| The antenna aperture of spaceborne radar is limited by the physical size of a single satellite carrier platform and launch is expensive,which reduces angular resolution,detection range and ability to detect moving targets.The distributed satellite array antenna system can break through the aperture limitation of traditional centralized satellite antennas,improve moving target detection capabilities,robustness,and reduce risk of launch.However,the distributed array will produces higher side lobe and more grid lobe in antenna pattern,because of the non-uniform arrangement of the elements,the layout of array elements needs to be optimized.The optimization of spaceborne non-uniform distributed array antennas is a complex nonlinear problem that is difficult to solve by traditional analytical methods.Therefore,this thesis focuses on how to apply intelligent optimization algorithms to reduce the peak sidelobe level of spaceborne distributed arrays.For spaceborne one dimensional non-uniform distributed arrays,this thesis discusses the two cases of equal sub-array size and variable sub-array size.In order to apply the genetic algorithm to optimization of large arrays,solve the problems of genetic algorithm that are prone to infeasible solutions,slow convergence speed,and large amount of calculations.This thesis designs a one-dimensional non-uniform distributed array with equal sub-arrays based on the position coding method of the sub-array interval and the improved single-point crossover method,the position of the distributed sub-array is optimized under the constraints of the minimum interval of the sub-arrays and total length of the array,and obtain lower peaks sidelobes of the one-dimensional non-uniform distributed array with the same subarrays.For arrays with variable sub-array size,this thesis designs a joint coding method based on wavelength quantization of sub-array position and sub-array size,modified crossover and mutation operation methods,and satisfies the minimum gap of sub-arrays,total number of elements,and maximum sub-arrays.Under the constraint of the minimum and minimum length,the position and size of the sub-arrays are optimized at the same time,finally one-dimensional non-uniform array with different sizes of sub-arrays with lower peak sidelobe levels is obtained.In this thesis,the modified genetic algorithm is extended to the optimization of spaceborne two dimensional non-uniform distributed arrays,and the sub-array-level optimization method is adopted to design the coding method based on the azimuth and range interval,the scale transform selection,and dominant two-dimensional crossover method.Then under the constraints of the minimum and maximum gap of the sub-array,optimizes the position of the sub-array,and obtains two-dimensional non-uniform array with a lower peak sidelobe level,and simulates the influence of the minimum gap of the subarrays on the peak sidelobe ratio of the array is discussed.This thesis verifies through simulation that the modified genetic algorithm effectively reduce the complexity and calculation amount of iterative operations when applied to onedimensional and two-dimensional array optimization,avoid the emergence of infeasible solutions,and effectively improve genetic algorithm's global optimization capability and the performance of the spaceborne distributed array antenna.In addition,this thesis introduces quantumfi-behaved particle swarm optimization to onedimensional and two-dimensional non-uniform distributed arrays respectively,and compares and analyzes the process with genetic algorithm and particle swarm algorithm.The results show this algorithm has better global optimization ability and fast convergence performance than genetic algorithm and particle swarm algorithm.Finally,through clutter simulation and target detection performance simulation experiments,the pattern performance of the optimized one-dimensional and two-dimensional nonuniform distributed antenna array configuration is verified respectively. |
PDF Full Text Request