The QPSO (Quantum-behaved Particle Swarm Optimization) algorithm is a novel swarm intelligent optimization algorithm. As it has powerful searching functions in an overall situation, fewer control parameters, faster calculating speed and other prominent characteristics; it has more advantages than those algorithms which are based on the intelligent optimization algorithm for searching for swarms and has good application prospects. However, the QPSO algorithm has many problems, such as poor partial searching ability, low searching accuracy and it is easy to get into the prime part during the solving process. So far, the research centering on the algorithm of quantum particle swarm has been very fruitful, and a series of the improved algorithms have, to different extent, improved the performance of the algorithm. In order to improve the performances of the algorithm further, this paper has done a comprehensive and systematic research about the Quantum-behaved Particle Swarm. Aiming at the problems of QPSO algorithm, it has improved the algorithm and also applied the improved algorithm to solving practical problems so that it could verify the validity of the algorithm.The main contents this paper studies are as follows:1. This paper has posed an adaptive local search quantum particle swarm optimization algorithm. During the process of the QPSO algorithm, especially in the initial stage of the search, the solutions that algorithm could find are not necessarily the best ones, at this time, the optimal strategy needs to search in larger areas. But in the late search stage, it has already included or approached the optional value. Thus, we should search in the adjacent areas, which will improve the search accuracy and effectiveness. After a great deal of analysis and experiment, we have determined adaptively to adjust the size of local search space strategy.2. This paper has put forward a breed group of adaptive local search quantum particle swarm optimization algorithm. Through the analysis and study of groups' randomness, diversity, overall situation and part capabilities, the QPSO has good convergence in an overall situation. But it also needs an effective local search mechanism so that it can search the optimal solution better. Based on this idea, this paper has raised a strategy combining classification method and adaptive local search to realize the improvement of QPSO algorithm's convergence speed and function. This strategy utilizes the QPSO algorithm's good convergence function, classify it according to the current search state of particles, and then make an adaptive local search to the classified particle swarm in order to achieve enhancing local search ability and search speed.3. This paper will apply the improved QPSO algorithm to transportation problems. The simulative experiment has showed its validity... |