Design a low price and high performance sensor network using constrained-based unscented Kalman filter

This paper presents a new methodology to design an instrumentation sensor network for nonlinear processes. The goal is to design a high performance and low price sensor network. The proposed approach utilizes constrained state estimation based on the unscented Kalman filter (UKF) approach to cater for some known signal information which is often either ignored or dealt with heuristically. The approach employs a clipping methodology to implement the constraints, leading to an optimal sensor network with less number of sensors. A benchmark continuous stirred tank reactor (CSTR) is used to evaluate the performance of the new method. The obtained simulation results validate the effectiveness of the new method.