Skip to Main Content
 

Global Search Box

 
 
 
 

Files

File List


30350.pdf (61.61 MB)

ETD Abstract Container

Abstract Header

Hybrid Genetic Fuzzy Systems for Control of Dynamic Systems

Stockton, Nicklas O
http://rave.ohiolink.edu/etdc/view?acc_num=ucin1523635312922039

Abstract Details

2018, MS, University of Cincinnati, Engineering and Applied Science: Aerospace Engineering.
Aerospace applications are composed of many dynamic systems which are coupled, nonlinear, and difficult to control. Fuzzy logic (FL) systems provides a means by which to encode expert knowledge into a set of rules which can produce highly nonlinear control signals; this is possible because FL, like many other soft computational methods is a universal approximator. While FL systems alone excel at encapsulating expert knowledge bases, when coupled with genetic algorithms (GA), they can learn the knowledge base from evolutionary repetition. It is the goal of this work to present the efficacy of hybrid genetic fuzzy systems (GFS) in a variety of applications. This will be achieved through exploring three specific use cases. First, a variation of a benchmark problem presented at the 1990 American Control Conference is used to demonstrate the robustness of FL control as well as the utility of GAs in the learning process. The results are a controller that is far more resistant to even large changes in the plant dynamics compared to a linear controller and a process by which a class of controllers may be quickly tuned for changes to the plant system. The next problem applies the same approach to an elevator actuator for pitch control of an F-4 Phantom. This controller is tuned for a nominal case and ten subjected to the same plant with degraded aerodynamic coefficients. It is compared to a well-tuned PID controller. The effort culminates in a practical application of a FL system to guide a small unmanned aerial system (sUAS) to a precision landing on a target platform moving with uncertain velocity. This was accomplished using custom developed Python software for GFS control in conjunction with Robot Operating System (ROS) and a simulation environment called Gazebo. Heavy emphasis was placed on using only software components which can be easily implemented on popular hardware platforms. ROS was critical to meeting this goal, as well as the open source flight controller project PX4. A controller is presented which is capable of exerting the control necessary to guide the vehicle to a successful landing on both a static and moving platform.
Kelly Cohen, Ph.D. (Committee Chair)
Manish Kumar, Ph.D. (Committee Member)
George T. Black, M.S. (Committee Member)
70 p.
Engineering
genetic algorithm; fuzzy logic; dynamic system; nonlinear control; adaptive control

Recommended Citations

Citations

  • Stockton, N. O. (2018). Hybrid Genetic Fuzzy Systems for Control of Dynamic Systems [Master's thesis, University of Cincinnati]. OhioLINK Electronic Theses and Dissertations Center. http://rave.ohiolink.edu/etdc/view?acc_num=ucin1523635312922039

    APA Style (7th edition)

  • Stockton, Nicklas. Hybrid Genetic Fuzzy Systems for Control of Dynamic Systems. 2018. University of Cincinnati, Master's thesis. OhioLINK Electronic Theses and Dissertations Center, http://rave.ohiolink.edu/etdc/view?acc_num=ucin1523635312922039.

    MLA Style (8th edition)

  • Stockton, Nicklas. "Hybrid Genetic Fuzzy Systems for Control of Dynamic Systems." Master's thesis, University of Cincinnati, 2018. http://rave.ohiolink.edu/etdc/view?acc_num=ucin1523635312922039

    Chicago Manual of Style (17th edition)

ucin1523635312922039
212
© 2018, all rights reserved.
This open access ETD is published by University of Cincinnati and OhioLINK.