摘要
在对微粒群优化算法PSO分析的基础上,提出了矢量微粒群优化算法VPSO。该算法通过矢量运算方法来定义微粒的运动,从而达到寻找最优解的目的。将VPSO和PSO分别用于常用测试函数的优化求解,结果表明:VPSO的优化性能明显优于PSO。基于VPSO构造的矢量微粒群神经网络(VPSONN)在丙烯腈收率软测量建模的应用中表明:基于VPSONN的丙烯腈收率软测量模型具有较高的精度,应用前景广阔。
Based on analyzing algorithm of particle swarm optimization （PSO）, the algorithm of vector particle swarm optimization （VPSO） is proposed. In VPSO, the definition of particle motion was given from the vector operation, so the optimal solution was found. The optimizations to common test functions using VPSO and PSO were compared. The results show that the optimization performance using VPSO is much better than that using PSO. The vector particle swarm optimization neural network （VPSONN） based on VPSO is applied in soft sensing modeling of acrylonitrile yield. The practice shows that the VPSONN based soft sensing model of acrylonitrile yield possesses higher precision. It is widely prosperous in application.
出处
《自动化仪表》
CAS
2006年第5期13-16,20,共5页
Process Automation Instrumentation
关键词
微粒群优化算法
丙烯腈收率
软测量
Particle swarm optimization Aerylonitrile yield Soft sensing
