CN104037755A - Optimization method for solving Pareto solution sets of wind-storage-thermal joint operation system in multiple time periods - Google Patents

Abstract

The invention relates to a method for solving Pareto solution sets of a wind-storage-thermal joint operation system in multiple time periods after access of large-scale wind power. The method adopts a means combining a traditional genetic algorithm and an NSGA-II algorithm in solving, and respectively solves the problem that the optimization result obtained by adoption of the traditional genetic algorithm is single and the problem that only the Pareto solution set of a single time period can be obtained through optimization by adoption of the NSGA-II algorithm. An optimization result of a wind-storage-thermal joint operation system mathematical model solved by the genetic algorithm is used as the initial value of solution by the NSGA-II algorithm so as to obtain the Pareto solution set of each moment of time through optimization. Compared with the traditional genetic algorithm and the NSGA-II algorithm, the method of the invention can effectively overcome the defects of the two algorithms in the solution process, more comprehensively search out optimization solution sets as many as possible, and provide comprehensive, clear and effective support for decision makers.

Description

An optimization method for solving the multi-period Pareto solution set of wind-storage-fire combined operation system

technical field

The invention belongs to the technical field of electric power dispatching analysis, and in particular relates to a method for solving a multi-period Pareto solution set of a wind-storage-fire combined operation system after large-scale wind power access.

Background technique

With the continuous expansion of my country's wind power scale, the impact of wind power grid integration on the power system is becoming more and more significant, and the phenomenon of wind abandonment continues to appear. Utilizing energy storage systems to participate in power grid peak regulation is an important way to improve the capacity of wind power consumption. As a large-scale energy storage device, the pumped storage power station can be used not only for power generation, but also for load consumption. It is an important measure to improve the peak-shaving pressure of the power grid and increase the capacity of wind power consumption. How to solve the optimal output of each unit in the dispatching period of the wind-storage-fired combined operation system is the premise of power system planning and dispatching. In solving multi-objective optimization problems, the traditional genetic algorithm can handle any form of objective function and constraint conditions, whether linear or nonlinear, discrete or continuous; The trade-off between the multi-objective problems is converted into multiple different single-objective optimization problems, and the optimization result is single. The NSGA-II algorithm does not need to know the relationship between the objective functions in advance, but uses its powerful global search ability to find possible optimal solutions for decision makers' reference; but its disadvantage is that it cannot deal with nonlinear constraints. Under this background, the present invention combines the advantages of these two algorithms to make up for the shortcomings of the two algorithms, and finally obtains the multi-period Pareto solution set of the wind-storage-fire joint operation system.

Contents of the invention

Aiming at the problems of traditional genetic algorithm and NSGA-II algorithm in analyzing and processing multi-objective optimization, the present invention proposes a method for solving multi-period Pareto solution sets of wind-storage-fire joint operation system, which can be processed in combination with traditional genetic algorithm Nonlinear constraints and NSGA-II algorithm can obtain the advantages of single-period Pareto solution set. Firstly, genetic algorithm is used to solve multi-objective multi-period optimization problems; then, according to the optimization result solved by genetic algorithm as the initial value of NSGA-II algorithm, respectively Optimize to get the Pareto solution set of 24 hours a day. This method cleverly combines the advantages of the two optimization algorithms, and makes up for the shortcomings of the two. problem with linear constraints.

To achieve the above object, the present invention takes the following technical solutions:

The present invention solves the technical scheme that the above-mentioned problem takes:

1. Establish wind power output uncertainty model. In order to better simulate the uncertainty of wind speed, the present invention adopts wind speed distribution to establish an uncertain wind power output model.

2. Establish the mathematical model of the wind-storage-fire combined operation system. Determine the multi-objective function and the corresponding constraints of the wind-storage-fire combined operation system.

3. Using traditional genetic algorithm to solve multi-objective optimization problems. The focus of the traditional genetic algorithm for solving multi-objective problems is to determine the adaptive degree function. The present invention analyzes the importance of each optimization target based on the linear weighting method, multiplies a set of weight coefficients according to the importance, and then adds them together as the objective function to complete Transformation of multi-objective problems to single-objective problems.

4. Use the NSGA-II algorithm to solve multi-objective optimization problems. In the present invention, the multi-objective optimization result obtained by solving the traditional genetic algorithm is used as the basis for solving the multi-objective optimization problem by the NSGA-II algorithm, and the result of the genetic algorithm is used as the initial value for the NSGA-II algorithm to solve the multi-objective problem, and then the NSGA-II algorithm is used Powerful global search capability to find possible optimal solutions and obtain Pareto solution sets for each time period for reference by decision makers.

The invention can be widely used in power system planning, operation and scheduling departments, forms a new solution method for power system operation analysis and scheduling decision-making, and provides decision makers with comprehensive, clear and effective support.

