摘要
在弹药试验品日常存储温度监测过程中,传统传感器测量存在滞后性。为解决这一问题并实现试验品储存室下一时刻温度的精准预测,文中提出一种基于SSA优化BP神经网络的智能算法。通过SSA算法与BP神经网络相结合的方法,在局部搜索中快速找出阈值更新的最优位置,为BP神经网络的训练提供更好的参数。利用Matlab仿真平台搭建SSABP温度预测模型,并与PSO-BP算法温度预测模型进行仿真对比。测试结果表明:SSA-BP神经网络算法稳定性好,鲁棒性强,收敛速度快;相比PSO-BP网络,该算法的MAE和MSE误差值分别减少2.31%和0.54%,预测精准度高。所提方法可为弹药试验品储存室温度精准预测提供重要依据和参考。
In order to overcome the hysteresis caused by the traditional sensor measurement and realize the accurate prediction of temperature at the next moment in the daily storage temperature monitoring process of ammunition test materials,an intelligent algorithm based on SSA to optimize BP neural network algorithm is proposed. By combining of SSA algorithm and BP neural network,the optimal location of threshold update can be quickly found in local search,providing better parameters for the training of BP neural network. Matalb simulation platform is used to build SSA-BP temperature prediction model and compare it with PSO-BP algorithm temperature prediction model. The testing results show that SSA-BP neural network algorithm has good stability,strong robustness and fast convergence speed. In comparison with PSO-BP network,the MAE and MSE error values of the algorithm are reduced by 2.31% and 0.54% respectively,and the prediction accuracy is high. The proposed method can provide important basis and reference for accurate prediction of the temperature of ammunition test article storage room.
作者
孙冲
刘沛然
伊猛
SUN Chong;LIU Peiran;YI Meng(School of Weapon Science and Technology,Xi’an University of Technology,Xi’an 710021,China;School of Electronic Information Engineering,Xi’an University of Technology,Xi’an 710000,China)
出处
《现代电子技术》
2023年第4期171-176,共6页
Modern Electronics Technique
基金
陕西省教育厅自然科学专项(19JK0411)
陕西省教育厅自然科学专项(16JK1387)。
关键词
SSA算法
PSO算法
BP神经网络
弹药存储
温度预测
模型搭建
仿真验证
SSA algorithm
PSO algorithm
BP neural network
ammunition storage
temperature prediction
model building
simulation verification
