Research On Plant-level Load Scheduling Optimization Based On Immune Algorithm

With the flourishing electric power industry and the intensified contradiction of coal resource, how to improve the operational efficiency and to reduce the production cost becomes the foucs which the power plants pay close attention to. Therefore, it is very important to take load optimization dispatch among the units to improve the operational efficiency and reduce the production cost. And the power production always leads to a lot of pollution to our environment, which causes for people concern. Meanwhile, with the development of the electrical market, the load dispatching not only reaches the load requirements of scheduling, but also satisfies the adjustment time. So it is meaningful to take the multi-objective optimization strategies to solve the load dispatching problems for power plant.This paper combines the non-specific immunity and specific immunity in immune theory with genetic algorithm, and proposes Adaptive Immune Vaccine Algorithm (AIVA). AIVA effectively overcomes the GA Defects, greatly prevents the degradation of population, and has perfect convergence stability and excellent global optimization capability. Then proves the convergence of the algorithm, verifies the testing functions and main steam temperature PID parameter optimization, which above have proved its effectiveness. Then, this paper proposes a new mathematical model for load scheduling optimization from three areas including economy, rapidity and environmental protection to consider. And AIVA is applied in load dispatch problem for single-objective optimization. Then compare the experimental results with others algorithm to verify the validation of algorithm and the new mathematical model.As solving the multi-objective load dispatch problems, the multi-objective evolutionary strategies based on the concept of Pareto optimal solution have obvious advantages compared with the traditional, such as weigh sum method. This paper designs Immune Multi-objective Algorithm (IMA), which have been verified by testing functions. Then compare the experimental results with NSGA-II, the effectiveness of Immune Multi-objective Algorithm and the new mathematical model of load dispatch has been verified.