Actuator linearisation for multivariable control applications

Fixed speed actuators have a characteristic nonlinear transient response. While this may not present a problem for single-input, single-output (SISO) control systems containing such actuators, the mixture of different fixed speed actuators is usually problematic in multivariable control applications. This paper presents two methods which allow such nonlinear actuators to present a linear response and therefore allow their incorporation in linear multivariable control design methods. The first method is based on a closed-loop describing function approximation, while the second uses an equivalent nonlinearity concept to linearise the relay which represents the fixed-speed limitation. A simulation study, demonstrating the performance of both schemes in a multivariable control setting, is presented.