0_Exams
To schedule a prelim exam or PhD defense, you need to submit a plan of work on the system so that the plan can be reviewed by the graduate school and approved by the committee. You cannot simply schedule a date for the prelim or defense by yourself without completing the process. The graduate school must be informed before the exam happens. Otherwise, the exam might be considered as an invalid one.
So for all students who plan to schedule a prelim/phd defense, you will need to do the following
1) please do check it out with previous phd students
2) make sure you meet the submission deadlines
3) hit the button to submit the information before the deadline
4) Inform me that the paperwork is all set a week before the exam
The grad school is relaxing its preference that PhD exams be held in person and will permit telcon defenses to be held.
This will still require submission of the remote defense form:
For PHD the link is: https://grad.ncsu.edu/wp-content/uploads/2015/11/sched-mas-oral-remote-exam.pdf.
For Master the link is: https://grad.ncsu.edu/faculty-and-staff/forms/graduate-school-forms/
Please check all guidelines at the Power Electronics and Power System Website.
New rules: There may not be a conditional pass. Instead you get a "fail". You can take it again though.
Need to take it within 3 semesters your receive your master degree
Select two out of the three courses: power electronics (taught by Dr. Bhattacharya), power system generation operation and control (taught by Dr. Baran), Control (taught by Dr. Chow).
Reproduce results from 1-2 papers and write a review paper (maximum 10-page)
Prepare a 25 minutes presentation
Selection of papers:
Need to have theoretical components (optimization algorithms, advanced modeling and control, large-scale simulation, data analytics, etc.)
Need to talk with me at the beginning of the semester so we agree that you work on reproducing the paper
You will meet with me every two weeks to report me your progress
You will prepare your results as short presentations (5-minute)
Presentation should be prepared as follows:
Outline
Background (What is the landscape of this research area you picked? why is this topic important?)
Literature review (What are the papers in the area and why do you pick this 1-2 papers to reproduce out of the papers you reviewed?)
Method 1:
Method 2:
Method 3:
Why you reproduced method 1?
Approach (What are the theory behind the algorithms/models/methods?)
modeling methods
assumptions
problem formulation
constraints
Simulation Results (Demonstrate that your can reproduce the results in the selected publication)
Side by side results (results in the paper, results you reproduced)
Discussions (What can be reproduced and what cannot be? If you achieve a better result, why, if you got a less promising result, why?)
Conclusion/Summary
When answering a question, make sure you fully understand what is behind the question.
Ask yourself, when you reproduce the algorithm, why you pick this paper? (wrong answers: because it is simple and I can do it... or because my adviser asked me to do it... or because my current project requires me to use this method... Appropriate answers: Because this algorithm has the following advantages over the other approaches: X, Y, Z. This modeling method matches measurement data well and has least modeling errors when I tried it on new data sets... etc) We want to see that you have independent thinking and want to explore suitable options for your future research.
Ask yourself, when I have reproduce the results, are the results match 100% or there are differences? Why?
If you did more than just reproducing the results, point out what you have done to make it perform better.
At NC State, we want oug. Find your passion.
Google the top tr PHD students to become independent thinkers and explorers of new ideas and new areas, so don't rely on your advisers to point you out to publications. Find one you are interested in reproducinen papers in the area (Literature review)
Compare their approaches and find 3~4 papers that you are interested in and will be useful for your future research.
Draft 2 slides to summarize your results of the literature review)
Come to my office and let's have a discussion on why your pick the 2-3 papers to reproduce their results.
We will access to what extent you can reproduce the results, whether or not the results is useful for your research in the future, how much you can add to the results, etc. My suggests are only a reference for you to consider.
Take this exam seriously as this is a great opportunity to work with your adviser, learn how to write a paper and how to present an idea.
Preparation for the courses review
Usually we will ask questions related with your course work and the paper you presented to make sure
This student has solid background in power systems, control or power electronics.
This student has sufficient research capability to be successful in the future.
This student show good analytical capabilities to react to challenging questions and think broadly instead of focusing on only what has been taught.
So we will ask you questions related with both your course work and your research and see how well you react to it.