Detailed ways

The present invention comprises the following steps:

1) Determine the wind power output within the dispatch period

The wind speed has the characteristics of fluctuation and uncertainty. In order to better simulate the change of the actual wind speed, the present invention adopts the wind speed distribution to establish an uncertain wind power output model.

The probability density function of wind speed is as follows:

Among them, c and k are scale parameters and shape parameters respectively;

The simple model used to describe the relationship between wind speed and output wind energy is as follows:

in, is the rated power of the fan, is the rated wind speed of the fan, is the cut-in wind speed, is the cut-out wind speed.

2) Determine the mathematical model of the wind-storage-fire combined operation system:

2.1) Determination of the objective function

In the formula: It is the economic benefit obtained by wind-storage-thermal joint operation within a dispatch period, that is, the joint electricity sales income minus the joint power generation cost. It means that the output power fluctuation of the wind farm is the smallest. is the calculation time in the scheduling cycle, is the number of all conventional units in the system, Indicates the time sequence number, Indicates the serial number of the unit, Indicates the first thermal power unit in time effort; Indicates thermal power unit exist The running state at all times, when , it means power-on state, when When , it means stop state; for Direct grid-connected power of wind farms at all times; is the average daily wind power output; for Momentary hydroelectric power; for Pumping power of water pump at all times; for On-grid electricity price for thermal power units at all times; is the unit price of coal used; for On-grid tariff for wind turbines at all times; for On-grid electricity price of hydropower units at all times; for pumping charges; , , for thermal power units coal consumption coefficient.

2.2) Constraints

(1) System power balance constraints:

In the formula: for time load value.

(2) Capacity constraints of thermal power units:

In the formula: , for conventional units The upper and lower limits of output.

(3) Spinning reserve constraint

In the formula: for the system The spinning reserve capacity required at any time is taken as 7% of the load value at each time.

(4) Generator climbing rate constraint:

In the formula: , respectively Decrease rate and increase rate (MW/h) of active power output of the unit.

(5) Minimum start-stop time constraints for conventional units:

In the formula: are the minimum running time and minimum stop time of the unit, respectively.

(6) Reservoir energy storage constraints:

In the formula: for Reservoir energy storage at all times; , is the upper and lower limit of energy storage of the reservoir.

(7) Reservoir energy conversion balance constraints:

In the formula: and Respectively moment and The energy storage situation of the pumped storage power station reservoir at all times, is the duration of each time interval, is the pumping efficiency of the pump, for hydropower efficiency.

(8) The upper and lower limits of water pumping:

In the formula: , It is the upper and lower limits of water pumping.

(9) Power constraints of hydropower generation:

In the formula: , It is the upper and lower limits of the generating power of the hydraulic unit.

(10) Equality constraints for pumped hydro generation conditions:

The pumping and power generation conditions of the pumped storage power station cannot be carried out at the same time, that is, pumping water does not generate electricity, and power generation does not pump water. The two are mutually exclusive.

3) Genetic algorithm solution:

After the mathematical model of the wind-storage-fire joint operation system is established, the traditional genetic algorithm is used to solve the multi-objective optimization problem. The present invention uses the linear weighting method to determine the adaptive degree function of the genetic algorithm. The specific steps are as follows:

3.1) Solve the single-objective optimization model with joint dispatch benefit and wind power output power fluctuation as the objective function by using the genetic algorithm, and obtain the maximum and minimum values of each objective function.

3.2) Then normalize each single target value, and then use the weighting method to obtain the fitness function, calculate the fitness value of each generation population, and save the relevant variable value corresponding to the maximum fitness value.

3.3) According to the genetic algorithm selection, crossover, mutation and other steps to achieve population evolution, rationally configure the optimal value of pumping and power generation of the pumped-storage power station according to the optimal value of wind power grid connection, and then according to the configured wind-storage joint operation value, And the characteristic constraint of the thermal power unit optimizes the output of the thermal power unit, and obtains the optimal value of the thermal power unit at each time period.

3.4) Repeat steps 3.1)-3.3) to obtain the optimization results for each time period.

4) NSGA-II algorithm solution:

The genetic algorithm multi-objective optimization result obtained by the weighting method has a certain unity, and the possible optimal solution set cannot be obtained comprehensively. The NSGA-II algorithm just solves this problem. Therefore, the present invention utilizes the traditional genetic algorithm to obtain the optimized The results are used as the basis for the solution of the NSGA-II algorithm, so as to obtain the Pareto solution set of each time period. The specific steps are as follows:

4.1) The start-stop state of the thermal power unit at the first moment obtained by the optimization of the weighting method is used as the initial value of the NSGA-II algorithm to solve the first moment, and the constraints of each unit are set. The minimum amount of abandoned wind power is optimized as the objective function, and the Pareto solution set of each unit at the first moment is obtained.

