Multi-target dispatching method considering wind electricity output for electrical power system

Abstract

The invention discloses a multi-target dispatching method considering wind electricity output for an electrical power system. The method comprises the steps of outputting high and low limits of a prediction interval based on high-low limit evaluation method of a neural network, generating a wind electricity frequency scene in a prediction interval sequence to describe the randomness of wind electricity power, simultaneously considering the total fuel cost, the pollutant discharge amount, the risk cost and the rotary standby cost in an objective function by taking load shedding punishment expense expectation as a risk index, constructing an optimized dispatching model including a wind electricity power system, and carrying out optimization solution on the constructed multi-target objective function based on a flower pollination algorithm and a differential evolution algorithm so as to obtain the optimized dispatching method. According to the method, the load shedding punishment expense expectation is taken as the risk index, meanwhile, the rotary standby cost is considered in the objective function, the optimized dispatching model including the wind electricity power system is constructed, conventional operation constraints of a unit are considered by the constructed model, and besides, reliability constraints are further added, so that the environment, the economy and the safety are simultaneously considered by virtue of a dispatching decision.

Description

technical field [0001] The invention relates to a multi-objective scheduling method of a power system considering wind power output. Background technique [0002] The economic scheduling problem of the power system refers to the scheme that minimizes the coal consumption of thermal power units under the conditions of satisfying the balance of system energy supply and demand and the constraints of unit output. The scheduling problem after wind power grid connection must consider the uncertainty of wind power, because the wind speed is random. When the wind power penetration rate reaches a certain proportion, it will inevitably increase the risk to the operation of the power system. The traditional wind power scene generation is generally divided into two steps: directly assume that the wind power obeys a certain distribution, and randomly simulate the sampling to generate the scene; However, the assumed distribution does not conform to the objective and actual distribution of...

AI Technical Summary

Problems solved by technology

[0003] The multi-objective economic scheduling problem including wind farms considers the safety issues brought about by wind power grid connection, the environmental pollution caused by coal-fired power units, and the cost of spinning reserves. This is a high-dimensional nonlinear multi-objective problem. When dealing with multi-objective optimization problems, the The multi-objective problem is transformed into a single-objective problem by weight addition or penalty function method, but there is no basis for selection of weight and penalty factor values, so it is difficult to obtain the optimal solution

Images