My observation is that professors turn to go easy on the students in recent years. As a result, a lot of students turn to try their lucks at the exam. In my opinion, it is important to review all your course works and try to get a hold of the big pictures to be ready to enter the second year of your PHD study. You should have learnt a lot of from all those relevant courses and you want to ask you what are the most common used concepts in this field - make sure you understand them thoroughly.
Don't try to get away from it by making your presentations exceptionally lengthy. As a PhD student, you want to face instead of dodging challenges.
Best of luck in passing the exams with a flying color!
Prelim Exam (send out a prelim report (it is a draft of your thesis) to faculty 2 weeks before the exam date, prepare a 50-minute presentation)
New rules: Graduate school will enforce passing pre-lim within 4-year
Examples (This is a private link and it is only open to students in my group)
You need to find committee members following the department guidelines
Three professors from the ECE department
1 professor from outside the department
can have 1 from industry
You need to go to the graduate school and set up the date for the exam
No other students are allowed to attend the prelim exam unless the committee approves the attendance of other students
Only committee members can ask questions so if other students are invited, they should not ask any questions.
After the exam, students need to leave the room and let the committee have a discussion on the exam.
Need to reserve the time at lease 1~2 months ahead of time
You will need to present the background, literature survey, draft outline of your thesis, the planned theoretical contributions, the path forward for completing your thesis.
Thesis Outline
Thesis outline and Research plan: Work done so far and Work to be done in the next 6 months
Which chapters have been completed/published
which chapter(s) will be finished and published in the next few months - from now to your planned graduation date.
Better show how your research works link together and what challenges you want to tackle in your field.
Background
Need to introduce the landscape
Configuration of the overall system
Literature review: state-of-the-art, challenges, proposed approach and the contributions
Method/Algorithm 1
Conclusions
Method/Algorithm 2
Conclusions
Future work
A list of publications and planned publications
Backup slides (put all your models and equations and papers in the same directory. In case a question is asked regarding the data, approach, methods, equations, you can quickly find a reference to answer those questions)
Final PhD Oral Defense Exam
Examples (This is a private link and it is only open to students in my group)
Scheduling the exam a few month ahead of time is always prudent.
The defense exam can be attended by students. Questions from students are asked at the end of the exam.
Your presentation is 50 minutes to an hour 1 hour. When Q&A is included, the whole defense can last for 2.5 hours.
Please try to skip the materials you have presented in the Prelim exam.
You will need to present the background, literature survey, draft outline of your thesis, the planned theoretical contributions, the path forward for completing your thesis.
Outline
Background
Briefly review the pre-lim presented material
New work after prelim
Literature review: state-of-the-art, challenges, proposed approach and the contributions
Motivation: what has not been done? Why you want to do it?
Contributions: What is the novelty? What is the expect improvements?
******************************
Method/Algorithm 1
Assumptions (a lot of students tend to forget this)
Introduce the model/concept/basics/definitions
Performance criterion
Mathematical formulation
Results
********************************
Method/Algorithm 2
Assumptions
Introduce the model/concept/basics/definitions
Mathematical formulation
Performance criterion
Results
Conclusions
Future work
A list of publications
Backup slides (put all your models and equations and papers in the same directory. In case a quesiton is asked regarding the data, approach, methods, equations, you can quickly find a reference to answer those questions)
You want to make sure you set up the defense dates with professors, reserve a conference room, send reminders, and fix graduate school documentations.
You want to dress up, arrive 20-30 minutes ahead of time to the conference room. You need to load the slides to the computer, start the conference call (if there are remote participants), print out a hard copy for the committee members, and try to see if the pointer works.
During the presentation, allow people to ask their questions and do not interrupt them when they are asking questions. Answer when they finish asking the questions.
Bring a notepad to write down the questions.
When preparing for the presentation
Focus on introducing the big picture: the importance of the work, what is the state-of-the-art, what is missing or insufficient, what you have done will make what kind of difference?
Don't read from slides. Especially don't read variables and equations. Focus on describing what the equation describes and how the formulation is different from others? Highlight and explain the differences and what makes your approach better.
Don't apologize (e.g. don't say that I am sorry, or opps, or etc) when you made a mistake. Carry on with the presentation.