Networked Battery System Management and Control for Active Diagnosis, Observability and Resilient Operation

电池是交通运输和电力行业实现清洁、高效和可持续能源发展的最关键技术之一。先进的汽车和电网电池系统由不同类型和特性的联网电池模块组成。系统级性能可能因子系统异常而中断或完全禁用。除了针对单个电池模块的典型电池管理系统（BMS）之外，该项目还将开发一种新型弹性电池管理系统（RBMS），用于网络电池系统的可靠运行。该RBMS将能够检测和定位异常，防止其传播，并重新配置电池系统以维持运行。这项关键的变革性技术将支持电动汽车延长电池寿命，在有限的能源资源下维持军事任务，并保持智能电网不中断。除了在电池特性、控制和管理方面的直接技术进步外，该项目还将对教育、社会和经济产生重大影响。它还将有助于推动这些储能和许多相关应用领域的技术和劳动力发展。在这个项目中，该团队将把研究结果整合到韦恩州立大学现有的电力驱动车辆工程本科和研究生课程的电池技术课程中。将为学生、工程师和技术人员开发先进电池管理系统的课程和培训研讨会，以加强车辆电气化、可持续能源发展和智能电网方面的劳动力培训。该项目的成功将有助于实现更清洁、更高效、更可靠的交通运输、电网等。电池异常情况包括特征变量漂移到不可接受区域所指示的潜在健康恶化、必须通过及时补救措施清除的故障以及不可恢复的故障。网络电池系统带来了独特而艰巨的挑战，包括快速故障诊断和定位的传感限制，快速准确地联合估计充电状态和表征参数，以及用于系统重构的先进电力电子设备。目前的电池系统及其管理不包括这些功能。采用离散事件系统、系统识别和状态估计、随机分析和先进电力电子控制的方法，该项目将引入一个新的RBMS框架，用于网络电池系统的集成控制、主动诊断和持续运行。该团队将与电池技术、电动汽车和智能电网领域的行业领导者合作，开发一种新的主动网络观察者理论、实时主动诊断和异常状态定位、重新配置和自适应BMS策略。提出的方法代表了一种革命性的技术，用于管理不同类型的电池和结构，包括新的或旧的电池，用于车辆和电网应用。本课题的主要研究方法具有以下特点：(1)基于离散事件系统（DES）的诊断策略和控制决策。采用系统的DES策略，可以实现异常状态的快速诊断和定位；(2)实时电池特性，实现模块级诊断。联合估计电池的状态和参数，实时表征电池；(3)主动系统重构以适应弹性操作和自适应电池管理系统。硬件重构可以防止局部异常扩散到电池系统的其他部分。电池管理系统然后自我调整以维持在新配置下的运行。

Batteries are one of the most critical enabling technologies for achieving clean, efficient and sustainable energy development in the transportation and power sectors. Advanced vehicle and power grid battery systems consist of networked battery modules of diverse types and characteristics. System-level performance can be interrupted or completely disabled by subsystem abnormality. Beyond typical battery management systems (BMS) for individual battery modules, this project will develop a novel resilient battery management system (RBMS) for reliable operations of networked battery systems. This RBMS will be capable of detecting and locating abnormality, prevent them from propagating, and of reconfiguring the battery system for sustained operation. This critical and transformative technology will support electric vehicles for extended battery life, sustain military missions under limited energy resources, and maintain smart electricity grids without interruption. Beyond its direct technological advance in battery characterization, control and management, this project will have significant educational, societal, and economic impact. It will also help advance technology and workforce development in these energy storage and the many associated fields of application. Within this project, the team will integrate findings into battery technology curricula in the existing Electrical-Drive Vehicle Engineering programs at the undergraduate and graduate levels at Wayne State University. Courses and training seminars on advanced battery management systems will be developed for students, engineers, and technicians to enhance workforce training in vehicle electrification, sustainable energy development, and smart grids. The success of this project will help achieve cleaner, more efficient, and more reliable transportation, power grids, and beyond.Battery abnormality conditions include potential health deterioration indicated by drifting of characterizing variables into unacceptable regions, faults that must be cleared by timely remedy actions, and failures that are not recoverable. Networked battery systems introduce unique and daunting challenges, including sensing limitations for prompt fault diagnosis and localization, fast and accurate joint estimation of state of charge and characterizing parameters, and advanced power electronics for system reconfiguration. Current battery systems and their management do not include such functions. Employing methodologies from discrete event systems, system identification and state estimation, stochastic analysis, and advanced power electronics control, this project will introduce a new RBMS framework for integrated control, active diagnosis, and sustained operation of networked battery systems. Collaborating with industry leaders in battery technology, electric vehicles, and smart grids, the team will develop a new theory of active network observers, real-time active diagnosis and localization of abnormal conditions, reconfiguration, and adaptive BMS strategies. The proposed methods represent a transformative technology for managing different types of batteries and structures, including new or old batteries, for vehicle and grid applications. The main approaches of this project have some distinctive and novel features: (1) Discrete event system (DES)-based diagnosis strategies and control decisions. By using systematic DES strategies, fast diagnosis and localization of abnormal conditions can be achieved; (2) Real-time battery characterization for module-level diagnosis. A battery's states and parameters will be jointly estimated for characterizing batteries in real time; and (3) Active system reconfiguration to accommodate resilient operation and adaptive battery management systems. Hardware reconfiguration will prevent local abnormality from spreading to other parts of battery systems. The battery management systems then self-adjust to sustain operation under the new configuration.