Scholarlink

Purpose <p>In order to realize the prediction of gear vibration characteristics under the influence of random uncertain parameters, a prediction model of gear transmission system vibration characteristics based on neural network is established.</p> Methods <p>The random uncertainty parameters of combined meshing stiffness and tooth surface friction coefficient are quantified, and the data set is constructed according to the bending-torsion coupling vibration model of gear transmission. Then, the BP neural network algorithm is used to establish the vibration characteristic prediction model of the gear transmission system, so as to realize the efficient prediction of the vibration characteristics of the gear transmission system. Finally, the accuracy of the model is verified by experiments.</p> Results <p>Under the influence of random uncertainty parameters of combined meshing stiffness and tooth surface friction coefficient, the prediction error of vibration acceleration is 2.26%, the prediction error of meshing force (time interval is 0.01&#xa0;s) is 2.22%, and the prediction error of meshing force along meshing line is 1.76%. The error between the test and prediction results is about 10%.</p> Conclusion <p>Mean-square error (MSE) and mean absolute error (MAE) values are all less than 0.1, and the absolute value of the average relative error is less than 5%, indicating that both the prediction ability and accuracy of the model meet the requirements. The vibration amplitude range and overall change trend of the predicted results are basically consistent with the experimental results, which verifies the accuracy of the prediction model.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Prediction of Vibration Characteristics of Gear Transmission Based on Parallel BP Neural Network

  • 期刊论文

  • 发表日期：

    2025年5月5日

查看全文
  • Shengping Fu,
  • Xudong Li,
  • Yaou Luo,
  • Jingyu Mo

摘要

Purpose

In order to realize the prediction of gear vibration characteristics under the influence of random uncertain parameters, a prediction model of gear transmission system vibration characteristics based on neural network is established.

Methods

The random uncertainty parameters of combined meshing stiffness and tooth surface friction coefficient are quantified, and the data set is constructed according to the bending-torsion coupling vibration model of gear transmission. Then, the BP neural network algorithm is used to establish the vibration characteristic prediction model of the gear transmission system, so as to realize the efficient prediction of the vibration characteristics of the gear transmission system. Finally, the accuracy of the model is verified by experiments.

Results

Under the influence of random uncertainty parameters of combined meshing stiffness and tooth surface friction coefficient, the prediction error of vibration acceleration is 2.26%, the prediction error of meshing force (time interval is 0.01 s) is 2.22%, and the prediction error of meshing force along meshing line is 1.76%. The error between the test and prediction results is about 10%.

Conclusion

Mean-square error (MSE) and mean absolute error (MAE) values are all less than 0.1, and the absolute value of the average relative error is less than 5%, indicating that both the prediction ability and accuracy of the model meet the requirements. The vibration amplitude range and overall change trend of the predicted results are basically consistent with the experimental results, which verifies the accuracy of the prediction model.