4.2) Select a set of optimized values from the optimized Pareto solution set at the first moment as the initial value at the second moment for optimization, and so on, to obtain the Pareto solution set for 24 periods.

Claims (5)

1. A method for solving the multi-period Pareto solution set of the wind-storage-fire joint operation system, which comprises the following steps: 1) Establishing a method based on 2) Establish a multi-objective mathematical model of the wind-storage-fired combined operation system; 3) Use the genetic algorithm to solve the multi-objective optimization problem of the wind-storage-fired combined operation system: 3.1) Solve the single-objective optimization model separately, and get The maximum and minimum values of each objective function; 3.2) Normalize each single objective value, then use the weighting method to obtain the fitness function, calculate the fitness value of each generation population, and save the relevant variables corresponding to the maximum fitness value 3.3) According to the genetic algorithm selection, crossover, mutation and other steps to achieve population evolution, rationally configure the optimal value of pumping and power generation of the pumped-storage power station according to the optimal value of wind power grid connection, and then operate according to the configured wind-storage value, and the characteristic constraints of the thermal power unit to optimize the output of the thermal power unit, and obtain the optimized value of the thermal power unit at each time period; Wind speed value, so that the wind power output is opposite to the load characteristics; 5) The start-stop state of the thermal power unit at the first moment obtained by the weighting method optimization is used as the initial value of the NSGA-II algorithm to solve the first moment, and the constraints of each unit are set. The maximum benefit of wind-storage-thermal combined operation and the minimum system abandoned wind power are optimized as the objective function, and the Pareto solution set of each unit at the first moment is obtained; The initial value of the second moment is optimized, and so on, the Pareto solution set of 24 periods is obtained. 2. A method for solving the multi-period Pareto solution set of a wind-storage-fire combined operation system as claimed in claim 1, characterized in that: in the step 1), the Uncertain wind speed is randomly generated by distribution, so as to obtain uncertain wind power output value. This processing method is more in line with the randomness characteristics of actual wind speed. The probability density function of wind speed is as follows: Among them, c and k are scale parameters and shape parameters respectively; The simple model used to describe the relationship between wind speed and output wind energy is as follows: in, is the rated power of the fan, is the rated wind speed of the fan, is the cut-in wind speed, is the cut-out wind speed. 3. A method for solving the multi-period Pareto solution set of a wind-storage-fired combined operation system as claimed in claim 1, characterized in that, in the step 2), due to the pumping and power generation conditions of the pumped storage power station, it is impossible to At the same time, on how to deal with the relationship between the two, the present invention introduces the constraint condition (10), that is, to ensure that at least one of the two is zero at the same time period, on the premise of ensuring that pumping water does not generate electricity, and power generation does not pump water In this case, the wind power that cannot be accepted by the grid is used to pump water, that is, the abandoned wind power is stored by pumping water, and power generation is performed during peak load hours. 4. The method for solving the multi-period Pareto solution set of a wind-storage-fire combined operation system as claimed in claim 1, characterized in that, in the step 3), since the traditional genetic algorithm solves the multi-objective optimization problem, For any form of objective function and constraint conditions, no matter it is linear or nonlinear, discrete or continuous, it can be processed. The minimum start-stop time constraint of conventional thermal power units is a nonlinear constraint. Genetic algorithm has certain advantages in dealing with this model. advantage, but it needs to determine the trade-off relationship between multi-objectives in advance, and convert the multi-objective problem into multiple different single-objective optimization problems, which leads to the problem of the singleness of the optimization results, and it cannot be fully optimized as much as possible. The results are for decision makers to refer to, so the present invention utilizes the advantage of genetic algorithm in dealing with nonlinear constraints, and uses its optimization results as a basis. How to obtain the Pareto solution set of each time period is completed by the NSGA-II algorithm. 5. A method for solving multi-period Pareto solution sets of wind-storage-fire joint operation system as claimed in claim 1, characterized in that, in step 4), when solving multi-objectives by weighting method, it is necessary to determine the number of multi-objectives in advance The weight relationship among them, the results obtained in this way have a certain degree of unity, and the optimization results are not comprehensive enough. In order to find out as many optimization results as possible, the present invention combines the start-stop state of the thermal power unit optimized by the weighting method, with the first moment The start-stop value is used as the initial value of the NSGA-II algorithm, and the constraint conditions of the wind, storage, and engine units are set, and the Pareto solution set at the first moment is obtained by optimization, and a set of optimized values is selected as the initial value of the next moment. value, and so on, to get the Pareto solution set of 24 periods in a day.