Adaptive Control of Unavoidable Hazardous Releases

不可避免的危险排放的适应性控制

• 批准号: 9220663
• 负责人: Nikolaos Katopodes
• 金额: $ 26万
• 依托单位: Regents of the University of Michigan - Ann Arbor
• 依托单位国家: 美国
• 项目类别: Standard Grant
• 财政年份: 1993
• 资助国家: 美国
• 起止时间: 1993-05-15 至 1996-12-31
• 项目状态: 已结题
• 来源: https://www.nsf.gov/awardsearch/showAward?AWD_ID=9220663&HistoricalAwards=false
• 关键词: Adaptive; Control; Unavoidable; Hazardous; Releases

The purpose of this project is to develop a computational procedure for the adaptive control of hazardous releases, either of water or toxic substances. In particular, the control strategy is aimed at deliberate releases from man-made storage or process facilities that are deemed unavoidable because alternative actions may lead to more severe catastrophic results. The procedure consists of combining two-dimensional transient models that simulate the dynamics of the release and its subsequent fate in the environment and an adaptive control scheme, which monitors the current state of the released matter and the ambient conditions. Real-time feedback is sent to the control mechanism, which in turn develops an optimum release strategy that minimizes a predetermined performance index. The simulation models that are employed for the direct solution of the release spreading have been repeatedly verified for their accuracy and reliability and standard optimization techniques have shown that control of water and toxic releases is possible. The novelty of the proposed research lies in an alternative formulation of the control problem that avoids iterative simulations of the direct problem, which, due to lengthy computations, hinder the real-time implementation of the procedure. In contrast, by solution of the adjoint equation for a variety of evolution problems described by an enlarging and deforming domain, it is possible to optimize a given performance index without repeated simulations. The result is an efficient and practical procedure for unavoidable hazardous releases that can minimize their environmental impact.

该项目的目的是开发一种计算程序，用于自适应控制水或有毒物质的有害排放。在 特殊的,控制策略的目的是故意从人为储存或处理设施中释放被认为是不可避免的，因为替代行动可能导致更严重的灾难性后果。该程序包括将模拟释放动力学及其在环境中的后续命运的二维瞬态模型与自适应控制方案相结合，它监测释放物质的当前状态和环境条件。实时反馈被发送到控制机制，控制机制进而开发出一个最佳的释放策略，使预定的性能指数最小化。直接求解释放扩散问题所采用的仿真模型的准确性和可靠性得到了反复验证，标准优化技术也证明了这一点控制水和有毒物质的释放是可能的。所提出的研究的新颖之处在于控制问题的替代公式，避免了直接问题的迭代模拟，由于冗长的计算，妨碍程序的实时执行。相反，通过求解由扩大和变形域描述的各种演化问题的伴随方程，可以在不重复模拟的情况下优化给定的性能指标。结果是一个有效和实用的程序，不可避免的危险释放，可以尽量减少其对环境的影响。