Multi-objective optimization design for trusses of spacecraft antenna

A multi-objective structure optimization method is presented for trusses of spacecraft antenna with appended structures. To establish an accurate finite element model for optimization design, the effect of appended structure stiffness on the truss dynamics characters is discussed. The structure topology configuration is decided using the surrogate model and the man-machine interactive method. For the surrogate model, optimal Latin hypercube sampling method is used as the design of experiments strategy, and radial basis functions are selected as the approximation method. The multi-objective evolutionary algorithm NSGA-II which produces a set of Pareto solutions is employed as the optimization strategy to trade off the minimization of weight and the maximization of natural frequency. The optimum solution is visually identified with the level diagrams which is a graphical representation of the Pareto front and the Pareto set. The results demonstrate that the weight is reduced by 29.66% without decreasing the natural frequency relative to the initial design scheme. This method not only can maintain good design efficiency, but also increase the global search capability. So it is very useful for the multi-objective optimization design of spacecraft structures.