The number and locations of deployed sensors are the key technology to be solved for obtaining high precision modal parameters of large satellite antenna reflector on-orbit. To avoid the limitations and one-sidedness of single-objective criterion used in the previous works, the double-objectives criterions are designed, which contain the largest orthogonality and the highest vibration energy of observed signals. NSGA-II algorithm is introduced in solving multi-objectives optimization problem of sensor deployment. Considering the constraints of continuous variables and the deficiencies of convergence and diversity in NSGA-II, the coding and genetic operators is improved. Comparing with the results of four cases in different single-objective and multi-objectives criterion, the solutions based on improved NSGA-II are the best in these four cases according to all indicators, which have greater flexibility and adaptability. This case can obtain the uniform distributed Pareto solution set, which includes all weight combination of multi-objectives. It is more significant for the design requirements in realistic engineering.
PDF(1814 KB)
