Research On Array Antenna Faulty Elements Diagnosis And Pattern Correction Algorithms

The array antenna is composed of a plurality of radiating antenna elements.When a certain number of array elements fail in the array,the array radiation pattern will be distorted to different extents,which affects the normal use of the array system seriously.Therefore,it is very necessary to analyze the influence of the failed array elements on the distortion of the array pattern,determine the position and number of the failed array elements in the array,and reduce the distortion of the pattern by using the optimization algorithm to restore the performance of the pattern.Based on the calculation of array antenna pattern,the influence of failure array elements on array pattern is analyzed.The failure element location method based on far field pattern is proposed,and the optimization algorithm of distortion pattern is designed.The effectiveness of the algorithm is verified by diagnostic and optimization examples.The final experimental results show that the neural networks algorithms based on BP neural network and extreme learning machine can effectively locate the array elements in the array.The population optimization algorithm using genetic algorithm and firefly algorithm can largely correct the distortion pattern,thus achieving the target of distortion analysis and diagnosis and optimization technology of array pattern.The research work in this thesis is as follows:First,the impact of the failed array elements on the array pattern is analyzed.The calculation method of uniform line array,circular array and area array is introduced.The influence of array element failure on uniform linear array,circular array and planar array pattern is analyzed.The distortion of the maximum sidelobe level,the first null beamwidth and the half power beamwidth of the three typical uniform array patterns with elements failed on different position are verified by simulation.Taking the uniform planar array of 1024 as an example,the effects of array element failure,T/R component failure and sub-array failure on the array pattern under different failure rates are analyzed,and the deterioration of the average sidelobe level of the pattern is discussed through simulation.Second,the failure array element diagnostic is researched.Based on the BP neural network and the extreme learning machine of the neural networks algorithms,the neural network model is trained by using the array pattern in different failure modes,and the trained network model is used to diagnose the position and number of the failed array elements in the array from the far field pattern with random errors.Finally,the performance comparison between the two algorithms for diagnosing the failed array elements is discussed.The simulation proves that extreme learning machine has faster learning speed than the traditional BP neural network algorithm and greatly improves the diagnostic efficiency.Third,the pattern correction optimization algorithm is researched.The genetic algorithm and the improved firefly algorithm are used to redesign the excitation of the remaining array elements through repeated iterative processes,so that the pattern of the array distortion can restore its performance to a large extent,thus achieving the correction of the pattern.Finally,the characteristics of the two array pattern optimization algorithms are compared and analyzed.The simulation results show that the firefly algorithm based on parameter variance adjustment has better optimization effect,which makes up for the low rate of discovery and low convergence rate of the standard firefly algorithm.It has a faster convergence speed than the genetic algorithm.