| WIKI | FAQ | Tech FAQ | |
 http://csua.com/feed/ | |
| ||||||
| 5/25 |
| 2009/9/23-10/5 [Academia/GradSchool] UID:53395 Activity:moderate |
9/23 I'm in grad school part time, and the professor I was trying to get
to advise me just sent me "What you suggest (remote, part-time, topic
formulation) doesn't fit my advising style..." Does anyone know
what "topic formulation" might mean in this context? I come with
funding from work, but the topics I can choose are somewhat limited.
Could be a reference to that...
\_ whatever it means, "doesn't fit my advising style" nullifies
the probability of having him/her as your advisor. Most advisors
hate remote (communication problems), most hate part-time
(full-timers just have more entropy), and most want to control
the topic. You should consider finding another advisor who is
more desperate for students and is willing to get someone
onboard for free. IMHO, I've seen a lot of remote & part-timers
and it's just difficult for both parties, and I can understand
why they don't want a remote & part-timer grad student even
for free.
\_ To be honest, I think it's academic snobbery. A coworker of
mine applied to and got accepted to both USC and UCLA for PhD.
When UCLA found out he wanted to be part-time while working
they sent him vibes similar to the above. USC didn't care.
The guy is brilliant and an excellent student who finished a CS
master's at the same time as his Aerospace Engineering PhD. UCLA
turned their nose up at him just because he didn't fit their
narrow definition of what they thought they wanted to deal
with. Screw that. They need to be more aware of who is paying
the bills, which is another reason why universities need less
state funding: students are an afterthought.
\_ What the hell is a part-time PhD? 12 year plan? Most PhDs
take 5-6+ years working 40-60 hrs/wk.
\_ Part-time PhD means taking a part-time load of coursework
and passing the exams. This takes maybe 2-3 years.
After that, you work on your research on a part-time
basis. The people I know who did a PhD while working
ended up working about 30 hours/week and spending about 30
hours/week on school. This is no big deal for a good
student. There are full-time students who screw around
and never make any progress. I remember one who went
full-time for 6 years and still wasn't done. That doesn't
mean you can't get it done while wasting less time. It
helps a lot, too, if some of your day job contributes
to your research.
\_ Point me to a C.V. of someone who earned a "part time"
PhD. 30 hrs/wk? Average PhD is 5-6 years @ 60 hrs/wk,
so sounds like a 10-12 year PhD. urlP = #f
\_ I know at least 5 people who earned their PhD
this way and it didn't take 10 years. Most of
them already had an MS before doing the PhD, though.
This is one of them:
http://tinyurl.com/yceq86t
I have no idea where his C.V. is online, but you
can see that it took him about 6 years based on
this update in 2002:
link:tinyurl.com/yccjl9m
"After graduation, I spent a year at the Lockheed
Martin Skunk Works doing structural dynamics
and flutter testing on the Joint Strike Fighter. In
November 2000, I joined NASA-JPL in the Navigation &
Mission Design section. This fall I am still at JPL
but am concurrently enrolled part time at Cal Tech as
a Ph.D. student in Mechanical Engineering."
He earned the PhD in 2008:
http://www.cds.caltech.edu/~murray/wiki/Main_Page
You sound as narrow-minded as the UCLA profs.
Luckily, Caltech is more than happy to educate
people who are smart and want to be educated.
\_ 6 years for a part-time PhD after 7 years of
full-time BS/MS. Seems about right to me.
(Source: http://linkedin.com)
\_ That's exactly right. This is the reason why you can do
part time MS in Stanford and Cornell and other nice schools
but good luck doing part time PhD in schools like Stanford
and Cornell. People laugh at the MS program so you can
get by a lot like MS, but many professors have a lot of
snobbery in that they think you should do full time research.
In their mind, if you do part time PhD, you're not really
serious about doing meaningful work. Now whether that is
true or not, I don't know, but that's just how it is
in schools like Stanford and Cornell. P.S. USC should
not be compared with Stanford and Cornell. Completely
different leagues.
\_ I think some of the most famous professors are those that already
have a lot of funding so don't care about having a freebie. They
want a grad student who is young and can slave away at the ideas
that the professor started in order to gain even more recognition
and funding. In academia, there isn't a shortage of hungry,
driven, young guys who will do whatever their advisors tell them
to do, so that's why schools like MIT and Caltech are full of
profs with this type of attitude. In the end, it's like a free
market. There is a lot of supply of cheap laborers but not enough
professors so they can afford to be total assholes.
\_ This guy is actually at the other end of the spectrum. He has
no students. He works pretty much exclusively on his own
little project that no one else cares about. But he has tenure,
so he doesn't have to do anything else. -op
\_ ah, tenure, that's the problem. Based on my experiences with
tenured profs, I'd say the majority of them burned out their
candles years ago and don't really give a damn today. Yes
there are a few good ones but they're rare. May I ask
which prof rejected the proposal and why are you requesting
an advisor who likes to work on things that no one else
cares about?
\_ I don't think I want to identify the professor. The motd
does show up google. He's not a prof at Cal.
