节点文献
基于BP神经网络的筒承式群仓自振特性预测研究
Research on Predication of Natural Vibration Characteristics of Cylinder Multi-silos Supported by Silos Wall Themselves Based on BP Neural Network
【摘要】 阐述了组合筒数N、高径比H/D、开洞率、组合方向与开洞方向的关系R及装料程度MF等因素对筒承式群仓自振特性的影响,并在此基础上构建了基于BP神经网络的筒承式群仓自振特性预测模型。将BP神经网络模型预测值分别与有限元模型、简化的杆系结构模型和简化公式模型的计算值对比分析,验证了该方法的可行性和准确性。计算结果表明,预测值与实测值吻合较好,该模型性能良好、预测精度较高,能满足工程的要求,具有一定的价值。
【Abstract】 The following influencing factors on the natural vibration characteristics were analyzed:the amount of combined silos N,height-diameter ratio H/D,ratio of opening hole ,relation between combined direction and opening direction R,and degree of material filling MF.Based on this,BP neural network prediction model of natural vibration characteristics of cylinder multi-silos supported by silos wall themselves was proposed.Compared with the calculations of the finite elecment model,simplified structure model of bar systems and simplified formula model,the feasibility and accuracy of the methodology were validated.The results indicated that the predicted data agreed well with the measured ones.It was shown that the method was feasible,effective and satisfied with engineering demand,which can be used in projects.
- 【文献出处】 特种结构 ,Special Structures , 编辑部邮箱 ,2007年02期
- 【分类号】TU311.3
- 【被引频次】1
- 【下载频次】69