摘要
从新增注入元的数量、因子道路树平均路径长度以及各点集快速前代乘加次数之和三个方面,比较研究了几种不同的节点编号算法。提出了粒子群优化智能算法,通过为不同的适应值设置相应的价值函数,分别用来衡量各种启发式算法的有效性。同时,针对启发式算法没有使确定的稀疏矢量非零元的道路树平均路径长度最短这一不足之处,给出了解决这一局限性的粒子群优化算法。研究结果为解决针对电力系统不同问题,提供了合理选取节点编号算法的依据。
Several node ordering algorithms were compared in three aspects as follows： the number of new fill-ins, the average length of overall nodes path and the times of fast forward for total nodes. The algorithm of particle swarm optimization was provided, which can be used to test the effectiveness of heuristic algorithms by means of set different functions for corresponding fitness value. Meanwhile, because of the limitation of heuristic algorithms which can＇t guarantee the average minimum path length of definite non-zero elements in the sparse vector, the PSO algorithm was provided to overcome the weakness. Advice was provided to choose suitable node ordering algorithm for different power system issues.
出处
《能源工程》
2014年第2期23-28,共6页
Energy Engineering
关键词
节点编号
新增元
道路集
粒子群优
适应值
node ordering algorithm
fill-in
the rode set
particle swarm optimization
fitness value
