仿生优化算法的研究与应用-模式识别与智能系统专业论文.docxVIP

11 11 Abstract Abstract Optimization is an age-old problem ，could be solved by so many classical mathematical methods. Wi伽 the development of science and technology，classical mathematical methods have been unable to meet the needs of technology.句 so evolutionary optimization algorithm appears 副 that moment. Evolutionary optimization algorithm is based on the bionics theory， which simulated the activities of individual animal to complete a series of operations. This paper describes the historical background and theoretical basis of optimization algorithm. It also introduces what the optimization and its related knowledge is. At the same time the particle swarm algorithm is applied to sensor networks and traffic lights controlling. The results areωfollow: (1) This paper explains what optimization problem and its related knowledge is，describe the historical background and research situation at home and abroad ofPSO. 。T)his paper gives the algorithm theory and implementation steps of PSO GA and AFSA.Mathematical methods 町e used to describe the optimization problem ，_and set out the test function. Linear system theory has been used in analyzing the reωons for PSO getting into local minimum. In order to jump out of local minimum to activate particles ，this paper draws on the crowding factor of AFSA and proposes an improved PSO method based on feeding forward disturbance .When count the number of p町ticles in the circle ，the current optimal value is the centre of the circle and the crowding factor is the radius. When the number is larger than one constant，it is considered to be trapped into local optimal value，so a disturbance is added to the system in advance. The simulation results have validated its effectiveness. This paper a忧emptsωanalyze the structure of the three optimization algorithms by analyzing the serial algorithm and parallel algorithm and so on. Finally the similarities and differences between the three optimization algorithms and the algorithm used i