Format

Send to

Choose Destination
Neural Netw. 2017 Sep;93:45-56. doi: 10.1016/j.neunet.2017.04.013. Epub 2017 May 5.

Hybrid neuro-heuristic methodology for simulation and control of dynamic systems over time interval.

Author information

1
Institute of Mathematics, Silesian University of Technology, Kaszubska 23, 44-100 Gliwice, Poland. Electronic address: Marcin.Wozniak@polsl.pl.
2
Institute of Mathematics, Silesian University of Technology, Kaszubska 23, 44-100 Gliwice, Poland. Electronic address: Dawid.Polap@polsl.pl.

Abstract

Simulation and positioning are very important aspects of computer aided engineering. To process these two, we can apply traditional methods or intelligent techniques. The difference between them is in the way they process information. In the first case, to simulate an object in a particular state of action, we need to perform an entire process to read values of parameters. It is not very convenient for objects for which simulation takes a long time, i.e. when mathematical calculations are complicated. In the second case, an intelligent solution can efficiently help on devoted way of simulation, which enables us to simulate the object only in a situation that is necessary for a development process. We would like to present research results on developed intelligent simulation and control model of electric drive engine vehicle. For a dedicated simulation method based on intelligent computation, where evolutionary strategy is simulating the states of the dynamic model, an intelligent system based on devoted neural network is introduced to control co-working modules while motion is in time interval. Presented experimental results show implemented solution in situation when a vehicle transports things over area with many obstacles, what provokes sudden changes in stability that may lead to destruction of load. Therefore, applied neural network controller prevents the load from destruction by positioning characteristics like pressure, acceleration, and stiffness voltage to absorb the adverse changes of the ground.

KEYWORDS:

Decision support; Dynamic systems; Heuristic methods; Neural networks; Simulation and control

PMID:
28544890
DOI:
10.1016/j.neunet.2017.04.013
[Indexed for MEDLINE]

Supplemental Content

Full text links

Icon for Elsevier Science
Loading ...
Support Center