Research And Application Of Interval Optimization Algorithm

Interval optimization algorithm using interval variables instead of point variables is a deterministic global method. Using interval optimization algorithm to optimize and control industrial process can prevent noise and disturbance from influencing system and provide a variety of feasible scheme for system. However, traditional interval optimization algorithm has many shortcomings which limit application scope of the algorithm. If search space is quite large, it is hard to solve the high-dimensional multimodal optimization problems.To overcome the shortcomings of traditional interval optimization algorithm, an interval-particle swarm optimization algorithm was proposed. Traditional interval optimization algorithm efficiency is greatly improved by random search capability and information sharing mechanism of Particle Swarm Optimization. High-dimensional multimodal optimization problem can be solved by using Particle Swarm Optimization as interval accelerated tools. This thesis includes the following aspects mainly:Basic concepts and idea of traditional interval optimization algorithm were introduced in detail. And basic process of the algorithm in solving specific examples was illustrated.According to basic idea and steps of interval optimization algorithm, an interval-particle swarm optimization algorithm was prodived.10diversiform functions were used to determine its reliability.The interval-particle swarm optimization algorithm was applied to optimize glutamic acid fermentation process. Neural network was used to found an interval model. Optimal interval of acid yield rate in every stage was got as well as the optimal interval of every variable.At the end of this thesis, research was summarized, and improving direction of interval algorithm and application in control theory was given.