000021778 001__ 21778
000021778 005__ 20170622131304.0
000021778 04107 $$aeng
000021778 046__ $$k2017-06-15
000021778 100__ $$aTairidis, Georgios
000021778 24500 $$aADAPTIVE NEURO FUZZY CONTROL FOR VIBRATION SUPPRESSION OF SMART STRUCTURES WITH ENBEDED PIEZOELECTRIC SENSORS AND ACTUATORS

000021778 24630 $$n6.$$pComputational Methods in Structural Dynamics and Earhquake Engineering
000021778 260__ $$bNational Technical University of Athens, 2017
000021778 506__ $$arestricted
000021778 520__ $$2eng$$aOne of the most important problems in engineering is, among others, the vibration suppression of smart composite structures, i.e. the reduction of the oscillations, which in turn are caused by external dynamic loadings. Embedded sensors and actuators, made of piezoelectric materials, along with a suitable control system provide intelligent behavior to the smart structures. Vibration suppression can be achieved using fuzzy controllers. Fine-tuning of the involved parameters may become necessary and can be achieved either by using the trial and error method, or via global optimization, such as genetic algorithms (GAs), particle swarm optimization (PSO), differential evolution (DE) etc. In this case the results are usually very satisfactory, however, the computational cost increases. On the other hand, adaptive, optimized neuro fuzzy controllers, like the ones which are presented here, can be even more effective in the design and application of smooth robust controllers. These smart controllers are able to follow a training procedure, and thus to increase their robustness, i.e. their adaptivity under several -and some times completely different- conditions. The modeling of neuro fuzzy controllers is quite similar to other system identification techniques. The first step is the construction of the control system and subsequently a set of training data is necessary for configuration. These data will be used for the tuning of the control parameters until an error criterion is met. In other words, the controllers are first designed, based on the dynamic characteristics of the modeled structures, and then an optimization process, which is based on adaptive neural fuzzy techniques, i.e. on neural networks, is applied. Some good results in this direction have already been presented in previous investigations of the authors [1, 2, 3]. An important feature of these controllers lies to the fact that they can achieve significant suppression of the oscillations without knowledge of the full state-space (measurements) of the problem and without the application of any classical optimization method. Recent developments and applications mostly on linear smart structures are presented here. More specifically, several plate models with different support scenarios and loadings are modeled and studied, in order to prove the robustness and the efficiency of the control.

000021778 540__ $$aText je chráněný podle autorského zákona č. 121/2000 Sb.
000021778 653__ $$aSmart structures, Control, Neuro fuzzy control, piezoelectricity.

000021778 7112_ $$aCOMPDYN 2017 - 6th International Thematic Conference$$cRhodes Island (GR)$$d2017-06-15 / 2017-06-17$$gCOMPDYN2017
000021778 720__ $$aTairidis, Georgios$$iStavroulakis, Georgios
000021778 8560_ $$ffischerc@itam.cas.cz
000021778 8564_ $$s684432$$uhttps://invenio.itam.cas.cz/record/21778/files/17898.pdf$$yOriginal version of the author's contribution as presented on CD, section: [RS13] Optimum design and control in structural dynamics and earthquake engineering
.
000021778 962__ $$r21500
000021778 980__ $$aPAPER