Measurement Delay Estimation for Kalman Filter in Networked Control Systems

This work presents a method to estimate an unknown varying delay of measurements and subsequent Kalman filtering in the networked control system (NCS) framework. The delay estimation algorithm is based on the Gaussian error model and a network delay model, either a probability distribution or a Markov-chain. This method is used to tackle the problem with varying delays in a NCS. The undelayed output of the plant is estimated with a Kalman filter and used for control. The probability of a wrong delay estimate is derived. The estimation and control performance is evaluated with simulations. From the control perspective, it has comparable performance to the case with known delays.