摘要
为了研究汽车灯泡光通量的变化规律,利用基于MATLAB的粒子群算法优化BP神经网络对汽车灯泡光通量试验数据进行非线性拟合分析,即用粒子群算法对目标函数进行改进,寻到最优权值和阈值应用于BP神经网络。比较改进神经网络(PSO-BP)算法与最小二乘法以及BP神经网络算法的拟合结果,结果表明改进神经网络(PSO-BP)算法的拟合能力显著提高。
In order to research the variation law of the luminous flux of vehicle lamp,a nonlinear fitting method based on the particle swarm optimization algorithm combining BP network was used to improve the objective function.The optimal weights and thresholds obtained by PSO were applied to the neural network.The comparison of the PSO-BP algorithm with least squares and BP neural network algorithm results shows that the nonlinear fitting ability of the PSO-BP neural network algorithm is significantly improved.
作者
黄高益
潘仕刚
郭锐
HUANG Gaoyi;PAN Shigang;GUO Rui(Liuzhou Customs,Liuzhou Guangxi 545001,China)
出处
《汽车零部件》
2019年第6期29-33,共5页
Automobile Parts
关键词
光通量
BP神经网络
粒子群算法
最小二乘法
Luminous flux
BP neural network
Particle swarm optimization
Least squares method
