基于LSTM神经网络的弹药消耗预测

Prediction of Ammunition Consumption Based on Long Short Term Memory Neural Network

针对当前考虑弹药消耗多重影响因素所反映的规律不够客观,长时间弹药消耗没有考虑其消耗规律等问题,提出了利用长短期记忆神经网络(long short term memory,LSTM)来分析弹药消耗的规律。通过示例数据的训练集和测试集,来进行弹药消耗的预测。通过对比RNN模型和BP神经网络模型在测试集上的平均绝对误差(mean absolute error,MAE)和均根方差(root mean square error,RMSE),LSTM神经网络在MAE和RMSE上的误差小,对于长时间序列弹药消耗数据有着很好的预测效果。

In view of the fact that the law reflected by considering multiple influencing factors of ammunition consumption is not objective at present,and the law of ammunition consumption is not considered for long-term ammunition consumption,etc.,a Long Short Term Memory neural network(LSTM)is proposed to analyze the law of ammunition consumption.Through the training set and test set of sample data,the ammunition consumption is predicted.The Mean Absolute Error(MAE)and Root Mean Square Error(RMSE)of RNN model and BP neural network model in the test set are compared,the error of LSTM neural network on MAE and RMSE is much smaller,so it can be seen that LSTM model has a good prediction effect on long-time series ammunition consumption data.