Optimization of Sparse Linear Arrays based on the algorithm of Multiple Objective Particle Swarm

For the multi-objective optimization problem of sparse linear arrays (SLA), the SLA is divided into a full array area adjacent to the aperture center and with the element interval equal to half wavelength, and a sparse array area far from the aperture center. Then, a set of random variables is introduced and optimized by the algorithm of multiple objective particle swarm optimization (MOPSO) to ensure that the element spacings in the sparse area are greater than half wavelength, and present a density tapering distribution during the optimization process. Consequently, the solution space is reduced, which is beneficial to the rapd convergence of the algorithm. Finally, Combined with the combination of two kinds of optimization objectives, the numerical simulation is perfomed based on SLA with different apertures. Comparing with several exsiting algorithms, the results show that this algorithm can reduce the level of sidelobe suppression by about 0.17-2.89dB and the beam width by 0.16 degrees at most. In addition, it can effectively reduce the level of the designated nulltrap region without obviously raising the sidelobe level. The running time of the algorithm is only 2~3 minutes.