Research On Array Antenna Beamforming Technology Based On Intelligent Algorithm

With the growth of communication requirements,the traditional single antenna is gradually unable to meet people’s performance requirements for high gain,high directivity,low side lobe level,beam control,beam forming,etc.For this reason,people put forward the scheme of array antenna.In the study of array antennas,beamforming theory has received more and more attention because it can produce customized patterns suitable for specific needs,and is widely used in wireless communications,radar systems,satellite communications,military communications,and mobile communications.The researchers used various intelligent optimization algorithms to optimize the antenna array to obtain the specified shape of the beam.In this paper,a set of Python-CST co-simulation program is written according to the Python3.6 programming interface of CST2020,so that it can modify the CST model in the Python environment,run the simulation and export the simulation results.On this basis,the genetic algorithm is used to optimize the excitation of the patch antenna array to obtain the radiation pattern of the specified shape.In order to be able to optimize the spacing of the introduced array elements under the premise of considering the mutual coupling between the array elements and the platform installation effect,and bring more degrees of freedom to the array optimization,this paper uses the Gaussian process regression model instead of CST for the antenna array pattern.calculate.And on this basis,the improved genetic algorithm is used to optimize the pixel spacing and excitation at the same time,so as to get better beamforming effect.The main work of this paper is as follows:1.Establish Python-CST Api according to the Python3.6 programming interface of CST2020,so that it can control CST in the Python environment for automatic simulation.2.Use the co-simulation program and genetic algorithm to optimize the patch antenna array in CST in the Python environment.In the Python environment,add the excitation current amplitude and phase of the array element to the CST antenna array through the Python-CST Api,run the simulation and export the pattern data,calculate the cumulative error value according to the target beam and optimize it using the genetic algorithm to get Specifies the shape beam.3.The improvement strategy of genetic algorithm is used to improve the performance of the algorithm when facing multi-objective complex function optimization problems.Improving the crossover operation to reduce the time for the optimization problem to converge to the optimal solution,and improve the mutation operation to improve the global search ability of the algorithm in the stable local optimum.4.Use Python to repeatedly change the spacing between the antenna array elements in the CST and run the simulation.After the simulation is completed,export the AEP of each element to generate a training data set.Use the training data to train the GPR,the input data is the array element spacing,and the output data is the AEP of each array element.By selecting the appropriate kernel function and training method,the GPR can obtain relatively high accuracy.The array pattern can be obtained by weighting and summing the AEP of each array element according to the array element excitation.On this basis,the improved genetic algorithm is used to optimize the spacing and excitation of the elements to obtain the beam of the specified shape.