LAUSR

Linear quadratic tracker (LQT) is usually employed to solve unconstrained tracking problems but falls short when dealing with systems exhibiting actuator saturation. This paper presents a novel quasilinear quadratic tracking method specifically designed to address this scenario. Firstly, the stochastic linearization (SL) approach is utilized to approximate the saturation nonlinearity with equivalent gains and biases using statistical properties of its input, which are thus incorporated into the system model so as to eliminate the nonlinearity. Then, different time scales are applied in the tracking controller and states in order to improve tracking accuracy. In addition, to reduce computational complexity, two algorithms are provided for approximating the equivalent gains and biases, catering to both scalar and vector control signals. Finally, the proposed algorithms are evaluated through numerical examples, demonstrating their effectiveness and superior tracking performances.