Autonomous Battery Management System (AutoBMS)

可充电电池是一种很好的能量存储形式。特别是，锂基电池已广泛应用于电动汽车、便携式电子设备、家用电器、电动工具、航空航天设备和可再生能源存储系统。电池管理系统（BMS）由电池电量计、电池平衡电路和最佳充电算法组成，对于电池组的安全、可靠和高效运行至关重要。BMS使用来自电池的三个非侵入性测量值（电压、电流和温度）来估计荷电状态（SOC）和健康状态（SOH）;这些估计值用于BMS功能，例如生成最佳充电波形、电池平衡以及激活安全保护器。 目前的BMS技术不足以准确预测电池的SOH;因此，选择要么过早更换电池，要么等到故障事件发生。这两种选择都有不良后果：过早更换将导致最终用户的成本增加和对环境的过度浪费;等待将对最终用户的安全和体验质量产生负面影响。此外，现有技术的BMS受限于其所表征的电池的特定化学、制造商和尺寸，即，目前的BMS不是通用的;这限制了电池的选择并导致成本增加;而且，这种限制性的BMS不允许人们重新使用旧的/新的电池组。此外，定制电池充电器会产生过多的电子垃圾和环境浪费。 拟议的研究有两个直接目标。第一个是发现一个独特的测量指标来准确估计SOH;为此，我们将使用机器学习算法来研究数千个观察结果，以识别作为SOH准确指标的简洁特征。第二个目标是开发通用BMS的必要算法基础，该BMS独立于电池的化学成分，制造商，尺寸和年龄;我们将利用云计算和信息融合算法的力量来实现这一目标。这项研究的一些成果将有助于改善最佳的电池充电算法，以减少充电时间，而不影响SOH。这项研究的长期目标是开发一种自主BMS，为最终用户提供效率，灵活性和安全性，并使他们能够以独特的创造性方式使用可充电电池来存储和使用可再生能源。

Rechargeable batteries are an excellent form of energy storage. Particularly, Lithium based batteries have been widely adopted in electric vehicles, portable electronic equipment, household appliances, power tools, aerospace equipment and renewable energy storage systems. A battery management system (BMS), consisting of a battery fuel gauge, cell balancing circuitry, and optimal charging algorithm, is essential for the safe, reliable and efficient operation of a battery pack. The BMS uses three non-invasive measurements from the battery, voltage, current and temperature, to estimate the state of charge (SOC) and state of health (SOH); these estimates are used in BMS functions, such as the generation of optimal charging waveforms, cell balancing, and to activate safety protectors. Today's BMS technology is inadequate to accurately predict the SOH of a battery; as a result, the choices are either to prematurely replace the battery or to wait until a failure event occurs. Both of these choices have undesirable consequences: premature replacement will result in increased cost to the end user and excessive waste to the environment; waiting out will negatively impact the safety and quality of experience of the end user. Further, the state of the art BMS is constrained to particular chemistry, manufacturer, and size of the battery to which it is characterized for, i.e., the present-day BMS is not universal; this restricts battery selection and results in increased cost; also, such a restrictive BMS doesn't allow one to repurpose old/new battery packs. In addition, custom battery chargers generate excessive electronic clutter and environmental waste. The proposed research has two immediate goals. The first one is to discover a unique measurement index to accurately estimate SOH; for this, we will employ machine learning algorithms to study thousands of observations to identify succinct features that are accurate indicators of SOH. The second goal is to develop the necessary algorithmic foundations of a universal BMS that is independent of the chemical composition, manufacturer, size, and age of the battery; we will make use of the power of cloud computing and information fusion algorithms to achieve this goal. Some outcome of this research will help to improve optimal battery charging algorithms to reduce charging time without affecting SOH. The long-term objective of this research is to develop an autonomous BMS that provides the end user with efficiency, flexibility, and safety and enables them to use rechargeable batteries in uniquely creative ways to store and use renewable energy.