摘要
针对污水处理过程中生化需氧量BOD难以实时在线测量的问题,建立了用于预估BOD的支持向量机(SVM)的软测量模型。考虑到该支持向量机模型的测量精度取决于其两个参数C、能否获得最优值,采用遗传算法和粒子群优化算法,实现对这两个参数的寻优。仿真结果表明:该软测量模型的测量精度较高,可用于污水处理厂对BOD进行在线测量。
In view of the hardship to get real-time and on-line of Biochemical Oxygen Demand （BOD） in sewage disposal process. A soft-sensing model based on support vector machine is established for estimating the Biochemical Oxygen Demand. The measurement accuracy of the support vector machine model depend on a good setting of the two parametres C, σ. Genetic algorithm and Particle Swarm Optimization （PSO） algorithm am applied to optimize the two parameters synchronously.The simulation results indicate that the soft-sensing model can be used for sewage disposal plant on-line measurement of BOD.
出处
《自动化与仪器仪表》
2009年第6期6-9,共4页
Automation & Instrumentation
关键词
软测量
支持向量机
污水处理
遗传算法
粒子群算法
Soft-sensing
Support vector machine
Sewage disposal
Genetic algotithm
Particle swarm optimization
