CAREER: Identifiability Optimization in Electrochemical Battery Systems

这个学院早期职业发展（Career）项目奖是为了检查电化学电池模型参数的可识别性，定义为一个人可以快速、独特、准确地从实验数据中确定这些参数的程度。这项研究将建立在PI之前在电池参数识别方面的研究基础上。第一个目标是探索量化可识别性的不同度量。该研究还将开发算法，用于从实验数据中估计锂离子电池的内部状态变量（例如，给定电池满或空的程度）和参数（例如，电池健康或不健康的程度）。总之，这两个贡献将作为可识别性优化的基础。在可识别性优化中需要获取的关键信息包括：安装在锂离子电池上的传感器类型，以提高其健康退化的检测速度和准确性，以及能够以最小的时间和成本获得最多关于电池模型参数信息的实验室循环实验类型。这项研究将对多个社会利益相关者产生更广泛的影响。整个社会将受益于电池参数可识别性的提高所带来的在线电池诊断的改进。电池实验学家将能够从更短、更便宜的实验室测试中获得更准确的参数。科学界将受益于弥合电化学、在线估计和最优控制文献之间目前差距的努力。最后，这项工作将集中在推广和教育活动上，包括：(i)通过宾夕法尼亚州立大学的施赖尔荣誉学院（Schreyer Honors College）吸引本科生，尤其是女性和代表性不足的少数族裔；（ii）在电池系统动力学和控制领域开设网络课程；（iii）通过ARPA-E等机构资助的补充性研究工作，让工业伙伴参与进来；（iv）为K-12 STEM推广创建以电池为重点的教育材料。

This Faculty Early Career Development (CAREER) Program award is to examine the identifiability of electrochemical battery model parameters, defined as the degree to which one can quickly, uniquely, and accurately determine these parameters from experimental data. The research will build on previous research by the PI in battery parameter identification. The first goal is to explore different metrics for quantifying identifiability. The research will also develop algorithms for estimating both the internal state variables (e.g., degree to which a given battery is full or empty) and parameters (e.g., degree to which a battery is healthy or unhealthy) of lithium-ion batteries from experimental data. Together, these two contributions will serve as foundations for identifiability optimization. Critical information to be acquired in identifiability optimization include: the type of of sensors to be install on a lithium-ion battery to improve the detection speed and accuracy of its health degradation and the type of laboratory cycling experiments that can yield the greatest amount of information about battery model parameters with minimal time and cost.The research will have broader impact on multiple societal stakeholders. Society at large will benefit from improvements in online battery diagnostics enabled by improvements in battery parameter identifiability. Battery experimentalists will be able to obtain more accurate parameters from shorter, less expensive laboratory tests. The scientific community will benefit from the efforts to bridge current gaps among electrochemistry, online estimation, and optimal control literature. Finally, the effort will focus on outreach and education activities including: (i) the engagement of undergraduate students, especially women and underrepresented minorities through Penn State University's Schreyer Honors College; (ii) the creation of web-based courses in the areas of battery system dynamics and control; (iii) the engagement of industrial partners through complementary research efforts funded by agencies such as ARPA-E; and (iv) the creation of battery-focused educational materials for K-12 STEM outreach.