摘要
选取具有不同结构参数的减震垫进行冲击实验。在此基础上,采用BP神经网络预测模型算法与SPSS多元线性回归方程拟合算法分别探究运动地板减震垫结构参数与反冲击力之间关系,通过冲击实验获取实验数据与两种算法结果对比。结果表明:BP神经网络预测模型算法得出的数据与实验结果较为接近;以反冲击力与体积占比代数和最小作为优化目标,应用BP神经网络预测模型算法对给定参数范围的减震垫进行反冲击力预测,获得最优结构参数,经实验证明优化后的减震垫具有良好的冲击载荷吸收性能。
From the shock absorbers with different structural parameters selected for impact test,the BP neural network prediction model algorithm and SPSS multiple linear regression equation fitting algorithm were used to explore the structural parameters and anti-impact force of the sports floor cushion.The relationship between the experimental data and the two algorithms was obtained by performing the impact test.The data obtained by the BP neural network prediction model algorithm was close to the experimental results;the anti-impact force and volume ratio algebra and minimum were the optimization targets,and the BP neural network prediction model algorithm was applied to the cushion of the given parameter range.The anti-impact force prediction was carried out to obtain the optimal structural parameters.Therefore,the optimized cushion has good impact load absorption performance.
作者
花军
张昊
王宏棣
刘一楠
刘宇宸
Hua Jun;Zhang Hao;Wang Hongdi;Liu Yinan;Liu Yuchen(Northeast Forestry University,Harbin 150040,P.R.China;Key Laboratory of Wood,Institute of Heilongjiang Province Wood Science Research)
出处
《东北林业大学学报》
CAS
CSCD
北大核心
2020年第6期69-74,共6页
Journal of Northeast Forestry University
基金
黑龙江省自然科学基金项目(C201045)。
关键词
运动地板减震垫
BP神经网络
预测模型
参数优化
Sports floor cushion
BP neural network
Prediction model
Parameter optimization