基于专家经验的单神经元串级温差控制研究

针对置换蒸煮锅内存在的温差及温差对象的非线性严重、大时滞、要求控制实时等特点,提出了保证蒸煮锅内温度一致的、基于专家经验的单神经元串级PID自适应控制策略。该控制策略引入专家经验调节单神经元增益K,同时将单神经元PID自学习和自适应的能力与串级优势相结合,可实时快速地进行温差控制。在Simulink中,调用该控制策略的s函数,进行温差对象特性的动态仿真实验,以验证其鲁棒性和模型失配响应。结果表明：该控制策略的鲁棒性与自适应性比常规温差-流量串级PID控制及单神经元PID串级控制更好；THJSK-1平台中的实时控制还表明,该控制策略具有可行性。

Single Neuron Cascade PID Control Based on Expert Experiences for Temperature Difference Control System of a Digester

Aiming at the temperature difference existing in digester, in order to ensure uniform temperature, this paper proposed a control strategy combining the advantages of the cascade control and single neuron PID control. It also introduced the gain K of single neuron regulated by the expert experiences thus could control the temperature difference moment by moment and rapidly. In the Simulink, the s-function of this control strategy was used to carry out dynamic simulation experiment of temperature difference characteristics, and verify the robustness and response to model mismatch of this control strategy. Compared to conventional temperature difference-flow PID cascade control and single neuron PID cascade control, this control strategy had better robustness and stronger adaptiveability. The results of real-time control on THJSK-1 experiment platform indicated this control strategy was feasible.