基于神经网络的逆控制法在热工过程中的应用仿真

Neural Network Based Inverse Control Method Used in Thermodynamic Process

逆控制法通过逆建模使串联控制器拥有与被控对象互为倒数的传递函数。将逆控制和神经网络相结合可以使逆控制法的控制效果都得到改进。以经典热工过程为控制对象,通过逆控制器与常规 PID 控制器的控制效果仿真比较,表明这种方法具有很好的快速反应性、精确的控制效果以及很强的自适应能力。

The inverse control method is a new approach for procedure control which makes the controller have a transfer function that is inverse to the controlled object by inverse modeling.The combination of inverse control and neural network improves the quality of inverse control.Based on a typically thermodynamic control object,the fea- sibility of thermodynamic process control method is explored by comparing the simulation result of the inverse con- troller and normal PID controller under the circumstance of Matlab.It is shown that the inverse control method has some properties of rapid reactivity,preciseness and strong adaptability.