CN-122026378-A - Water-light complementary scheduling method and related device based on improved multi-target genetic algorithm

CN122026378ACN 122026378 ACN122026378 ACN 122026378ACN-122026378-A

Abstract

The invention provides a water-light complementary scheduling method based on an improved multi-target genetic algorithm and a related device, and belongs to the technical field of water-light complementary scheduling. The method comprises the steps of carrying out normalization processing on an original multi-target fitness function to obtain a multi-target fitness function after normalization processing, constructing a multi-target fitness function for water-light complementary scheduling based on the multi-target fitness function after normalization processing, adopting a uniform random strategy initialization population mode, a self-adaptive crossover operation and a self-adaptive mutation operation to generate a new population mode and an elite retaining mechanism to improve a multi-target genetic algorithm to obtain an improved multi-target genetic algorithm, adopting the improved multi-target genetic algorithm to solve the established multi-target fitness function for water-light complementary scheduling to obtain an optimal value of the fitness function, and carrying out water-light complementary scheduling according to the optimal value of the fitness function. The invention solves the problem of low precision of complementary scheduling of water and light.

Inventors

Yan Ningjun
TIAN JIAWEI
CHEN BIN
WU BAOFU
CUI MENG
ZHOU ZHITONG
HUANG XINXIN
FENG JIANTONG

Assignees

大唐水电科学技术研究院有限公司
中国大唐集团科学技术研究总院有限公司
中国大唐集团科学技术研究总院有限公司水电科学研究院

Dates

Publication Date: 20260512
Application Date: 20260119

Claims (10)

