论文部分内容阅读
提出了一种改进的针对高维优化问题的自适应多粒子模拟退火(AMSA)算法,通过多个粒子对整个高维空间进行随机分割和相对独立的局部退火.当每个局部于当前温度下达到稳态后,随着温度的降低,粒子依据自身状态和相互之间的关系自适应地减少粒子数目,以降低复杂度.该算法用于解决通用移动通信系统自动小区规划问题.仿真结果显示,对比其他用于解决高维优化问题的启发式算法,AMSA算法能在预定的时间内取得更理想的结果.
An improved adaptive multipartite Simulated Annealing (AMSA) algorithm for high-dimensional optimization problems is proposed, in which multiple particles are used to segment the entire high-dimensional space randomly and to perform local annealing separately.When each local is at the current temperature After reaching the steady state, with the decrease of the temperature, the particles reduce the number of particles adaptively according to their state and the relationship between them, so as to reduce the complexity.The algorithm is used to solve the automatic cell planning problem of the universal mobile communication system.The simulation results show , Compared with other heuristic algorithms used to solve high-dimensional optimization problems, AMSA algorithm can achieve better results in a predetermined time.