Research On Subarray Partition Method Of Phased Array Antenna Based On Pattern Approximation

Phased array antenna is an antenna that changes the directional pattern shape by feeding phase of each radiation element in the array antenna.By controlling the phase,the maximum direction of the antenna pattern can be changed to achieve the purpose of beam scanning.Phased array antenna is widely used in modern communication,radar,navigation,remote sensing,radio astronomy,biomedical imaging and other fields because of its fast scan speed,easy configuration and flexible beam control.Each element of a conventional phased array antenna is equipped with a transmit/receive device,which allows independent control of the amplitude and phase delay of the received or transmitted signal.Therefore,conventional phased array antennas are expensive.In order to meet market demand and reduce antenna cost,phased array antenna needs to be optimized through subarray technology,which can achieve a good balance between cost and antenna performance.The contents of the dissertation is mainly divided into the following sections:1.Understanding the basic theory of phased array antenna,taking one-dimensional linear array as the research object,the far field pattern theory calculation formula is discussed and given.On this basis,the Taylor array,flat-top beam and cosecant-squared beam are simulated experimentally.Subsequently,the basic idea of subarray technology is discussed,and an intelligent optimization algorithm is used to partition one-dimensional linear array using genetic algorithm(GA)as an example.The validity of this method is verified by applying the genetic algorithm to the subdivision of one-dimensional linear array.2.As a classical clustering algorithm,K-Means clustering algorithm(KMM)is characterized by simple principle,fast convergence and good clustering effect.Therefore,it is applied to the subarray partition of one-dimensional linear array.In the dissertation,the KMM clustering method is applied to three typical subarray partitioning problems of one-dimensional linear arrays under the excitation matching strategy,and the subarray partitioning results are compared with the GA method(far-field pattern,array performance parameters and matching error).The simulation results show that KMM is superior to GA in all dimensions.3.Considering that KMM is a local composite strategy,the result has a lot to do with the initial cluster center selection.So a new strategy is designed so that it is not affected by initialization.KMM is a clustering grouping of two-dimensional space.Hilbert curves map twodimensional planes to one-dimensional curves while maintaining the local characteristics of the data.Therefore,the dissertation uses Hilbert curve to sort the subarrays,and then uses onedimensional clustering grouping algorithm for its results.In the dissertation,one-dimensional linear array with complex value excitation is used as an example to divide the subarray using Hilbert curve,and the simulation results are compared with those of KMM.The simulation results show that the subarray partition based on Hilbert curve is closer to the ideal array in radiation performance than that based on traditional KMM.