No self-introduction
Loading
Profile
Follow
codes (2)
Standard NSGA-II algorithm based on real and binar
4.0
The characteristics of NSGA are as follows: & nbsp; L & nbsp; & nbsp; & nbsp; non dominated sorting: & nbsp; & nbsp; & nbsp; & nbsp; in the process of evolution, the current parent population is crossed and mutated to obtain a sub population, and the two populations are merged. In the objective space, individuals in the group are compared according to their objective function vectors according to the Pareto optimal relationship, and all individuals in the group are divided into multiple front layers which are controlled in turn. L & nbsp; & nbsp; & nbsp; fitness sharing: & nbsp; & nbsp; & nbsp; in the process of evolution, some strategies must be adopted to maintain the diversity of the population and prevent the population from eventually converging to a few individual solutions (i.e. premature convergence); & nbsp; & nbsp; & nbsp; NSGA method specifies the same fitness for individuals in the same layer, so as to ensure the distribution diversity of the population. For individuals close to a certain distance, the method of discount fitness is used to improve the coverage of frontier; Because individuals at the forefront have the greatest fitness, the more likely they are to be passed on to the next generation. The improvement of NSGA-II compared with NSGA method: l in NSGA-II, except for non dominated sorting, the operation rules are completely different from NSGA. In NSGA-II, the concept of archive is introduced. Compared with NSGA, Pareto frontier has become more reliable in its advance and expansion. Because the parent exploration population is generated from archive according to elimination selection, it exerts large selection pressure on individuals with high Pareto superiority (described in detail later). "& nbsp; & nbsp; & nbsp; this feature is manifested in the high forward ability of Pareto frontier. As an alternative method of fitness sharing in NSGA, "crowding distance" and“
pch20087333 0 0
1
Standard NSGA-II algorithm based on real
no vote
NSGA_2 whole, based on the example of real numbers, with the goal to facilitate the optimization and multi-objective optimization can be. Engineering design and decision-making often encounter problems under multiple criteria or design goals, and these goals are often Is contrary to find the best design to meet these goals, it is necessary to solve the multi-objective and multi-constraint optimization Problem, propagation algorithm provides a system for solving complex optimization problems common framework, which does not depend on the problem Field and type "on the need for a practical application of optimization, generally can follow the steps above To construct a genetic algorithm to solve the problem "can be seen from the above steps, you need to consider when constructing genetic algorithm The two main issues are the encoding method and the design of genetic operators feasible solution, which is designed to GA Two key steps. "...
pch20087333 1 0
1
No more~
Follows0
Fans0