Approaches For Computation Service Selection And Scheduling In Multidisciplinary Design

Complex product design usually needs machinery, control, electronics, and many other fields of knowledge, the process is a complex system which includes a number of subsystems. From the system view, the performance of Complex product design not only depends on the single subsystem, but also on the interaction between these subsystems. Hence multidisciplinary design optimization is needed. The synergy of design is an important research direction in multidisciplinary design, so the computation service scheduling becomes an important link to be optimized.Computation service scheduling is a NP-hard problem in most applications. It can be generally described as:a task set to be allocated to a group of servers, under certain restrictions, how to map the computing services and processors to achieve the minimum runtime, that is, to achieve load balancing. So the research of computation service scheduling has great realistic significance.Based on the analysis of the current status of computation service scheduling and various research methods, a new hybrid algorithm based on heuristics rules and GA is proposed to solve the multi-constrained hybrid computation service scheduling problems. A two-step scheduling scheme is formulated in which the first step is selecting and the later is scheduling. On the first step, a designed simple genetic algorithm, which cancelled mutation and modified the selection and recombination strategy, is used to select the computation services. On the second step, a based on heuristics rules method is used to schedule the selected computing services. The scheduling results could be displayed in the form of Gantt chart. MATLAB is used to program the algorithm. The simulation results and comparison with others'verified the effectiveness of the proposed algorithm.