| In recent years, thermal power units didn't operate under the full load conditionor its vicinity along with the power grid peak-valley increased gradually in our country. The method which could meet the requirements of reducing costs for power plant and the demand of real time dispatching for power grid is critical to use optimal load dispatch of power units widely.Therefore, firstly, the online monitoring method of main steam flow rate of a steam turbine based on the main condensate flow rate was proposed to modify the coal consumption characteristic curve. And its accuracy and practicability were confirmed by an application example. Secondly, according to the demand of real time dispatching for power gird, the total time of optimal load dispatch was proposed. The calculation formulas were given out to establish mathematical model, to choose optimal algorithm and to provide the basis for finishing load command of power grid dispatching quickly. Furthermore, multi-objective mathematical model based on the coal consumption and the pollutant emission and the load adjustment time wasestablished. And it was improvingly treated as single-objective mathematical model. Thirdly, the particle swarm optimization algorithm was discussed and improved. And the specific issues that algorithm applying in optimal load dispatch were processed.According to the contradiction between optimization computing and solution quality of intelligent algorithm, the real time load optimal dispatch method based on the improved BP network was proposed. BP network was trained by the sample data thatmultiple mutation adaptive genetic algorithm was introduced for calculation, then it can dispatch load in real time. Finally, the feasibility of the mathematical model and the effectiveness of the optimal algorithm are further verified by an application example. Meanwhile, it can meet the needs of reducing costs for power plant and thedemand of real time dispatching for power gird.
|
PDF Full Text Request