基于生物激勵(lì)機(jī)制的算法研究.pdf

1、湘潭大學(xué)碩士學(xué)位論文基于生物激勵(lì)機(jī)制的算法研究姓名：彭琰申請(qǐng)學(xué)位級(jí)別：碩士專(zhuān)業(yè)：控制理論與控制工程指導(dǎo)教師：鄭金華20070501II ABSTRACT The nature is our source of solving various problems. For hundreds of years, it is proved that it is a successful way that use solutions which

2、provided by biology to solve practical problems. Now biology simulation has been a part of computer science. Many optimization algorithms based on biology prompting appear. For example, Genetic Algorithm, Artificial Immu

3、ne Systems, Ant Colony Optimization, Particle Swarm Optimization, Artificial Neural Network, Culture Algorithm, etc. These algorithms learning from life phenomenon mimic the behavior, function and characteristic of life

4、system. Now they have been widely applied in many other areas and multi-objective optimization problems. Population-based genetic algorithm is a kind of random searching method using evolutionary theory and genetic theor

5、y. Particle Swarm Optimization Algorithm is such a new optimization method which is inspired by social behavior of bird flocking or fish schooling. It is a population-based, self-adaptive search optimization technique.

6、This paper gives the researchs on the above two algorithms in single-objective, and applys genetic algorithm to multi-objective optimizations. We improve the A Restricted Genetic Algorithm based on Ascending of Tangent P

7、lanes. Multi-parent fitness-weighted crossover is adopted to improve the algorithm’s ability of global search. The experiment indicates that the improved algorithm can solve more complicated multi-variant and multi-modal

8、 problems. Particle Swarm Optimization is combined with genetic operator (crossover and mutation). So it could escape from the local optima. Compared with standard Particle Swarm Optimization in four typical test functio

9、ns, results show that our algorithm has potential to find a better solution.In multi-objective optimization, we describe A Fast Genetic Multi-objective Algorithm based on Random Operator. Arena’s Principle (AP) is used t

10、o construct the nondominated set so quicken the running efficiency. Random Operator is used to construct next population. Experiment results in different dimensions indicate that it has a better distributing than NSGA2,