Adaptive Function of Genetic Algorithm Optimization and Application
Performance of genetic algorithms is dramatically influenced by algorithmic settings. To improve the research performance of genetic algorithm and avoid its limitation of local optimization, a new adaptive genetic algorithm is applied to optimize three st
- PDF / 1,870,685 Bytes
- 7 Pages / 439.37 x 666.142 pts Page_size
- 99 Downloads / 207 Views
Adaptive Function of Genetic Algorithm Optimization and Application Jiang-Bo Huang
Abstract Performance of genetic algorithms is dramatically influenced by algorithmic settings. To improve the research performance of genetic algorithm and avoid its limitation of local optimization, a new adaptive genetic algorithm is applied to optimize three standard benchmark functions selected in this paper. The comparison between the results of the present algorithm and that of the simple genetic algorithm shows that the technique has improved the performance of genetic algorithm. Keywords Genetic algorithm
Adaptation Optimization Stereo matching
96.1 Introduction Genetic algorithms, genetic operations on the convergence of the algorithm performance have a large impact. The algorithm crossover operation and mutation operation combined to form the new model, so as to continuously open up new exploration of space, easy to jump out of local optimal solution, and get the global solution. Crossover probability, the greater the algorithm in the more frequent cross operation in the optimization process, the faster the update of the gene string groups that can enhance the genetic algorithm to open up the ability of the new search area, but high-performance model was the possibility of damage which also increased. If the crossover probability is set too low, the algorithm may be too small, where new space exploration rate stagnated. Variation in the genetic
J.-B. Huang (&) Yangtze Nornal University School of Physics or Electron Engineering, Chongqing, China e-mail: [email protected]
Z. Zhong (ed.), Proceedings of the International Conference on Information Engineering and Applications (IEA) 2012, Lecture Notes in Electrical Engineering 218, DOI: 10.1007/978-1-4471-4847-0_96, Ó Springer-Verlag London 2013
781
782
J.-B. Huang
algorithm assisted search operation, its main purpose is to maintain the diversity of the solutions group. In general, low-frequency variability prevents groups, a small number of gene loss, high-frequency variability in turn makes the algorithm tends to a pure random search [1–3]. Variation in the genetic algorithm assisted search operation, its main purpose is to maintain the diversity of the solutions group. In general, low-frequency variability prevents groups, a small number of gene loss, and high-frequency variability in turn makes the algorithm tends to a pure random search. The genetic algorithm, search performance algorithm, and depth search algorithm in the search space, and breadth of the search for balanced decision, this balance by the algorithm parameter settings of algorithm parameters has been an important branch of this field. The traditional genetic algorithm parameters are generally taken as a fixed value, to enable the algorithm is relatively simple, but often cannot get a satisfactory search results. Genetic algorithm taking into account the essence is dynamic and adaptable, many scholars have proposed a variety of adaptive mechanisms, including the adaptive operator and parameter
Data Loading...
