1. Home
  2. Embedded Programming
  3. Review My Book
  4. Page 2

Review My Book

I don't have a good book on optimal control, and I should get one. I have a good book on robust control which I suspect comes at the subject from a very different direction from normal, it is "Robust Control, The Parametric Approach" by Bhattacharyya, Chapellat & Keel.

It's heavy on theory, but they spend a lot of time extending frequency- domain approaches to formal robust control theory, and I've spent a lot of time using frequency-domain approaches in my work. I've done projects with it propped up next to my computer screen and had good success.

Beware of "optimal control" that doesn't mention robustness, though -- you'll get "optimal control" of a nonexistant plant, and lousy control of what you actually have (which is the whole point of robust control theory, by the by).

All in all, though, if you take a big pot, brown the traditional control methods on all sides and simmer well with a couple of quarts of experience, you'll make a better soup than you can get by heating up a can of optimal robust control fresh from the supermarket.

When I was about to start my master's degree program I inquired about getting a PhD in control theory (this was at WPI, which is good but small). I was sent to the mathematics department. While I was a grad student there I was a teaching assistant (twice!) for a 3rd-year control systems class that was taught by a biomechanical engineer and a mechanical engineer, through the ME department.

Most recently, as part of a contract I helped to hire a woman who got her bachelor's degree in economics, during which she took her first course in control theory through _that_ department, then emigrated to the United States and finished up with a PhD in mathematics -- specializing in control theory -- here in the states.

So control theory is a discipline that finds a lot of homes. If endocrinologists can't take a control theory class as part of medical school they should be able to, and if I found out that there was a school of archeology out there that required controls to graduate I would be amused, but not amazed.

--
My liberal friends think I'm a conservative kook.
My conservative friends think I'm a liberal kook.
Why am I not happy that they have found common ground?

Tim Wescott, Communications, Control, Circuits & Software
http://www.wescottdesign.com
Reply to
Tim Wescott
Loading thread data ...

Its interesting what kind of classes you can take. In the last year of my bachelor EE study I wasn't happy with the packages of classes they put together so I went to the head of the faculty and asked if I could put my own package together. Much to my surprise his answer was 'yes, ofcourse' and he pulled up a book with all the classes I could take for graduation. I was tempted by 'playing the piano' but I choose 'compression algorithms' which was actually part of the IT department.

--
Failure does not prove something is impossible, failure simply
indicates you are not using the right tools...
nico@nctdevpuntnl (punt=.)
--------------------------------------------------------------
Reply to
Nico Coesel

If their ME is closer to Aerospace engineering it would make a lot of sense- there's a lot of related math involved in that. If it's regular knuckle-dragging ME, then maybe not so much.

Reply to
Spehro Pefhany

I've always thought that's because negative feedback is natural and pretty universal, so sorting out how to model and manage it applies pretty broadly.

Eric Jacobsen Anchor Hill Communications

formatting link

Reply to
Eric Jacobsen

that

understand

topic

both

it

I have it. I wish it had a bit more theory. Better coverage of the z-transform in particular would help; but there are whole textbooks on that larger then Tim's book.

I guess i should get into it again and what i have to say.

?-)

Reply to
josephkk

me)

that

All too often, reviews in serious technical publications are by one = cranky old person. =20

?-)

Reply to
josephkk

You got something against cranky old people?!?!?!

Eric Jacobsen Anchor Hill Communications

formatting link

Reply to
Eric Jacobsen

these kids nowadays. they've never seen a 1/2 watt resistor. have they ever smelled the solder resin?

i have, in my life, soldered 2-watt resistors, vacuum tube sockets (both PC-board and chassis mounted), buss and wire harnesses, point-to-point wiring, etc. i've even gotten a taste (or a touch) of tube B+ voltage (300 and once 700 volts). but i haven't done that for maybe over 4 decades.

maybe 3 decades ago was using a brake to bend sheet aluminum into a chassis/box and riveting it together. drilling holes and mounting brackets (that one also makes with the brake) and standoffs and thru-hole PC boards. painting and lettering (i used ruboff decals you could buy then and then acrylic lamenated over it).

what do *they* know about cranky old people?

--

r b-j                  rbj@audioimagination.com

"Imagination is more important than knowledge."
Reply to
robert bristow-johnson

Rivets? You had rivets? Talk about being privileged! We had to use old sheet metal screws, that we salvaged from junk car radios!

Reply to
Michael A. Terrell

yah. they were those cheapo blind-rivets where you vice-grip the pieces together, drill the hole, stick un-expanded rivet in with some fancy tool, pull lever and it pulls a nail-like thing back (expanding the rivet behind the hole) and severs the rest of the nail shaft flush with the outside of the rivet. like this tool:

formatting link

this technology existed in the late 70s. perhaps earlier, i dunno.

i also didn't have to walk to school 5 miles in the snow barefoot.

but the truth is that my eldest brother and sister one year younger (4 years older than me) *did* go to a one-room school house in the country for their first 2 and 1 years, respectively. (in 1959, they started going to a new "modern" school in town, where i began elementary in 1961.)

eewww! i wanted to make a solid box that i could paint over the fasteners except those holding on the bottom plate.

if i could've, i would have wanted to spot-weld the chassis together. can't do it real easy with aluminum. building Heathkits, i saw some steel chassis that were bent and spot-welded and then galvanized or dipped in something. built like a brick shit-house.

i think the bottom plate went on with simple self-tapping sheet metal screws. had to be careful not to strip the aluminum chassis.

