...
  • 主题
高级搜索  >
历史搜索 清空历史
首页> 外文期刊>ASHRAE Transactions >Gray-Box Modeling of Multistage Direct-Expansion Units to Enable Control System Optimization
【24h】

Gray-Box Modeling of Multistage Direct-Expansion Units to Enable Control System Optimization

机译：多级直接扩展单元的灰箱建模，以实现控制系统优化

获取原文
获取原文并翻译 | 示例
           
页面导航
  • 摘要
  • 著录项
  • 相似文献
  • 相关主题

摘要

Gray-box or black-box models that are trained using on-site data typically do a better job of capturing actual system performance than forward models that are based on physical parameters due to deviations from design assumptions or uncertainties. However, accurate, site-specific models are needed for developing optimal control strategies for existing systems. In this study, two different gray-box, quasi-steady-state modeling approaches for multistage direct-expansion (DX) units with variable air volume (VAV) were developed and compared for application to supervisory control optimization. Firstly, the ASHRAE Toolkit model was modified for multistage DX units and trained using data from a field site. Secondly, a component-based, gray-box modeling approach was developed and trained using on-site data. The models were validated using measured data not included in the training data set. The advantage of the component-based approach is that it requires less data for training and provides better extrapolating performance. However, it requires significantly more computation. Therefore, a metamodel that correlates outputs from the gray-box model was developed for application to supervisory control optimization. This overall approach provides good accuracy over a wide range of conditions with limited training data and computational requirements. In order to test the application potential, optimization of supply air temperature setpoints was performed for the case study using both the metamodel and modified Toolkit model. Although the two different modeling approaches gave some differences, both indicate significant energy savings potential.
机译：与基于物理参数的正向模型（由于偏离设计假设或不确定性）相比，使用现场数据训练的灰盒或黑盒模型通常在捕获实际系统性能方面做得更好。但是，需要精确的，针对特定地点的模型才能为现有系统开发最佳控制策略。在这项研究中，针对可变风量（VAV）的多级直接膨胀（DX）单元，开发了两种不同的灰箱准稳态建模方法，并将其进行比较，以应用于监督控制优化。首先，针对多级DX单位修改了ASHRAE Toolkit模型，并使用了现场数据进行了训练。其次，开发了基于组件的灰盒建模方法，并使用现场数据进行了培训。使用未包含在训练数据集中的测量数据来验证模型。基于组件的方法的优势在于，它需要较少的数据进行培训，并提供更好的推断性能。但是，它需要更多的计算。因此，开发了一个与灰箱模型的输出相关的元模型，以应用于监督控制优化。在有限的训练数据和计算需求的情况下，这种整体方法可提供良好的精度。为了测试应用潜力，使用元模型和修改后的Toolkit模型对案例研究进行了送风温度设定点的优化。尽管两种不同的建模方法存在一些差异，但两者都显示出巨大的节能潜力。

著录项

  • 来源
    《ASHRAE Transactions》 |2015年第2期|203-216|共14页
  • 作者

    Jie Cai; James E. Braun;

  • 作者单位

    School of Mechanical Engineering, Purdue University, Lafayette, IN;

    School of Mechanical Engineering, Purdue University, Lafayette, IN;

  • 收录信息
  • 原文格式 PDF
  • 正文语种 eng
  • 中图分类
  • 关键词

相似文献

  • 外文文献
  • 中文文献
  • 专利
获取原文

客服邮箱：kefu@zhangqiaokeyan.com

京公网安备：11010802029741号 ICP备案号：京ICP备15016152号-6 六维联合信息科技 (北京) 有限公司©版权所有

  • 客服微信

  • 服务号