Worst-Case Tolerance Synthesis for Low-Sidelobe Sparse Linear Arrays Using a Novel Self-Adaptive Hybrid Differential Evolution Algorithm

A worst-case tolerance synthesis problem for low-sidelobe sparse linear arrays is solved by using a novel self-adaptive hybrid differential evolution (SAHDE) algorithm. First, we establish a worst- case tolerance synthesis model for low-sidelobe sparse linear arrays, in which random position errors are considered and assumed to obey the Gaussian distributions. Through the random sampling, the random model is converted to a deterministic optimization problem. Then, a novel SAHDE algorithm is presented for solving the problem. As a modification to the existing hybrid differential evolution algorithm, a simplified quadratic interpolation (SQI) operator is used to tune the control parameters self-adaptively, establishing the connections between control parameters and the fitness values. In order to determine appropriate control parameter values quickly, a selection operation is also used. Detailed implementation procedure for the SAHDE algorithm is presented, and some numerical results show its effectiveness. Finally, for the deterministic optimization problem, we present a fast way for calculating its fitness values. The SAHDE algorithm is used to obtain optimal nominal element positions. Simulated results illustrate that the worst-case peak sidelobe levels for the sparse linear arrays are improved evidently. The SAHDE algorithm is efficient for solving the worst-case tolerance synthesis problem.