空调滞后对象的神经网络预测控制

ANN predictable control of air conditioning time-delay system

在空调控制过程中,广泛存在惯性滞后对象,用常规的控制方法很难取得良好的控制效果。文章结合神经网络和预测控制的优点,提出了神经网络预测控制方法,并对某一空调系统滞后模型进行了仿真。从仿真结果看,此种方案在对不同对象的适应性、响应速度、抗干扰及稳定性等方面均有明显的改善,更符合实际工业过程控制的特点。

There is a lot of inertia time-delay system,in the process of air conditioning control and it is difficult to get good effect with normal control measure. The paper presents a ANN predictable control measure based on the advantage of ANN and predictable control, simulates the air conditioning time-delay system model. According to the result of simulation, the project is consistent with the characteristic of practical process control in adaptability, response velocity, anti-jamming and stability for different system.