A Method for Predicting Active Scattering Pattern of Planar Array Antenna

Abstract

The invention discloses a method for predicting the active scattering pattern of a planar array antenna. The method includes the following steps: 1) dividing the scattering field of a large planar array into several small planar sub-array scattering fields according to the active pattern method; 2. ) to establish a BP neural network correction model; the structure of the BP neural network correction model is divided into three major layers: an input layer, a hidden layer, and an output layer; calculating the output error; and returning the difference between the output result and Δ, According to the error gradient descent method, the correction amount of the weight value is corrected sequentially; the model training is carried out, and the training is completed when the network reaches the convergence target or the predetermined number of iterations; the active direction diagram method based on the trained neural network correction model is used to give a large planar array. Scatter pattern. The invention optimizes the traditional active scattering pattern prediction method of the array antenna through the optimization algorithm of the neural network, can better control the error, and provides a prediction means for grasping the scattering pattern of the large-scale array antenna.

Description

technical field [0001] The invention relates to the communication field, in particular to a method for predicting the active scattering pattern of a planar array antenna. Background technique [0002] Predicting the Scattering Direction of an Array Antenna Figure 1 Generally, numerical algorithms based on finite element and method of moments are used for prediction. However, for larger planar array antennas, the computational time and computational memory resources required by the numerical algorithm increase dramatically, making it difficult to estimate the scattering pattern. Some researchers at home and abroad have proposed an active pattern method for large arrays based on the scattered field calculation of small arrays, which greatly reduces the algorithm's demand for computing time and computing memory resources. However, due to the approximation error of this method, the error of its calculation result is relatively large. Since the approximation error of the activ...

AI Technical Summary

Problems solved by technology

However, due to the approximation error of this method, the error of the calculation result is relatively large.

Since the approximation error of the active pattern method is a nonlinear error, the conventional linear error elimination method and regression method are invalid

Images