--

r b-j                  rbj@audioimagination.com

"Imagination is more important than knowledge."
Reply to
robert bristow-johnson
[Snipped by Lyons]

Hi Tim, I completely understand the frustration you had with only one review, and a negative review at that, for your book on Amazon.com.

It looks like you now have some good reviews on Amazon. I'm happy to see that.

I've learned to be suspicious of book reviews on Amazon. Here's why. Years ago I wanted to learn about the digital modulation technique of OFDM. I found an OFDM book on Amazon that had very good reviews, "OFDM Wireless LANs: A Theoretical and Practical Guide."

That book was the worst example of technical writing I have *EVER* seen! I cannot tell you how many technical errors I found in that book. There were, I estimate, two typos for every three pages in the book. I went to the Sams Publishing website to download an errata for the book. Get this,... the first three entries in the errata contained two typographical errors! How discouraging. They should have published an errata to the errata. (I also downloaded the MATLAB code which, woefully, crashed on line 19 of the very first routine.)

Being ticked off because I wasted $35, I went back and looked more closely at the "glowing" reviews of that book on Amazon. I soon became convinced that most of those reviews are fraudulent. I very carefully, and in great detail, chronicled my powerful evidence for suspecting reviewer fraud.

Tim, I monitor Amazon's signal processing book reviews pretty carefully. I'd estimate that a technical book takes 9 -to- 12 months to receive a half dozen reviews. (You can see this by looking at the dates of the reviews of various books.) That OFDM book received seven reviews within a two week period! Now what in the heck is the chance of that happening in a valid way? I'll tell ya', it's zero. In addition, five of those seven reviews contained the word "excellent." What are the chances of that happening legitimately?

It's my opinion that someone who'll profit from increased book sales wrote those misleading reviews.

Anyway, I posted my opinion in the form of a review of that book on Amazon. My scathing review lasted one week on Amazon, and then suspiciously disappeared forever. So that's why I'm much more careful now in believing book reviews on Amazon. By the way Tim, I thought you were very brave in submitting your original post requesting reviews of your book.

[-Rick-] PS. I don't have a copy of your book, so I'm not able to post a review of it on Amazon.com.
Reply to
Rick Lyons

Actually, they make it. . Crazy things going on out there. Not to disagree with your general point, though.

Mel.

Reply to
Mel Wilson

It seems to me that we are "improvising ourselves" at various tasks all the time. Otherwise Jerry would not have learned so much!

Read a little, simulate or model a little ......

I can but think of the embedded programmers who have worked for me. Not control systems guys. Now what if they had a situation that fit a "control system" context? Would I give them Tim's book or hire a new engineer? Obviously, "it depends" but I'd first be wanting to give them the book.

Fred

Reply to
Fred Marshall

"Pop" rivets were used at least as far back as WWII to stop metal airplanes falling apart.

Can be set from just one side, useful for fitting, say, wing panels.

Original patent was held by a company (British?), called Tucker.

"Nail" down the center designed to force itself up inside the hollow rivet, expanding it, then break off, leaving just the head inside the rivet. Made in aluminum, steel, monel, copper.

"Genuine" (aerospace qualified) pop rivets ain't exactly cheapo.

There were a few lookalikes, designed to get round the Tucker patent by wrinkling, rather than expanding the rivet. They were crap.

--
"For a successful technology, reality must take precedence 
over public relations, for nature cannot be fooled."
                                       (Richard Feynman)
Reply to
Fred Abse

Of course the ideal solution is to give them my book, then if that fails bring me on board.

:)

--
Tim Wescott
Control system and signal processing consulting
www.wescottdesign.com
Reply to
Tim Wescott

I was urged, by a good friend who is impeccably honest when marketing isn't involved, to get on and write my own review for the book. I'm actually surprised that more publishers don't hire shills to do just that.

I dunno. It's kind of hard to mail someone a pipe bomb over USENET.

--
Tim Wescott
Control system and signal processing consulting
www.wescottdesign.com
Reply to
Tim Wescott

That's weird, because nearly all of the Ph.D. MEs I work with are almost completely ignorant of controls.

To make matters worse, their primary software developer only knows PID loops.

--
_____________________
Mr.CRC
crobcBOGUS@REMOVETHISsbcglobal.net
SuSE 10.3 Linux 2.6.22.17
Reply to
Mr.CRC

Control systems is very much an optional course for ME.

And you can go a long way with loops that -- in the end -- can be described as "just" PID. But if all you know is one flavor, and how to tune them up by the seat of your pants -- yup, that can get pretty limiting.

--
Tim Wescott
Control system and signal processing consulting
www.wescottdesign.com
Reply to
Tim Wescott

[Snipped by Lyons]

Hi Dennis, Ha ha. That's another reason why I like you Aussies!

[-Rick-]
Reply to
Rick Lyons

[Snipped by Lyons]

Shame shame shame on that guy!!! Tim, ...I suggest you do NOT turn your back on that dishonest bastard.

[-Rick-]
Reply to
Rick Lyons

ElectronDepot website is not affiliated with any of the manufacturers or service providers discussed here. All logos and trade names are the property of their respective owners.