CAREER: Feedback Control of Micro-Fluidic Systems and the Bio-Chemical Particles Inside Them

职业：微流体系统及其内部生化颗粒的反馈控制

• 批准号: 0348251
• 负责人: Benjamin Shapiro
• 金额: $ 40万
• 依托单位: University of Maryland, College Park
• 依托单位国家: 美国
• 项目类别: Standard Grant
• 财政年份: 2004
• 资助国家: 美国
• 起止时间: 2004-02-01 至 2009-01-31
• 项目状态: 已结题
• 来源: https://www.nsf.gov/awardsearch/showAward?AWD_ID=0348251&HistoricalAwards=false
• 关键词: CAREER; Feedback; Control; Micro; Fluidic

Intellectual MeritResearch will focus on feedback control of bio-molecules and of liquid packets inside micro-fluidicsystems. The goal is to control the path and shape of micro-fluidic packets, and the trajectories andchemical reactions of bio/chem particles inside the packets. This will facilitate new micro-fluidic systemssuch as miniaturized drug delivery systems, and it will allow existing systems to function in noisy realworld conditions. Specific control tasks will include: steering of many particles at once for targetedcollisions between different cells, viruses, and bacteria; precision moving, splitting, and joining ofdroplets by electrically induced surface tension forces; and shape control of individual particles. For example, it is known that fluid flow can straighten DNA chains. Our grand challenge is to create a flow field that only unwraps a small portion of the DNA chain and makes a specific protein hit the start of that unwrapped section. Feedback control requires the integration of devices, sensing, control algorithms, and actuation. Such system integration raises fundamental research challenges. We will address four key areas that are open and which match our core expertise in fluid dynamics and control. 1) Real time sensor processing: Infer the position, type, shape, and properties of fluid packets and bio particles from the sensor data in real time. For example, we will infer the shape of cells down to tens of nanometers resolution by measuring the surrounding fluid flow using PIV (particle image velocimetry) and by efficiently solving an inverse inflow-to-shapel mathematics problem. 2) Control algorithm design: Based on the sensor data, compute the appropriate actuator response. For observed particle positions, find the electrode voltages to move the particles in the desired directions. 3) Control implementation: Our controllers must function in real time. This raises significant computational issues including controller reduction and state estimation from limited sensing. 4) Modeling: All three steps above rely on a quantifiable understanding of the systems at hand.These four steps will be implemented on micro fluidic systems in our lab and on systems in the labs ofour collaborators. The research will focus primarily on topic two: control algorithm design.Broader ImpactThis proposal aims to unite research from different disciplines. The outreach plan reflects this aim:1) Micro fluidics design competition supported by the Hinman undergraduate entrepreneurship CEOprogram: Each team in the competition will consist of students from engineering, physics, chemistry,biology, and the business school. Undergraduate teams will design and fabricate micro systems aftertaking pre-requisite micro fabrication courses. Successful teams will transition their ideas intobusiness plans through the CEO program (the program provides undergraduates with the toolsrequired to start and manage a business, see www.hinmanceos.umd.edu). In collaboration with theWomen In Engineering (WIE) program, the competition will be used as a recruiting tool to attractwomen and under-represented minorities to science, systems research, and entrepreneurship.2) Strong focus on multi-disciplinary undergraduate research: Undergraduate students will be involvedin creating the experiments, developing the control algorithms, and testing the devices, and they willundertake internships at the companies and government labs with which my group collaborates.3) Integrating the languages of control and micro/nano: The controls and micro/nano community speakdifferent technical languages. For example, existing micro fluidic models are not suitable for controldesign. I will chair a micro fluidic ilmodeling versus designli workshop at the next AIAA AerospaceSciences Meeting and Exhibit which will bring together researchers from these two communities.Future workshops will be organized at micro-systems and control conference. These workshops willfocus on translating physical micro-systems control challenges into tractable control theory questions.

智力MeritResearch将专注于生物分子和微流体系统内液体包的反馈控制。其目标是控制微流体包的路径和形状，以及包内生物/化学颗粒的轨迹和化学反应。这将促进新的微流体系统，如小型化药物输送系统，它将允许现有的系统在嘈杂的现实世界条件下发挥作用。具体的控制任务将包括：一次操纵许多粒子，用于不同细胞、病毒和细菌之间的目标碰撞;通过电诱导表面张力精确移动、分裂和连接液滴;以及单个粒子的形状控制。例如，已知流体流动可以拉直DNA链。我们最大的挑战是创造一个流场，只解开DNA链的一小部分，并使一种特定的蛋白质击中解开部分的开始。 反馈控制需要设备、传感、控制算法和致动的集成。这种系统集成提出了基础研究的挑战。我们将解决四个关键领域是开放的，这符合我们在流体动力学和控制的核心专业知识。 1)真实的时间传感器处理：从真实的时间传感器数据推断流体包和生物颗粒的位置、类型、形状和属性。例如，我们将通过使用PIV（粒子图像测速仪）测量周围的流体流动并有效地解决逆流动到形状数学问题来推断细胞的形状，分辨率可达数十纳米。 2)控制算法设计：基于传感器数据，计算适当的执行器响应。对于观察到的粒子位置，找到使粒子沿所需方向移动的电极电压。 3)控制实现：我们的控制器必须在真实的时间内运行。这引起了重大的计算问题，包括控制器减少和状态估计有限的传感。4)建模：以上三个步骤都依赖于对现有系统的量化理解，这四个步骤将在我们实验室的微流体系统和我们合作者实验室的系统上实现。研究将主要集中在主题二：控制算法设计。更广泛的影响这个建议旨在联合不同学科的研究。该推广计划反映了这一目标：1）由Hinman本科创业CEO计划支持的微流体设计竞赛：参赛的每个团队将由来自工程，物理，化学，生物和商学院的学生组成。本科团队将在完成微制造课程后设计和制造微系统。成功的团队将通过CEO项目（该项目为本科生提供创业和管理企业所需的工具，见www.hinmanceos.umd.edu）将他们的想法转化为商业计划。在与妇女在工程（WIE）计划合作，竞争将被用作招聘工具，以吸引两名女性和代表性不足的少数民族的科学，系统研究，和奖学金。2）强烈关注多学科的本科研究：本科生将参与创建实验，开发控制算法，并测试设备，他们将在与我的团队合作的公司和政府实验室实习。3）集成控制和微/纳语言：控制和微/纳社区使用不同的技术语言。例如，现有的微流体模型不适合于控制设计。我将在下一届AIAA航空航天科学会议和展览会上主持一个微流控建模与设计研讨会，届时这两个团体的研究人员将聚集在一起。未来的研讨会将在微系统和控制会议上组织。这些研讨会将侧重于将物理微系统控制的挑战转化为易于处理的控制理论问题。