Synthesis of unequally spaced linear arrays using modified differential evolution algorithm

Unequally spaced linear arrays are synthesised for the sidelobe level (SLL) suppression, the null control in some specified directions and the beamwidth (BW) set within predefined limits. Based on the vector mapping method and the position perturbation technique, the constrained vector projection (CVP) is defined to offer flexibility in controlling null depths with multiple constraints. A modified differential evolution (DE) algorithm with the CVP method (in short MDE-CVP) is then performed in the array synthesis, which effectively reduces the optimisation space, avoids the infeasible solutions and enhances the global search ability. Three examples are presented and the first two results are compared with those obtained by learning particle swarm optimisation, invasive weed optimisation and wind-driven optimisation and a modified version of the central force optimisation algorithms. Comparison results indicate that the MDE-CVP algorithm outperforms the existing algorithms in terms of minimum SLL, BW control, null control and convergence characteristics.