On Plant Load Distribution Strategies Under The Two Rules

According to the electric power grid scheduling rules, in order to achieve the demand of unit rapid load and to snatch more load in the moment of raising load in the morning, power plant should decide what rate should be taken to raise/drop load, how to choose unit to improve rate of raise/drop load. This thesis presents two models which considers both the rate of raise/drop load and economic income, and integrates them to one model as the guiding method of operating workers to choose units.Based on the practical point of view, this thesis tries to solve the problem of optimal load distribution at plant level by improving modeling and algorithm design under the framework of existing plant-level optimal load distribution. The main results are as follows:Firstly, this thesis explains the basic principle of genetic algorithm, including encoding mechanism, selection, crossover, mutation and other basic genetic operation process, and introduces the elitism strategy and infeasible solution processing method.Secondly, according to the "Two Rules", electric power grid scheduling rules, of increasing requirements to power plant, which leads to contradiction and dispute between coal consumption load and all units in power plant snatches load. This thesis includes simulation experiment for the power plant which equips optimal load distribution on how to choose rate to raise load, and designs a single objective genetic algorithm which has NSGAII algorithm with elitist preservation strategy, and determine the optimum unit which should be chosen. Aiming at the choice of power plant in speediness and economy index, it should build up an optimization model with the goal of power plant’s economic income and adjustment time and optimize by using hierarchical and on-demanding NSGAII algorithm.Finally, this thesis gives a set of optimal load distribution software by using Visual C#and describes the software communication, computing, security structure and other parts in detail. And furthermore, we explain the implementation mechanism of the third chapter and the fourth chapter method in the program.