Optimal mixed H2/H∞ control based on initiative Markovian packet dropout

Considering the network resource constraints, the networked control systems with periodic packet dropouts are treated as a varying sampling rate system. Refer to the non-periodic packet dropout that could be described as a Markov process, the varying sampling rate networked control system could be modeled as a discrete-time Markov jump linear system. A relation between the system´s H₂ performance and the Markov chain´s transition matrix is established, and an algorithm is proposed to address an optimization problem whose solution yields the transition matrix that optimizes closed loop H₂ performance for fixed average dropout rates. Meanwhile according to the transition matrix, an optimal mixed H₂/H_∞ controller has been designed. Finally, a simulation of a networked flight control system is given to validate the effectiveness of the proposed method.