Hello, my name is Gregory Platt and it's my pleasure to welcome you to this course on battery state of charge estimation. This course is the third in a specialization that investigates the proper management and control of battery packs. Usually, battery packs comprising many battery cells. A significant challenge for a battery management system algorithms is that they must estimate quantities that describe the condition of a battery pack, but which are not readily measurable. The big picture for battery management system algorithms is shown in the illustration on this slide. It describes the critical functionality of a battery management system from the perspective of the algorithm design engineer. When the battery powered load is connected to the battery pack or the battery management system is turned on, the algorithms must all initialize their parameter values and their state values. Then the algorithms enter a loop which repeats periodically while the battery powered load is active and connected to the battery pack. The first step in this loop is to measure cell voltages and pack current and module temperatures. A subsequent step is to use these measured values to update estimates of this state of charge of all of the cells and the battery pack. An additional step is to update the estimates of the states of health of all the cells in the battery pack, and furthermore, the cells must be balanced and finally we must compute power limits for the battery pack. The repetition rate for this algorithm loop is part of the design process of creating a battery management system. The typical rates are between one and 50 times per second. When the battery powered load is disconnected from the battery pack and the battery management system is turned off, the algorithm store their status information in non-volatile memory for use the next time the battery pack is activated. Considering this entire programmatic algorithm loop, this course focuses specifically on methods to estimate state of charge. The next course in this specialization will focus on methods to estimate state of health and the final course in this specialization will focus on how to balance cells in a battery pack and how to compute power limits. In this course, you will learn some preparatory definitions that are required to be able to talk consistently and precisely about the problem of state of charge estimation and we will also cover some background math concepts that will be used throughout the course. You will learn some simple methods to estimate battery state of charge and how these methods can fail because they lack robustness. A better estimate method is based on a technology that is called Kalman Filtering. So, you will learn how to derive the steps of the linear Kalman Filter. Because this technology is probably unfamiliar to you at this point, we are also going to spend some time discussing how to think about the steps executed by this Kalman Filter and how to implement those steps in active code and how to evaluate the results from that Kalman Filter execution. Now, it turns out that the standard Kalman Filter works only for linear systems and our batteries are non-linear systems, but we can generalize the Kalman Filter to work with non-linear systems as well. So, the next step is for you to learn how to derive and implement one variety of non-linear Kalman Filter that is known as the extended Kalman Filter. The extended filter can work very well in some situations, but it's not the best non-linear generalization of the standard Kalman Filter. So, you will also learn how to derive and implement something that is known as a Sigma-Point Kalman Filter. It's also sometimes known as an Unscented Kalman Filter. Finally, you will also learn how to solve some real world problems using some side information that is provided automatically to you by the Kalman Filter. After completing this course, you will be able to implement simple voltage and current based state of charge estimators and you will understand their limitations. You will be able to explain the purpose for every sequential probabilistic inference step and this is the basis for the implementation and the derivation of the Kalman Filter. You will be able to execute octave code that I provide to you for a linear Kalman Filter and you will know how to evaluate the results that it computes. You will also be able to evaluate and execute octave code that I provide to you for state of charge estimation using an extended Kalman Filter and you will use this on laboratory test data to implement a state of charge estimator and you will be able to evaluate those results as well. Additionally, you will be able to execute octave code that I provide to you for state of charge estimation using a Sigma-Point Kalman Filter and you will implement that on lab test data and evaluate those results. Finally, if you decide to pursue the honors section of this course you will learn how to implement methods to detect and discard faulty voltage sensor measurements and you also learn methods to implement the non-linear Kalman Filter is a very efficiently for multi-cell battery packs. This course is the third course in our specialization on algorithms for battery management systems. The first two courses in this specialization are prerequisite to this course. You must complete those courses before attempting this one. Now, you should remember that the first course was titled, "Introduction to Battery Management Systems." The prerequisite knowledge from that course that we will depend on in this course is a set of basic background understanding in battery management system requirements and in technologies for sensing voltage and current and temperature and so forth. The second course in the specialization is titled, "Equivalent Circuit Cell Model Simulation." The prerequisite knowledge from that course that we depend on very, very heavily comprises a detailed specific background in the enhanced self-correcting circuit model that describes the dynamics of lithium ion battery cells. The required background includes knowing how to use the specific octave toolbox of various codes that I have shared with you earlier of helper routines that work with this cell model. You will also depend on a background in probability theory from other courses in your undergraduate curriculum and especially a background in what is called a Gaussian random variable or sometimes a normally distributed random variable. As an optional resource, you may be interested in the book, "Battery Management Systems Volume 2" and "Equivalent-Circuit Methods fromArtech House." This course will be based on chapter three from that textbook. Again, the book is not required to succeed in this course, but I believe that it could be valuable to you especially for your future study to have this optional resource. That concludes this introductory lesson, and again, welcome to the course. I trust that you will learn some very valuable and helpful skills for implementing state of charge estimators for battery management systems.