论文部分内容阅读
针对遗传算法(GA)在MF-TDMA资源规划中易出现陷入局部最优和搜索效率低的问题提出一种改进的遗传算法。通过改进精英保留策略以及交叉和变异概率对的选择方法降低了陷入局部最优的概率,并且基于模式定理和积木块假设提出的分块搜索提高了搜索效率。仿真结果表明改进遗传算法结合自适应罚函数和适应度函数在MF-TDMA资源规划中相对于简单遗传算法、自适应遗传算法和分层遗传算法全局搜索能力更强和搜索效率更高。
Aiming at the problem that Genetic Algorithm (GA) tends to fall into local optimum and search inefficiency in MF-TDMA resource planning, an improved genetic algorithm is proposed. By improving the elitist retention strategy and the crossover and mutation probability pair selection methods, the probability of falling into the local optimum is reduced, and the search efficiency is improved by the block search proposed based on the model theorem and the block hypothesis. The simulation results show that the improved genetic algorithm combined with adaptive penalty function and fitness function in MF-TDMA resource planning relative to the simple genetic algorithm, adaptive genetic algorithm and hierarchical genetic algorithm more powerful global search and search efficiency.