Model following adaptive control based on least squares method

This paper describes a design method of adaptive control based on least squares (LS) method to the second-order plant with known parameters. To use the LS algorithm properly, the system must be discretized by a zero-order hold. Although adaptive control based on LS without a parallel feedforward compensator (PFC) may be unstable, the proposed system tracks the reference model, and it has no steady-state error in the disturbance response. The PFC improves the plant based on the almost strictly positive real condition, and an integrator improves the steady-state response. Furthermore, the proposed system has the advantage of simpler design compared with the conventional simple adaptive control system. The proposed method does not require the satisfaction of the perfect model following condition.