1. The water-light complementary scheduling method based on the improved multi-target genetic algorithm is characterized by comprising the following steps of: acquiring an average power generation water head of each time period of the cascade hydropower station, a unit output efficiency coefficient of the cascade hydropower station, power generation flow of each time period of the cascade hydropower station and total power of hydropower and photovoltaics of each time period of the cascade hydropower station; Constructing an original multi-target fitness function based on an average power generation water head of each time period of the cascade hydropower station, a unit output efficiency coefficient of the cascade hydropower station, power generation flow of each time period of the cascade hydropower station and total power of hydropower and photovoltaics of each time period of the cascade hydropower station; normalizing the original multi-target fitness function to obtain a normalized multi-target fitness function; Constructing a water-light complementary scheduling multi-target comprehensive fitness function based on the multi-target fitness function after normalization processing; initializing a population mode by adopting a uniform random strategy, generating a new population mode by adopting self-adaptive cross operation and self-adaptive mutation operation, and improving a multi-target genetic algorithm by adopting an elite retention mechanism to obtain an improved multi-target genetic algorithm; Solving the established water-light complementary scheduling multi-target comprehensive fitness function by adopting an improved multi-target genetic algorithm to obtain an optimal value of the fitness function; and carrying out water-light complementary scheduling according to the optimal value of the fitness function.
2. The improved multi-objective genetic algorithm-based hydro-optical complementary scheduling method according to claim 1, wherein the original multi-objective fitness function expression is: Wherein, the As an original multi-objective fitness function, The objective function is stored for the original steps of scheduling individual k, For the original system total output fluctuation objective function of schedule individual k, The objective function weighting coefficients are stored for the steps of the original schedule individual k, The method comprises the steps of (1) providing a system total output fluctuation objective function weight coefficient for an original scheduling individual k; step energy storage objective function of original scheduling individual k The expression is: Wherein, the For the power generation flow of the t scheduling period of the cascade hydropower station i, For the average power generation head of the cascade hydropower station i in the t scheduling period, The method comprises the steps of providing a unit output efficiency coefficient for the step hydropower station i, wherein N is the number of the step hydropower stations, and T is the total scheduling period number; original system total output fluctuation objective function of scheduling individual k The expression is: Wherein, the To schedule individual k for the total power of both hydropower and photovoltaic at time t schedule period, Total power of hydropower and photovoltaics at the time of scheduling period t-1 for scheduling individual k; Scheduling individual k total power of hydropower and photovoltaics at t scheduling period The calculation formula of (2) is as follows: Wherein, the To schedule individual k The schedule output of the seatpost hydropower station in the t schedule period, A predicted sequence of output for the photovoltaic at the t-th schedule period.
3. The improved multi-objective genetic algorithm-based hydro-optical complementary scheduling method according to claim 2, wherein the scheduling of individual k is the first The upper and lower limit constraints met by the dispatching output of the seat step hydropower station in the t dispatching period are as follows: Wherein, the And The upper and lower limits of the output of the cascade hydropower station i are set.
4. The improved multi-objective genetic algorithm-based hydro-optical complementary scheduling method according to claim 2, wherein the scheduling of individual k is the first The vibration area avoidance constraint met by the dispatching output of the seat step hydropower station in the t dispatching period is as follows: Wherein, the The upper limit and the lower limit of the vibration section of the cascade hydropower station i are set.
5. The improved multi-objective genetic algorithm-based water-light complementary scheduling method according to claim 1, wherein the expression of the water-light complementary scheduling multi-objective comprehensive fitness function is: Wherein, the The multi-objective comprehensive fitness function is scheduled for the water-light complementation, Step energy storage objective function weight coefficients for the scheduled individual k after normalization processing, For the system total output fluctuation objective function weighting coefficient of the schedule individual k after the normalization process, For the step-wise stored objective function of the scheduled individual k after the normalization process, A system total output fluctuation objective function of the scheduling individual k after normalization processing; step energy storage objective function of scheduling individual k after normalization processing The expression is: Wherein, the The objective function is stored for the original steps of scheduling individual k, And Respectively obtaining a maximum value and a minimum value of a step energy storage objective function in the current population; system total output fluctuation objective function of scheduling individual k after normalization processing The expression is: Wherein, the For the original system total output fluctuation objective function of schedule individual k, And The maximum value and the minimum value of the total output fluctuation objective function of the system in the current population are respectively.
6. The improved multi-objective genetic algorithm-based water-light complementary scheduling method according to claim 1, wherein the calculation formula of the adaptive crossover operation is: Wherein, the The cross probability is complementarily scheduled for scheduling the water light of individual k, The multi-objective comprehensive fitness value is complementarily scheduled for scheduling the water light of the individual k, And Respectively carrying out complementary scheduling on the multi-target comprehensive fitness values of the population minimum value and the average water light, And The upper limit and the lower limit of the cross probability of the water-light complementary scheduling are respectively adopted, Is a smoothing factor; the calculation formula of the adaptive mutation operation is as follows: Wherein, the To schedule individual k's water-light complementary schedule variation probabilities, The multi-objective comprehensive fitness value is complementarily scheduled for scheduling the water light of the individual k, And Respectively carrying out complementary scheduling on the multi-target comprehensive fitness values of the population minimum value and the average water light, And The upper limit and the lower limit of the cross probability of the water-light complementary scheduling are respectively adopted, Is a smoothing factor.
7. The improved multi-objective genetic algorithm-based hydro-optical complementary scheduling method according to claim 1, wherein the elite retention mechanism has a calculation formula: Wherein, the The amount is reserved for the elite and, For the elite proportion of the population, M is the population size.
8. An improved multi-objective genetic algorithm-based hydro-optical complementary scheduling system, comprising: The data acquisition module is used for acquiring the average power generation water head of each period of the cascade hydropower station, the unit output efficiency coefficient of the cascade hydropower station, the power generation flow of each period of the cascade hydropower station and the total output of hydropower and photovoltaics of each period of the cascade hydropower station; The original fitness function construction module is used for constructing an original multi-target fitness function based on the average power generation head of each period of the cascade hydropower station, the unit output efficiency coefficient of the cascade hydropower station, the power generation flow of each period of the cascade hydropower station and the total output force of hydropower and photovoltaics of each period of the cascade hydropower station; The normalization processing module is used for carrying out normalization processing on the original multi-target fitness function to obtain a multi-target fitness function after normalization processing; the comprehensive fitness function construction module is used for constructing a water-light complementary scheduling multi-target comprehensive fitness function based on the normalized multi-target fitness function; The multi-target genetic algorithm improvement module is used for adopting a uniform random strategy to initialize a population mode, a self-adaptive cross operation and a self-adaptive mutation operation to generate a new population mode and an elite retention mechanism to improve the multi-target genetic algorithm so as to obtain an improved multi-target genetic algorithm; the fitness function solving module is used for solving the established fitness function of the water-light complementary scheduling multi-objective comprehensive by adopting an improved multi-objective genetic algorithm to obtain the optimal value of the fitness function; and the scheduling module is used for carrying out water-light complementary scheduling according to the optimal value of the fitness function.
9. An electronic device comprising a memory and a processor, the memory storing a computer program, characterized in that the processor, when executing the computer program, implements the steps of the improved multi-objective genetic algorithm based hydro-optical complementary scheduling method according to any one of claims 1 to 7.
10. A storage medium having stored thereon a computer program, which when executed by a processor, implements the steps of the improved multi-objective genetic algorithm-based hydro-optical complementary scheduling method according to any one of claims 1 to 7.

