Research Of Mind Evolutionary Computation Multi-modal Optimization Performance And Of Mind Evolutionary Computation Parameters Effecting Efficiency

Evolutionary computation has four main branches: Genetic Algorithm, Evolutionary Strategy, Evolutionary Programming and Genetic Programming.They are stochastic optima algorithms that simulate biology'evolutionary process and mechanism.Becouse of its particular theory and ability to solve some problems, evolutionary computation has been paid many attentions and studies by many researchers in the world. In the 1960s, genetic algorithm (GA) was proposed by professor J. Holland from Michigan University. In the passed 30 years, the theoretical study and application of GA are most active, and has rapidly become a highlight in the fields of international academy. However, there are many shortcomings in GA, which should not be neglected. In the early stage, researchers have noticed the problem of GA's prematurity. Another crucial problem is computing efficiency of GA; due to non-directionality of its evolutionary mechanism.To overcome the problems of GA, Mind Evolutionary Computation (MEC) was proposed by Chengyi Sun in 1998, which imitates two phenomena of human society - similartaxis and dissimilation. After several years' studies on its theory andexperiment, MEC has been made great progress. So far, a preliminary system has already been established for MEC.In this thesis, based on the review of EC and MEC, two-level MEC based on two objects optimization is constructed, multimodal optimization performance of MEC is researched by using two-level MEC deeply, and so MEC's framework is enriched. The effect that relation among MEC's parameters to MEC's searching efficiency is discussed, and so we can use MEC to solve problems better in the future.Firstly, in order to improve the experiment efficiency and conquer the discommodiousness of adjusting parameters by hands , two-level MEC based on two objects optimization is constructed. In high-level MEC, a group of parameters is an individual and is evaluated by low-level MEC, meanwhile its fitness is returned. Thought about the number of evaluation is the important standard that evaluates the algorithm's performance, on the other hand the solutions' quality (that is solutions' precision) is thought, the high-level MEC's fitness is the result that the low-level MEC's function's number of evaluation adds the absolute error multiplied a power. The absolute error is between the found solution and real solution. Then we can get the trade-offs of the number of evaluations and the solution' precision in low-level MEC. The low-level MEC optimizes function.Secondly, multimodal optimization performance of MEC is researched deeply using two-level MEC . Ten multimodal functions are tested, including equal or unequal peak values. The performance of MEC is compared with several NGAs and SCGA, and the result shows MEC's multimodal optimization performanceis very high. Especially to the complicated deceptive function, the MEC's efficiency is higher 92% than the compared algorithm.Finally, in order to use MEC to solve problems better and solve one dimension and two dimensions numeric problems , we use two-level MEC to discuss the parameters 1) Nm 's effect to the found solution's precision . The result shows that when the S(i is fixed about numeric problems ,the bigger the Arm is, the higher the found solution's precision is. 2) 5G's effect to MEC's searching efficiency too . On the condition of the found solutions getting the precision , the scale of SG is got when MEC searching computing cost is less .