| 王冬磊,张智禹,尹爱军.双基球扁药工艺优化仿真研究[J].装备环境工程,2018,15(7):29-32. WANG Dong-lei,ZHANG Zhi-yu,YIN Ai-jun.Simulation on Process Optimization of Double-based Oblate Spherical Powder[J].Equipment Environmental Engineering,2018,15(7):29-32. |
| 双基球扁药工艺优化仿真研究 |
| Simulation on Process Optimization of Double-based Oblate Spherical Powder |
| 投稿时间：2018-05-09 修订日期：2018-07-25 |
| DOI：10.7643/ issn.1672-9242.2018.07.007 |
| 中文关键词: 双基球扁药 工艺优化 BP 神经网络 成球质量 |
| 英文关键词:double-based oblate spherical powder process optimization BP neural network pelletization quality |
| 基金项目:国防预研基金项目（9140A17050115JW20001）；重庆市人工智能技术创新重大主题专项重点项目（cstc2017rgzn-zdyfx0007） |
|
|
| Author | Institution |
| WANG Dong-lei | Institute of Chemical Materials, China Academy of Engineering Physics, Chengdu 621900, China |
| ZHANG Zhi-yu | State Key Laboratory of Mechanical Transmissions, College of Mechanical Engineering, Chongqing University, Chongqing 400044, China |
| YIN Ai-jun | State Key Laboratory of Mechanical Transmissions, College of Mechanical Engineering, Chongqing University, Chongqing 400044, China |
|
| 摘要点击次数: |
| 全文下载次数: |
| 中文摘要: |
| 目的 优化双基球扁药成球工艺参数，解决由于目前双基球扁药理论研究不充分、控制模型不明确、生产工艺参数调控依靠人工经验所导致的药品成球后直径、弧厚偏差大的问题。方法 利用 BP 神经网络在处理复杂非线性映射问题上的强大的能力，对成球关键工艺参数与成球质量指标进行建模，并应用成球工艺过程仿真数据对其进行训练，将训练得到的 BP 神经网络模型用于优化成球工艺参数。同时利用仿真数据进行检验模型的可靠性。结果 训练后 BP 神经网络均方误差为 0.001，成球直径误差率为 1.27%，成球弧厚误差率为 2.08%，成球质量参数误差均很小，可以满足工艺要求。结论 该 BP 神经网络模型具有较高精度，适用于含能材料工艺优化，提出的成球工艺优化方法能有效降低成球试制成本，缩短生产周期。 |
| 英文摘要: |
| Objective To optimize pelletization process parameters of double-based oblate spherical powder, and solve the problems of large deviation in diameter and web size of double-based oblate spherical powder after pelletization process caused by insufficient theoretical research, unclear control model and regulation of production process parameters depended on the artificial experience. Methods The ability of BP (back propagation) neural network in handling complex nonlinear mapping problems was applied to build model between pelletization process parameters and pelletization quality index, and train the model with pelletization process simulation data. The BP neural network model obtained in training was used to optimize the process parameters. At the same time, the simulation data was used to test the reliability of the model. Results After training, the mean square error of the BP neural network was 0.001, the error rate of the pelletization diameter was 1.27% with 2.08% for web size. The errors of the quality parameters were small. The technological requirements were met. Conclusion The BP neural network model has high precision and is suitable for the optimization of the energetic material process. The proposed palletization process optimization method can effectively reduce the trial cost of the palletization and shorten the production cycle. |
| 查看全文 查看/发表评论 下载PDF阅读器 |
| 关闭 |
