摘要
在热分散过程中,由于温度控制具有非线性、慢时变、大滞后的特点,并存在不可测干扰。采用神经网络PID预测控制策略,预测下一时刻的模型输出,并经过反馈校正,结合最优化指标调整BP网络的权值校正PID的调节参数,可以优化对热分散温度控制系统的有效控制。该方法已应用在某纸业制浆控制系统中,取得了温度控制平稳,调节时间短,热分散效率明显提高的效果。
In the heat dispersion process, temperature control has the characteristic of nonlinear, slow time-varying, lagging behind the major and simultaneously unpredictable disturbance. Using the neural network PID predictive control strategy to ＇predict the output of next model time,through feedback correction,combined with the optimal index to adjust the right value of BP network and to correct the regulation parameter of PID,can optimizes the effective control to the heat dispersion temperature control system. This method applied in thick liquid control system of some paper mill makes temperature control steady, control time short, and the heat dispersion efficiency be obviously enhanced, and receives quite good control result.
出处
《纸和造纸》
北大核心
2009年第5期56-59,共4页
Paper and Paper Making
关键词
热分散
温度控制
神经网络PID
预测控制
heat dispersion
temperature control
neural network PID
predictive control
