Combinatorial scheduling rules optimisation has been a focus of most of the studies about flexible manufacturing systems (FMS), in the design and control of which, the performance optimisation of a machine cell is of great importance. To optimise the performance of a machine cell, the paper describes the design of a simulation optimisation system. It is constructed according to the principle of separating the models from the algorithms. The system consists of a simulation module and an optimisation module. The optimisation algorithms are the core of the paper, a new hybrid optimisation algorithm is proposed by combining the features of region-shrunk genetic algorithms with an enhanced continuous tabu search. The paper provides an example to demonstrate the validity of the system.