Cyberphysical Smart Cities Infrastructures

4Analysis of Different Regression Techniques for Battery Capacity Prediction

Param Khakhar¹ and Rahul Kumar Dubey²

¹Indian Institute of Technology Delhi, Department of Computer Science and Engineering, Delhi, India

²Robert Bosch Engineering and Business Solutions Ltd., Bengaluru, India

4.1 Introduction

Lithium‐ion batteries are used increasingly nowadays in a wide variety of applications. This can be attributed to the fact that lithium is the lightest of all the metals, has great electrochemical potential, and therefore provides the largest energy density for weight, which is twice that of the traditional nickel–cadmium battery. The maintenance for the Li‐ion battery is low as compared with its other counterparts. However, Li‐ion battery has disadvantages as well. They are structurally more fragile than their other counterparts and are prone to aging. Consequently, the batter would have reduced battery capacity. After a certain number of charge and discharge cycles, the battery would have to be replaced. Battery capacity is a measure of the amount of energy stored in the battery, and it can have the units such as ampere‐hour, milliampere‐hour, etc. It depends on several factors such as cycling, ambient temperature, battery chemistry, application, maintenance, etc. This chapter used machine learning algorithms to model the relation of battery capacity with various charge and discharge cycles. The widespread use of Li‐ion batteries over many different scales makes the end‐of‐life ...