摘要
微生物燃料电池(microbial fuel cell,MFC)反应底物浓度的控制问题是整个系统优化控制的重要环节,其控制效果的优劣对系统的输出电压有很大的影响。针对MFC输出电压在常规控制策略下超调量大和响应速度慢的特点,对MFC系统模型中输入量、控制量的变化对系统输出的影响进行动态仿真;将负载电流作为扰动量,提出了针对MFC系统阳极进料流量进行控制的神经网络预测控制策略。仿真结果表明,与PID控制方法相对比,利用神经网络预测控制策略的系统输出电压响应速度快且超调量小,其动态性能得到了较大的改善。
The control of substrate concentration for microbial fuel cell（MFC） is an important part of the entire MFC system, which have a great effect on the output voltage of MFC. The effects of input variables and control variables on the output voltage of MFC are studied, and a neural network predictive control strategy for anode feed flow of MFC is proposed, in which the load current is regarded as disturbance, aiming to solve the problem of overshoot and slow response of output voltage under conventional control strategy. The simulation results show that, compared with the PID control method, the system output voltage response of the neural network predictive control strategy is fast, the overshoot is small, and the dynamic performance of system is greatly improved.
出处
《化工学报》
EI
CAS
CSCD
北大核心
2017年第3期1090-1098,共9页
CIESC Journal
基金
国家自然科学基金项目(61563032)
甘肃省自然科学基金项目(145RJZ024
145RJYA313)~~
关键词
微生物燃料电池
神经网络
动态仿真
预测控制
microbial fuel cell
neural network
dynamic simulation
predictive control
