本文阐述了自适应干扰对消的基本原理,详细讨论了利用神经网络求解自适应滤波器最佳加权矢量的方法,并在此基础上介绍了一种基于神经网络的用于对消引信泄漏信号的自适应对消器。这种神经网络可以在电路时常数数量级内给出与最佳加权矢量任意接近的近似加权矢量,而且不存在编程复杂性问题。 This paper describes the basic principles of adaptive interference cancellation, discusses in detail the use of neural network to solve the optimal weight vector adaptive filter method, and on this basis introduced a neural network based on the dereferencing signal leakage signal Adaptive canceller. This neural network can give an approximate weight vector that is arbitrarily close to the optimal weight vector within the time-constant order of the circuit, and there is no programming complexity issue.