Constrained zonotopic Kalman filtering based state estimation algorithm for nonlinear system under unknown disturbance

A constrained zonotopic Kalman filtering is pro-posed to solve the state estimation problem of nonlinear system with state constraints under unknown disturbance. The nonlinear model is firstly linearized by Stirling interpolation and the difference of convex (DC) programming method is used to outer-bound the linearization error. After linearizarion, the state constraints of the nonlinear system are extended to the output vector of the linearized system. Then, the zonotope is used to represent the error boundary and the feasible state set. In the update step, the strip incorporating the linearization error is intersected with the predection zonotope to obtain the update zonotope. The effectiveness and accuracy of this algorithm are demonstrated by the electrothermal coupled model of Li-ion battery.