基于α阶神经网络逆系统孵化过程解耦控制

Decoupling Control of Poultry Hatching Process Based on α-order Neural Network Inverse System

针对家禽孵化过程是一个复杂的生物过程,温度、湿度和氧气浓度之间耦合严重,常规的控制方法难以保证3个参数之间的稳定,提出一种基于α阶神经网络逆系统家禽孵化过程解耦控制方法。将3个参数进行耦合度分析分组,采用α阶神经网络逆系统控制耦合度较强的2个参数,另一个则通过模糊控制器控制,从而保证3个参数的稳定。实验结果表明了该方法的有效性。

The poultry hatching process is a complex biological process, in which coupling between temperature, humidity and oxygen concentration is serious, and conventional control methods can not guarantee the stability of three parameters. An a-order inverse system decoupling control method for poultry hatching process was proposed. The three factors were grouped by coupling degree analysis, using a-order neural network inverse system to control the two parameters whose coupling degree is stronger, another parameter would be controlled by fuzzy controller, thus ensuring the stability of the three parameters. The experimental results show the effectiveness of this method.