摘要
针对发酵过程的时变性和强非线性,本研究采用神经网络对发酵过程建模,在线预测某些重要生物参数,并以这些生物参数的预测值为指导,使用遗传算法(GA)在线优化主要的环境参数,确定各参数的最优轨线来控制发酵过程,并构建了由PC机和单片机组成的发酵优化控制的监控系统。实验表明,该控制系统应用于实际发酵生产过程能有效提高发酵生产效率。
Aiming at the time-vary and nonlinear features of microbe fermentation process, a neural network model is established for modeling fermentation process, predicting some important biologic parameters on-line. Using the predicted biologic parameters, several key physical parameters by which the whole fermentation process is under control can be optimized with the Genetic Algorithm. One experiment of fermentation was carried out on Penicillin. Emphasis was given to the training and update of the neural network model in a real-time PC. Study was also focus on tracing the optimized physical parameter trajectory. Furthermore, the entire fermentation control system has been successfully constructed.
出处
《微计算机信息》
北大核心
2008年第25期313-314,156,共3页
Control & Automation
基金
江苏省科学技术厅项目名称:2006科技攻关计划项目(工业部分)(BE2006090)总线化水环境智能多参数高精度检测及控制仪表
关键词
发酵
神经网络
遗传算法
fermentation
neural network
genetic algorithm
