Fault-tolerant control of hydraulically-powered actuators using fractional-order PID schemes

dc.contributor.authorMaddahi, Ali
dc.contributor.examiningcommitteeBalakrishnan, Subramaniam (Mechanical Engineering) Kinsner, Witold (Electrical and Computer Engineering) Zhao, Qing (Electrical and Computer Engineering, University of Alberta)en_US
dc.contributor.supervisorSepehri, Nariman (Mechanical Engineering)en_US
dc.date.accessioned2019-03-06T21:47:42Z
dc.date.available2019-03-06T21:47:42Z
dc.date.issued2019-02-27en_US
dc.date.submitted2019-02-28T03:14:52Zen
dc.degree.disciplineMechanical Engineeringen_US
dc.degree.levelDoctor of Philosophy (Ph.D.)en_US
dc.description.abstractHydraulic actuators are used extensively in various equipment including aircraft, robots and heavy-duty machinery. One of the common problems in hydraulic systems is occurrence of faults such as the actuator internal leakage. Some faults need to be detected properly to prevent actuator malfunctioning. For example, the internal leakage deteriorates the system performance as it causes the fluid to move between cylinder chambers. Model-based fault detection techniques require a model of the hydraulic actuator and faults or an estimation of system parameters. In practice, there is no accurate model to describe some faults. Application of data-driven fault detection techniques is more desirable. One of the main challenges, due to the faults, in hydraulic actuation systems is the controller design. If a controller is not able to compensate for fault effects, it may lead to considerable losses. Therefore, there is a need for designing fault-tolerant controllers (FTCs). Data-driven techniques of controller design are good solutions to overcome challenges associated with obtaining models or estimating system and fault parameters. In this thesis, the applications of data-driven techniques of the fault detection and controller design to the hydraulic actuators are investigated and exemplified with detection of the internal leakage fault and design of a control system tolerant to the leakage. First, a set of multiscale measures is quantified in various healthy and faulty operating modes and a comparison is carried out, between the measures, to identify the most reliable indicator(s) in detecting various levels of the leakage. Comparison results indicate that the wavelet transform is the best technique, amongst the employed multiscale measures, for the internal leakage detection. Next, a fractional-order PID (FOPID) position controller is designed using experimental data. Efficacy of the designed controller in tracking various reference inputs, in presence of different system uncertainties, is examined through experiments. Finally, a methodology is developed for designing an FOPID-based FTC that does not require a prior knowledge about the model and parameters of the system and fault or emulation of the fault in experimental setup. The methodology is based on introduction of a set of synthetic errors to the hydraulic actuator. Experimental results prove that the methodology works well for a hydraulic actuator experiencing the internal leakage.en_US
dc.description.noteMay 2019en_US
dc.identifier.urihttp://hdl.handle.net/1993/33781
dc.language.isoengen_US
dc.rightsopen accessen_US
dc.subjectFault-tolerant controlen_US
dc.subjectRobust controlen_US
dc.subjectFractional-order PID controlleren_US
dc.subjectMultiscale analysisen_US
dc.subjectIterative feedback tuningen_US
dc.subjectHydraulic actuatorsen_US
dc.subjectFault detectionen_US
dc.subjectWavelet transformen_US
dc.subjectFractal dimensionsen_US
dc.subjectInternal leakage detectionen_US
dc.titleFault-tolerant control of hydraulically-powered actuators using fractional-order PID schemesen_US
dc.typedoctoral thesisen_US
Files
Original bundle
Now showing 1 - 1 of 1
Loading...
Thumbnail Image
Name:
Maddahi_Ali.pdf
Size:
4.61 MB
Format:
Adobe Portable Document Format
Description:
License bundle
Now showing 1 - 1 of 1
Loading...
Thumbnail Image
Name:
license.txt
Size:
2.2 KB
Format:
Item-specific license agreed to upon submission
Description: