(Sept 2020 - Present)
(July 2018 - June 2020)
(July 2012 - June 2016)
For the expertise in Chirp-based EIS & Machine Learning
For the expertise in Electrochemical systems & Li-ion Batteries
For the expertise in Signal & Systems
Picture a patient in a hospital. To determine their health, doctors don't simply assign a percentage or a number. They monitor the patient's symptoms, conduct tests, and study medical reports - all to observe the patient's 'behavioral' changes that could indicate underlying health issues. Now, consider a lithium-ion (Li-ion) battery as that patient.
In our technology-driven world, we're more dependent than ever on devices powered by Li-ion batteries, from our smartphones an d laptops to electric vehicles and renewable energy storage systems. However, just like humans, batteries also have a 'health' that needs to be monitored. Over time, due to cell degradation, growth of the solid electrolyte interface (SEI) layer, lithium plating, and other factors, a battery's performance changes, signaling an alteration in its health status. This detoriated health lead to various faults and undesirable incidents of combustion within the battery.
But, how can we observe these 'behavioral' changes in a battery? How can we predict when a battery is about to 'fall ill' and perhaps even take preventive measures?
Enter Electrochemical Impedance Spectroscopy (EIS). EIS is like the medical reports for our battery 'patient.' It's a powerful tool that gives us insight into the processes occurring inside a battery. However, classical
EIS techniques, while informative, are not without challenges - they are time-consuming, complex, and not suitable for real-time health monitoring.
This is where my research comes into the picture. I aim to revolutionize the way we monitor and predict the 'health' and 'well-being' of Li-ion batteries, making them more reliable and efficient for our everyday use. By improving our ability to prognose possible cell degradation and faults in batteries, we can make batteries more safe from unwarranted fires, enhance their lifespan, cut down waste, and contribute to a more sustainable future.
The focus of first objective is the design and development of a novel Chirp-Based Electrochemical Impedance Spectroscopy (EIS) method. The chirp-based EIS is well defined and established for the linear-time-invariant (LTI) systems. It is shown that chirp-based EIS can construct nyquist plot(EIS) within 2-3 seconds. However, efficacy and applicability of chirp-based EIS on nonlinear systems is not yet studied.
Recognizing the need for rapid and real-time applications, an essential aspect of this research is significantly reducing the time required to construct an EIS. Initial studies have been conducted on a nonlinear liquid tank system, which served as an ideal preliminary model to investigate the efficacy of the chirp-based EIS technique. The results obtained from these simulations are providing valuable insights that will be crucial in shaping the application of this technique to inherently Li-ion batteries.
Once the chirp-based EIS is developed, intention is to validate this method using real-world Li-ion batteries. In the research process, the chirp-based EIS technique will be critically examined for its effectiveness in capturing essential battery behavior across various states of charge (SoC) and different operating temperatures. This validation will be achieved through extensive simulations using PyBaMM, an open-source battery modeling software.
After successful simulation, the subsequent phase of validation involves cross-validating the simulation results with experimental observations. An experimental setup, comprising a function generator, signal analyzer, and potentiostat, has been designed for this purpose. This setup facilitates the real-world application of both the Chirp-based and conventional EIS techniques on a Li-ion battery, and the data obtained is subsequently utilized for comparison and validation.
The cross-validation between the simulation and experimental results provides a robust framework for future work. It not only enhances the reliability of the EIS plots generated from the PyBaMM model but also paves the way for the subsequent development of a machine learning model in the next objective of the research.
Moving into the realm of predictive analysis, third objective revolves around creating a machine learning (ML) algorithm informed by EIS data. The goal here is to design an efficient machine learning algorithm that can effectively parse and interpret EIS data to recognize and predict faults associated with battery cell degradation.
In this phase, the PyBaMM model, refined with the addition of cell degradation mechanisms such as Li-plating, SEI layer growth, and cracking, will generate intricate EIS datasets. The objective is to create a realistic simulation of various fault characteristics observed in real-world battery systems.
Harnessing these datasets, an image-based pattern recognition approach will be implemented for the development of the machine learning model. This model will learn to associate specific patterns within the EIS data with different states of battery health and degradation, thereby enabling precise battery performance predictions.
The final objective of the research is the integration of the developed chirp-based EIS and machine learning algorithm into an online tool for real-time battery health monitoring and fault prognosis. By translating the results obtained from the simulations and the machine learning model into a user-friendly interface, the tool aims to provide immediate, actionable insights about the health of Li-ion batteries. This online prognostic tool is expected to pave the way for proactive battery health management, preventing potential failures, enhancing the lifespan of batteries, and contributing to efficient and sustainable energy use.
ME181403
Materials Science
Level UG
Department of Mechanical Engineering
January 2023 - May 2023
January 2022 - May 2022
ME181502
Machine Design - I
Level UG
Department of Mechanical Engineering
July 2022 - December 2022
Dept. of Chemical Engg,
IIT Guwahati, 781039
+91-8830626164
created with
Static Website Generators .