power generation flow constraints: （45）； in, For the scene s Down t Shike Hydropower Station i The unit n Power generation flow; For hydropower stations i The unit n The upper limit of power generation flow; For hydropower stations i The unit n The lower limit of power generation flow, hydropower station i The power generation flow is expressed as: （46）； (8) Unit vibration zone constraints: （47）； in, For hydropower stations i The unit n Vibration zone k Output limit; For hydropower stations i The unit n Vibration zone k The lower limit of output; (9) Unit start-up and shutdown duration constraints: （48）； in, for t Shike Hydropower Station i The unit n The startup operation variable, 1 means the startup operation; for t Shike Hydropower Station i The unit n The shutdown operation variable, 1 means shutdown operation; For hydropower stations i The unit n Minimum power-on duration; For hydropower stations i The unit n Minimum downtime duration; For hydropower stations i The unit n The maximum number of startup times within the scheduling cycle; (10) Unit output ramp constraints: （49）； in, For the scene s Down t+1 Shike Hydropower Station i unit n of efforts, For hydropower stations i The unit n Climbing ability; (11) Unit generating head constraints: （50）； in, For the scene s Lower Hydropower Station i exist t-1 The reservoir water level at the moment, For the scene s Down t Shike Hydropower Station i The unit n The generating head; For the scene s Down t Shike Hydropower Station i The unit n Head loss; (12) Head loss constraint: （51）； in, For power stations i Head loss coefficient; For power stations i The head loss constant; (13) Unit dynamic characteristics relationship: （52）； in, For hydropower stations i The unit n Head-discharge-output function; (14) Section transmission constraints; （53）； in, I o For cross section o the number of hydropower stations included; W o For cross section o the number of wind farms included; G o For cross section o The number of PV plants included; For cross section o The transmission capacity, For the scene s Downstream wind farm w exist t The effort of every moment; (15) Power generation plan constraints: （54）； in, is a smaller deviation threshold, For hydropower stations i exist t The effort of every moment, For photovoltaic power stations g exist t The predicted output of the moment, For wind farms w exist t The predicted output of the moment, Indicates the power generation plan issued by the power grid. G is the number of photovoltaic power stations in the water-wind-solar system, I is the number of cascade hydropower stations in the water-wind-solar system, W is the number of wind farms in the water-wind-solar coefficient.

14. An electronic device, characterized in that: The invention comprises a memory (1) and a processor (2), wherein the processor (2) and the memory (1) are connected via a communication bus (3), and the memory (1) is used to store a computer program. When the computer program is executed by the processor (2), the method for establishing a day-ahead scheduling model for water, wind and solar power as described in claim 1 is implemented.

Images