Application of Multisensor Data Fusion Based on RBF Neural Networks for Drum Level Measurement

Abstract

Data fusion is the process of combining data from multiple sensors to estimate or predict entity states. The data from individual sensors are noisy, uncertain, partial, occasionally incorrect and usually inherent. Multisensor data fusion seeks to combine data to measure the variables that may not be possible from a single sensor alone, reducing signals uncertainty and improving the accuracy performance of the measuring. In this paper, radial basis function (RBF) neural network and multisensor data fusion are combined and used in drum water level measurement. It is applied several sensors to measure the process variables related with boiler water level, such as drum pressure, temperature, differential pressure, ambient temperature, water inflow and steam outflow, etc, and their relationships always represent the characteristics of nonlinear. The RBF neural network can be thought of as a nonlinear mapping between input variables and output variables. By using the combination method the results of level measurement are more accurate and reliable than the traditional method. The simulation results illustrate that this method is feasible and more effective; the drum level measurement precision can be improved by using this method