Recursive estimation for reduced-order state-space models using polynomial chaos theory applied to vehicle mass estimation

The main contribution of this paper is to present a recursive estimation/detection technique for reduced-order state-space systems. The recursive state and parameter estimator is built on the framework of polynomial chaos theory and maximum likelihood estimation. The estimator quantifies the reliability of its estimate in real-time by recursively calculating a signal-to-noise ratio. The signal-to-noise ratio (SNR) indicates how well the output of the reduced-order estimation model matches the actual system output. A detection algorithm makes decisions to trust or distrust the current estimate by comparing the current value of the SNR ratio against a threshold value. This paper applies the proposed techniques to estimate the sprung mass of an actual vehicle. It uses a reduced-order model to approximate the complex ride dynamics of the vehicle. Despite the modeling approximations and simplifications, the proposed technique is able to reliably estimate the sprung mass of the vehicle to within 10% of the true value.