为解决对弹道初始发射参数多变量辨识困难的难题,提出了基于神经网络逆模型的辨识方法.采用对象-正模型-逆模型学习法,用正模型来模拟弹道方程组机理模型,然后结合逆模型进行辨识.神经网络选用动态回归网络,利用MATLAB仿真结果表明,该方法有效可行. In order to solve the problem of multivariate identification of ballistic initial launch parameters, an identification method based on inverse neural network model is proposed. Using the object-positive model-inverse model learning method, the positive model is used to simulate the mechanism model of ballistic equations, The model is identified.The neural network chooses the dynamic regression network, the result of MATLAB simulation shows that the method is effective and feasible.