Delay Time Identification and Dynamic Characteristics Study on ANN Soft Sensor

Abstract

Soft sensor software based on ANN (artificial neural network) using BP or RBF was developed to estimate unmeasured variables such as product quality online. Some important topics including how to determine the delay time, how to simulate the dynamic system were discussed and solved. We applied a 3 layers BP network to identify the delay time of nonlinear system, feedback output variables to input layer, and weight of all the input variables to describe dynamic characteristics of the system. This makes the ANN soft sensor reflect truly both the static and dynamic characteristics of the system and provide more adaptability