The optimal solution of multi-objective control problem (MOCP) isn't unique, so it is hard for traditional method to obtain these optimal solutions in one simulation process. Based on this background, Multi-Objective Compatible Control (MOCC) algorithm was presented by Lihong Xu in . MOCC is a compromise method, which hunts for suboptimal and feasible region as the control aim rather than precise optimal point. The controller of MOCC is optimized by GA in its interval, namely its controller lacks concrete controller structure. Due to the controller without concrete structure, the system model must be accurate and without input disturbance; however, it is impractical in practice. Besides, the control problem is different from the optimization. Different user has different preference and users' preference information plays a key role in control performance. In this paper, the feedback control law u = Lx and users' preference information are incorporated into MOCC algorithm. An improved MOCC (IMOCC) algorithm is presented. The simulation result demonstrates its superiority and advantage over the MOCC algorithm.