摘要
针对发酵过程中一些难以或者无法在线测量的关键物化参数和生物参数等变量,提出了改进的PSO-FNN软测量建模方案。通过改进的粒子群优化算法(PSO)寻优算法与模糊神经网络(FNN)相结合,建立发酵过程的软测量模型,再结合实际数据进行仿真研究。仿真结果表明,与传统PSO-FNN软测量相比,改进的模型测量精度更高,可以满足实际工程中的要求。
In fermentation process,some of the physical,chemical,and biological parameters are difficult or impossible to be measured online,to solve the problem,the soft sensing modeling strategy based on improved PSO-FNN is proposed. Through combining improved particle swarm optimization( PSO) algorithm with fuzzy neural network( FNN),the soft sensing model of fermentation process is established; then the simulation of model is researched with actual data. The results of simulation show that comparing with traditional PSO-FNN soft sensing,the improved model possesses higher measurement accuracy,and meets requirements of practical projects.
出处
《自动化仪表》
CAS
2016年第3期62-64,共3页
Process Automation Instrumentation
基金
国家高技术研究发展计划基金资助项目(编号:2011AA09070301)
江苏省农业科技支撑基金资助项目(编号:BE2010354)
江苏省自然科学基金资助项目(编号:BK2011465
SBK2014042351)
江苏省高校自然基金资助项目(编号:12KJB210001)
江苏大学高级人才启动基金资助项目(编号:12JDG108)