Statistical process adjustment of multivariate processes with minimum control efforts

In controlling a multiple-input-multiple-output (MIMO) process, usually all control variables have to be adjusted at each step, which may incur high adjustment cost. This paper proposes a Lasso adjustment algorithm, which minimizes the number of variables to be adjusted at each step. Simulation results show that the proposed algorithm can maintain acceptable output deviations while reduce the number of variables need to be adjusted significant.