Description

Water-light complementary scheduling method and related device based on improved multi-target genetic algorithm Technical Field The invention belongs to the technical field of water-light complementary scheduling, and particularly relates to a water-light complementary scheduling method based on an improved multi-target genetic algorithm and a related device. Background In a novel power system mainly comprising new energy, the new energy gradually takes the dominant role in a power supply structure. However, the photovoltaic output has volatility, randomness and intermittence, so that the impact on the safe and stable operation of the power grid is easy to cause, especially under the condition of large-scale grid connection. The hydropower is used as a carbon-free clean energy source with excellent regulation performance, is used as an electric power type power source and is also used as an electric power type power source, and the flexible and rapid peak regulation capability can fully make up the defects of randomness, fluctuation and intermittence of new energy. The scheduling system needs to coordinate the flexibility of hydroelectric power generation and the fluctuation of photovoltaic power generation, so that economic, efficient, safe and stable joint scheduling is realized. However, the existing scheduling method still has significant limitations, and the current method for complementary scheduling of water and light mainly has the following problems: (1) Most of the traditional scheduling methods are mainly optimized by a single target, and the trade-off between multidimensional targets such as economy, reliability and the like cannot be considered at the same time, so that the accuracy of complementary scheduling of water and light is low. (2) When facing the problem of high-dimensional complex constraint, the traditional heuristic algorithm is often in local optimum, and the algorithm convergence speed is low, so that the real-time requirement of actual scheduling is difficult to meet. Disclosure of Invention The invention aims to provide a water-light complementary scheduling method based on an improved multi-target genetic algorithm and a related device, which are used for solving the problem of low precision of water-light complementary scheduling in the prior art. In order to achieve the above purpose, the present invention adopts the following technical scheme: in a first aspect, the present invention provides a water-light complementary scheduling method based on an improved multi-objective genetic algorithm, comprising the steps of: acquiring an average power generation water head of each time period of the cascade hydropower station, a unit output efficiency coefficient of the cascade hydropower station, power generation flow of each time period of the cascade hydropower station and total power of hydropower and photovoltaics of each time period of the cascade hydropower station; Constructing an original multi-target fitness function based on an average power generation water head of each time period of the cascade hydropower station, a unit output efficiency coefficient of the cascade hydropower station, power generation flow of each time period of the cascade hydropower station and total power of hydropower and photovoltaics of each time period of the cascade hydropower station; normalizing the original multi-target fitness function to obtain a normalized multi-target fitness function; Constructing a water-light complementary scheduling multi-target comprehensive fitness function based on the multi-target fitness function after normalization processing; initializing a population mode by adopting a uniform random strategy, generating a new population mode by adopting self-adaptive cross operation and self-adaptive mutation operation, and improving a multi-target genetic algorithm by adopting an elite retention mechanism to obtain an improved multi-target genetic algorithm; Solving the established water-light complementary scheduling multi-target comprehensive fitness function by adopting an improved multi-target genetic algorithm to obtain an optimal value of the fitness function; and carrying out water-light complementary scheduling according to the optimal value of the fitness function. The invention is further improved in that the original multi-objective fitness function expression is: Wherein, the As an original multi-objective fitness function,The objective function is stored for the original steps of scheduling individual k,For the original system total output fluctuation objective function of schedule individual k,The objective function weighting coefficients are stored for the steps of the original schedule individual k,The method comprises the steps of (1) providing a system total output fluctuation objective function weight coefficient for an original scheduling individual k; step energy storage objective function of original scheduling individual k The expression is: Wherein