In petroleum chemic industry, device is synthetic and complicated. Its operating parameters are interacted and related one another. So, any one changed irrationally, the device would be affected. With a number of parameters, it is difficult to make petroleum chemic device optimum, only on artificial adjustment.Prediction-controlling is a optimum controlling method constrained multiple variables. Since 1970's, it is applied to petroleum chemic industry widely.About complicated and nonlinear system, the model of Prediction-controlling is not perfect.With back promulgated neural network, a nonlinear model of double linear system is set forth in this paper. It would be used in single- value-Prediction-controlling.
|