Analysis of Li-Ion Battery Characteristics in a Wearable Patient Monitoring Device Using Equivalent Circuit Model

Abstract

Hospital care is rapidly transforming with the proliferation of Wearable Patient Monitoring devices. These "plug and play" devices are very useful in situations where remote monitoring is required. They are usually battery operated, which render them portable and rechargeable. The rechargeable devices are cost effective and environment friendly. However, the rechargeable batteries which are used in medical devices exhibit non-linear characteristics. The non-linear behaviour is typically due to the chemical reactions which takes place within the cell, and it makes it hard for the designer to predict the battery behaviour. In order to design efficient monitoring systems, it is important to understand the nature of the battery in combination with the power needs of the device. As a solution battery models are implemented which give the designer and developer a tool to gauge battery behaviour in various circumstances. This thesis focuses on sensor devices and their power source i.e. sensor batteries. The rechargeable battery pack consists of three lithium-ion cells that are modeled to predict the performance of the device. Experimental test data is collected using the Cadex C8000 battery analyzer test setup. Data collected from battery testing is used to validate a modified Thevenin's Equivalent Circuit model. The experimental test data and simulated output from the battery model are compared and evaluated to create an optimized model which can estimate the performance of a battery pack.