Performance improvement of multivariable linear system with unmeasured external disturbance

In this paper, we present a system design criterion which improves the performance of the general multivariable linear system with unmeasured external disturbances. We use the Lyapunov function to find the upper bound of the integral square error (ISE) which serves as a measurement of system performance in process control. We point out the relationship between the upper bound of ISE and the related convergence rate. The design criterion shows how the system matrix plays an important role in reducing overall ISE. The results presented can be used as a general system design criterion for a multivariable linear system with better disturbance rejection