摘要
该文提出一种综合考虑规划区域的业务量,覆盖率和成本的并行遗传算法,通过理论分析和计算机模拟对该算法的性能进行研究。该算法将基站位置,天线高度和发射功率结合进行三维向量分布编码,利用并行遗传算法的孤岛模型,在每个孤岛中独立的运行具有保优能力的选择算子、具有方向搜索能力的单纯形算子、具有自适应变异概率的变异算子,同时各个孤岛之间增加了种群多样性的双向环迁移策略。最后,为防止产生局部收敛,还加入了爬坡算子。结果表明:该算法有效的解决了基站选址问题,并且具有速度快,效率高,且不易陷入局部最优解等优势。
This paper presents one kind of parallel genetic algorithm,which considered a region's traffic planning,radio coverage and maintenance costs.A theoretical analysis and computer simulation were undertaken to research the performance of this algorithm.The algorithm uses three-dimensional distributed vector code with base station location,antenna height,transmit power.It uses a island model of the parallel genetic algorithm.Between each island runs selection operator of high grade,simplex operator of direct search capabilities,mutation operator of adaptive mutation probability,and increased the diversity of population migration strategy bidirectional between the various island.Finally,it added the climbing operator to prevent local convergence.The results proved,the algorithm could solve the problem of base station location,which has rapid speed,highly efficient,and could prevent fall into local optimal solution.
作者
牟晓晔
MOU Xiao-ye(College of Computer Science and Information,GuiZhou University,Guiyang 550025,China)
出处
《电脑知识与技术（过刊）》
2010年第33期9529-9532,共4页
Computer Knowledge and Technology
基金
“贵州省优秀青年科技人才培养计划,产品智能化创新设计关键技术研究及应用”资助(黔科合人字(2007)08号)
关键词
遗传算法
并行计算
爬坡算子
基站选址
多目标优化
GA
parallel computing
climbing operator
base station location
multi-objective optimization
