simulation@uflorida.cis.ufl.edu (Moderator: Paul Fishwick) (09/23/88)
Volume: 5, Issue: 4, Fri Sep 23 09:22:15 EDT 1988 +----------------+ | TODAY'S TOPICS | +----------------+ (1) Neural Net Simulations for HVAC? (2) Network Simulation on the MAC? (3) Qualitative Dynamics and QP Theory (4) Re: State and Change Moderator: Paul Fishwick, Univ. of Florida Send topical mail to: simulation@uflorida.cis.ufl.edu ----------------------------------------------------------------------------- To: comp-simulation@rutgers.edu Path: phoenix!englandr From: englandr@phoenix.princeton.edu (Scott Englander) Newsgroups: comp.ai.neural-nets,comp.ai,comp.simulation Subject: Neural nets for energy management Keywords: neural, HVAC, energy management Date: 21 Sep 88 16:20:32 GMT Organization: Princeton University, NJ Does anyone know of work being done on a neural net-based energy management system for controlling heating, ventilating and air conditioning (HVAC) systems in buildings? -- - Scott ------------------------------ Return-Path: <tikal!news@beaver.cs.washington.edu> To: uw-beaver!comp-simulation@beaver.cs.washington.edu Path: tikal!sigma!uw-nsr!haynor From: uw-nsr!haynor@beaver.cs.washington.edu (David Haynor) Newsgroups: comp.simulation,comp.sys.mac Subject: Looking for Mac queuing simulation package. Date: 21 Sep 88 15:31:23 GMT Reply-To: haynor@nsr.bioeng.washington.edu (David Haynor 548-6730) Distribution: na Organization: UW-Bioengineering, Seattle, WA Is anyone aware of/does anyone have any software that enables the Mac to do a queue-oriented simulation of a network (ie multiple nodes, with a server at each node)? Please reply by mail; if there's sufficient interest, I'll post a brief summary. David Haynor Internet: haynor@nsr.bioeng.washington.edu University of Washington RC-05 UUCP: uw-nsr!haynor Seattle, Washington 98195 Dial: (206) 548-6294 ------------------------------ To: simulation@bikini.cis.ufl.edu Subject: Time and State: QP theory, and Nonlinear Dynamical Systems Reply-To: aboulanger@bbn.com Date: Thu, 22 Sep 88 15:10:37 EDT From: aboulang@WILMA.BBN.COM Sender: aboulang@WILMA.BBN.COM There are people, including myself who are interested in the relationship between QP theories and the program of qualitative dynamics started by Poincare. Elisha Sacks at Princton got started with Kuiper's QSIM then got into dynamical systems. He has done some recent work in studying QP theory in the light of dynamical systems. Gerry Sussman is also interested in this question. There is also the question of the relationship of a branch of mathematics called variational inequalities and QP theory. Some of the discreteness in the world that is capitalized by symbolic theories of process (including cognitive process) is due to the emergent properties of non-linear dynamical systems. Albert Boulanger aboulanger@bbn.com BBN Systems & Technologies Inc. ------------------------------ Date: Thu, 22 Sep 88 14:51:04 EDT From: Lindsay Patten <lindsay@watnext.waterloo.edu> To: simulation@bikini.cis.ufl.edu Subject: Re: SIMULATION DIGEST V5 N3 Newsgroups: comp.simulation In-Reply-To: <18322@uflorida.cis.ufl.EDU> Organization: U. of Waterloo, Ontario Cc: >(3) State and Change...continued I would like to add some input from the area of expertise of our lab, namely robotic vision. The point I think Paul is missing is stated in John's first posting >>No-one can acquire the initial conditions or integrate the >>equations sufficiently rapidly. It is quite feasible for a computer vision system to recognize the existance and location of a glass and even that it contains water. It is probably feasible to detect that the glass is tipping over and in what general direction. Given the limitations of tv cameras alone I am quite sure that it would be impossible to acquire proper initial conditions so as to allow application of the NS equations. It is possible that using sufficiently advanced technology one might be able to acquire images at a sufficient rate and with sufficient information content to allow the initial conditions to be derived. I am dubious that that information could be derived using computational methods using anything resembling current technology in the relevant time. Whether or not common sense methods could make use of the available information (location and approximate velocity) to allow the robot to react in some useful manner therefore seems to me to be a valid question. -- Lindsay Patten "People are package deals - No substitutions allowed" Pattern Analysis & Machine Intelligence Group lindsay@watnext Department of Systems Design Engineering lindsay@watnext.waterloo.edu University of Waterloo {utai|decvax|uunet}!watmath!watnext!lindsay ------------------------------ +--------------------------+ | END OF SIMULATION DIGEST | +--------------------------+