|
Identification
of Parametric Models for Systems and Signals
PURPOSE
The design of control systems and the
tuning of performant controllers require a good knowledge of
system input/output models, and of disturbance models.
WinPIM allows to identify parametric models for systems from
experimental data ; these data may have a very small
magnitude, thus allowing to identify plant models without
disturbing the normal operation of the plant.
Furthermore, as an open loop operation mode should often be
avoided in practice, WinPIM provides a methodology for
identification in closed loop operation. This is very useful
for improving the performances of an existing controller
(thanks to the complementary software for controller design
: WinREG).
This software was developped in cooperation with the
Laboratoire d'Automatique de Grenoble (C.N.R.S./I.N.P.G.) to
be used for systems of varied nature : robots, mechanical
systems, chemical and petro-chemical plants, process
industries (cement, paper, phosphate dryer, metallurgy),
servomechanisms, engines, thermal processes, aerospace,
semi-conductor cristal growing, detection and control of
vibrations, power systems, spectral analysis, teaching,
research,...
GENERAL DESCRIPTION
Although this software uses
computational algorithms that were already included in a
corresponding DOS software, named PIM, WinPIM has been
developped for Windows. It takes advantage of the ergonomics
of Windows' standard, which makes it easy to use, manages
printers, colors, ... WinPIM allows to include numerical and
graphical results in a report, thanks to Windows clipboard.
Contextual help using hypertext management is of course also
available.
Functionalities of WinPIM may be classified in 6 groups :
- Files Management for preprocessing of input/output data :
visualisation of data, centering, scaling, filtering...
- Estimation of model complexity, which enables to estimate
the pure time-delay and the degrees of a plant model
directly from the experimental input/output data, and
indicates the over estimation risk.
- Identification of model parameters from input/output data
file, implementing four types of discrete linear parametric
models and nine recursive identification methods.
- Validation of identified model using statistical tests to
find the best structure and the best identification method
for a given system.
- Analysis in time and frequency domains of the identified
model : step response, frequency characteristics, spectral
density, computation of models poles and zeros, and their
continuous-time equivalents.
- Simulation of system responses, with or without
disturbances, for various input signals : step, sinusoid,
square wave, PRBS, ... or even read from a file.
WinPIM belongs to WinPIM+ software package, which also
contains a software for controller design and analysis,
named WinREG.
Furthermore, this package may be completed by a module for
acquisition and digital control for Windows, which allows to
collect input/output data through an Input/Output board.
DISTRIBUTION KIT
- floppy disks containing
program, help file and example data files,
- user's manual (with theory, operating instructions and
examples).
MINIMAL HARDWARE
CONFIGURATION
PC and compatibles
Memory : minimum 1 Mo (2 Mo recommended)
1 hard disk + 1 floppy disk driver
Windows version 3.1 or more
EGA or VGA graphic screen
Mouse (highly recommended)
(Opt.) printer
(Opt.) mathematic coprocessor
|