A New Process Neural Network Algorithm of Soft Sensor for Time-Varying System

• Published in: 2006 International Conference on Computational Inteligence for Modelling Control and Automation and International Conference on Intelligent Agents Web Technologies and International Commerce (CIMCA'06) p. 166
• Main Authors: LIU, Zaiwen, WANG, Xiaoyi, LIAN, Xiaoqin, Wang, Zhengxiang, HOU, Chaozhen
• Format: Conference Proceeding
• Language: English
• Published: IEEE 01.11.2006
• Subjects: Artificial neural networks; Chaos; Computational intelligence; Data processing; Mathematical model; Mathematics; Neural networks; Neurons; Sensor systems; Time varying systems
• ISBN: 0769527310; 9780769527314
• DOI: 10.1109/CIMCA.2006.21

A new algorithm of data processing and a method of soft sensor based on process neural network (PNN) for time-varying system are represented in the paper. Process neural network is an extension of traditional neural network, in which the inputs and outputs are time-variation. An aggregation operator is introduced to process neuron, and it makes the neuron network has the ability to deal with the information of space-time two dimensions at the same time, so the data processing enginery of biological neuron is imitated better than traditional neuron. Through analyzing the reasons of low-speed constringency of the network, two improvements on the foundation of original algorithm were put forward, and an improved algorithm for BP network and algorithm research based on function orthogonal basis expansion in process neural network for soft sensor were discussed. A good training result of soft sensor was obtained by simulation in sewage disposal system.