I wanted to work with him because he has a good reputation
as an adviser. I knew student who was advised by him and
I understand that he demands good work, but does not
abusively move the goal posts like many profs do. He
is also one of the only profs at the school that works
in my research are.
in my research area.
\_ what area is that? |
| 5/25 |
|
| tinyurl.com/yceq86t -> etd.caltech.edu/etd/available/etd-05072008-131735/ Caltech Student Instructions Carson III, John Maurice (2008-04-17) Robust model predictive control with areactive safety mode. edu/CaltechETD:etd-05072008-131735 Title Robust model predictive control with a reactive safety mode Degree PhD Option Mechanical Engineering Advisory Committee Advisor Name Title Richard M Murray Committee Chair Behcet Acikmese Committee Member Douglas G MacMynowski Committee Member Joel W Burdick Committee Member Keywords * Linear Matrix Inequalities * Robust Control * Nonlinear Systems * Uncertain Systems * Model Predictive Control * Receding Horizon Control * Safety Mode Date of Defense 2008-04-17 Availability unrestricted Abstract Control algorithms suitable for online implementation in engineering applications, such as aerospace and mechanical vehicles, often require adherence to physical state and control constraints. Additionally, the chosen algorithms must provide robustness to uncertainty affecting both the system dynamics and the constraints. As further autonomy is built into these systems, the algorithms must be capable of blending multiple operational modes without violating the intrinsic constraints. Further, for real-time applications, the implemented control algorithms must be computationally efficient and reliable. The research in this thesis approaches these application needs by building upon the framework of MPC (Model Predictive Control). The MPC algorithm makes use of a nominal dynamics model to predict and optimize the response of a system under the application of a feedforward control policy, which is computed online in a finite-horizon optimization problem. The MPC algorithm is quite general and can be applied to linear and nonlinear systems and include explicit state and control constraints. The finite-horizon optimization is advantageous given the finite online computational capabilities in practical applications. Further, recursively re-solving the finite-horizon optimization in a compressing- or receding-horizon manner provides a form of closed-loop control that updates the feedforward control policy by setting the nominal state at re-solve to the current actual state. However, uncertainty between the nominal model and the actual system dynamics, along with constraint uncertainty can cause feasibility, and hence, robustness issues with the traditional MPC algorithm. In this thesis, an R-MPC (Robust and re-solvable MPC) algorithm is developed for uncertain nonlinear systems to address uncertainty affecting the dynamics. The R-MPC control policy consists of two components: the feedforward component that is solved online as in traditional MPC; and a separate feedback component that is determined offline, based on a characterization of the uncertainty between the nominal model and actual system. The addition of the feedback policy generates an invariant tube that ensures the actual system trajectories remain in the proximity of the nominal feedforward trajectory for all time. Further, this tube provides a means to theoretically guarantee continued feasibility and thus re-solvability of the R-MPC algorithm, both of which are required to guarantee asymptotic stability. To address uncertainty affecting the state constraints, an SR-MPC (Safety-mode augmented R-MPC) algorithm is developed that blends a reactive safety mode with the R-MPC algorithm for uncertain nonlinear systems. The SR-MPC algorithm has two separate operational modes: standard mode implements a modified version of the R-MPC algorithm to ensure asymptotic convergence to the origin; safety mode, if activated, guarantees containment within an invariant set about a safety reference for all time. The standard mode modifies the R-MPC algorithm with a special constraint to ensure safety-mode availability at any time. The safety-mode control is provided by an offline designed control policy that can be activated at any time during standard mode. The separate, reactive safety mode provides robustness to unexpected state-constraint changes; eg, other vehicles crossing/stopping in the feasible path, or unexpected ground proximity in landing scenarios. Explicit design methods are provided for implementation of the R-MPC and SR-MPC algorithms on a class of systems with uncertain nonlinear terms that have norm-bounded derivatives. Further, a discrete SR-MPC algorithm is developed that is more broadly applicable to real engineering systems. The discrete algorithm is formulated as a second-order cone program that can be solved online in a computationally efficient manner by using interior-point algorithms, which provide convergence guarantees in finite time to a prescribed level of accuracy. This discrete SR-MPC algorithm is demonstrated in simulation of a spacecraft descent toward a small asteroid where there is an uncertain gravity model, as well as errors in the expected surface altitude. Further, realistic effects such as control-input uncertainty, sensor noise, and unknown disturbances are included to further demonstrate the applicability of the discrete SR-MPC algorithm in a realistic implementation. |
| www.cds.caltech.edu/~murray/wiki/Main_Page IST Note: I am on sabbatical from Caltech until Fall 2010. During this time, I will not be hosting visitors, taking on new postdocs or sponsoring new undergraduate projects (including SURF). I will be continuing to advise my existing students and participate in ongoing projects. If you are a Caltech student and would like to meet with me, I'll be returning to campus a few days every month or two. Search Research My group's research is in the application of feedback and control to networked systems, with applications in biology and autonomy. Current projects include verification and validation of distributed embedded systems, analysis of insect flight control systems, and biological circuit design. Visitors Teaching The list below is the courses that I have taught at Caltech (or at least most of them). The course links will take you to the (current) course homepage, where you can find the syllabus, handouts, and homework sets. The links for specific terms take you to the course homepage for that term. |