基于神经网络的不确定电路网络电流预测算法

随着人工智能的快速发展,机器学习已广泛应用于军事和民用领域。针对复杂、存在不确定器件且难以用常规分析方法进行分析的电路网络。结合BP神经网络和某型号的部分不确定电路网络的分析需求,提出一种基于神经网络的不确定电路网络电流分析预测方法。首先采用神经网络对不确定电路网络进行建模,然后根据建立的模型对已有电路网络数据进行训练分析。最后根据神经网络的分析数据,完成整个电路网络在含有不确定器件的情况下,不同位置点的电流极端值范围。同时,将预测的数据与采用Multisim仿真的数据进行对比来验证上述方法的有效性。

The Algorithm of Current Prediction for Uncertain Circuit Network Based on Neural Network

With the rapid development of artificial intelligence,machine learning has been widely used in military and civilian fields.Aiming at the complex and uncertain circuit network which is difficult to be analyzed using con-ventional analysis method,this paper proposes a method of current analysis and prediction for uncertain circuit net-work based on neural network.Firstly,the uncertain circuit network was modeled using neural network,and then the existing circuit network data were trained and analyzed according to the established model.Finally,the current ex-treme value range of the whole circuit network with uncertain devices at different locations was completed according to the analysis data of neural network.At the same time,the predicted data were compared with the data of Multisim simulation to verify the effectiveness of the method.