Doctor of Philosophy
Electrical and Computer Engineering
This work deals with the modeling and control of automated drilling operations. Advances in drilling automation are of substantial importance because improvements in drilling control algorithms will result in more efficient drilling, which is beneficial from both economic and environmental points of view. While the primary application of the results is extraction of natural resources, potentially there exists a wide range of applications, including offshore exploration, archaeological research, and automated extraterrestrial mining, where implementation of new methods and control algorithms for drilling processes can bring substantial benefits.
The main contribution of the thesis is development of new methods and algorithms for control of drilling processes in industrial drilling systems, ensuring stability and high performance characteristics. The problems of regulation of vertical penetration rate and drilling power in rotary drilling systems are solved; as a result, stability and vibration mitigation is ensured. A number of challenges is addressed, such as complexity and nonlinearity of the drilling model, lack of information about environment and parameters of the drilling system itself, and poor communication between downhole sensors and ground-level equipment. Several cases are considered, depending on the amount of information that is available in advance or in real time. Two mathematical models of the drilling system are investigated: one is finite-dimensional, and another is a distributed parameter model. Several solutions are proposed for both of them, using methods of adaptive, robust, and sliding mode control, and comparisons are made. Feasibility and efficiency of the proposed control algorithms are confirmed by simulations in MATLAB/Simulink.
Summary for Lay Audience
Nowadays, extraction of natural resources is essential for generation of energy and raw materials. Drilling is an important part of this process, because these resources often lie deep underground. Automation of drilling processes (i.e., making a drilling system work with minimum human supervision and intervention) makes a positive impact from economic and environmental points of view, because it may improve safety of operations, reduce the number of breaks, failures, and wearouts, as well as make operators' work easier and more effective.
In this thesis, we propose new methods for automatic control of drilling processes, assuming that our knowledge about characteristics of the material that we are drilling is extremely limited, and measurements from the bottom end are often unavailable. Utilizing a number of mathematical models that describe a conventional drilling system with acceptable precision, we design several control algorithms that generate control signal (in our case, electrical voltage) in real time. This control signal, when applied to the motor on the ground level, successfully brings the whole system into a certain desired stable mode of operation.
Results presented in this thesis correspond to gradual relaxation of simplifying assumptions, and complexity of control schemes described in different chapters depends on how much we know about the drilling system and environment where it is operating. Effectiveness and feasibility of the results are confirmed by simulation using the software package MATLAB.
Faronov, Maksim, "Nonlinear Adaptive Control of Drilling Processes" (2023). Electronic Thesis and Dissertation Repository. 9474.