论文部分内容阅读
作者提出了观察方向差分约束多副瓣抵消问题的Hopfield神经网络求解方法:对常规Hopfield神经网的基本模型加以改进,得到了适合于求解差分约束多副瓣抵消问题的一种Hopfield神经网络模型。为了进一步提高实时处理能力,避免在运用神经网之前求解空间采样协方差矩阵,我们还引入了一种新的结构模型。分析和仿真实验结果表明,Hopfield神经网络方法比传统的线性自适应方法有更快的收敛跟踪速度。
The authors proposed a Hopfield neural network method to solve the problem of multi-sidelobe cancellation based on the observation of differential constraints. This paper improves the basic model of the conventional Hopfield neural network and obtains a Hopfield neural network model suitable for solving the multi-sidelobe cancellation problem with differential constraints. In order to further improve the real-time processing ability and avoid solving the spatial sampling covariance matrix before using neural network, we also introduce a new structural model. Analysis and simulation results show that the Hopfield neural network method has faster convergence rate than the traditional linear adaptive method.