Richard C. Hill

Modular Supervisory Control with Equivalence-Based Abstraction and Covering-Based Conflict Resolution

A modular approach to control is one way to reduce the complexity of supervisory controller design for discrete-event systems (DES). A problem, however, is that modular supervisors can conflict with one another. This paper proposes requirements on coordinating filters that will resolve this conflict. Abstractions are employed to reduce the complexity of the filter construction. Our specific approach is unique in that it employs a conflict-equivalent abstraction that offers the potential for greater reduction in model size than those abstractions employed in previous works on conflict resolution. The resulting control implemented by the modular supervisors in conjunction with coordinating filters meeting the proposed requirements is shown to be safe and nonblocking. Approaches for constructing these filters are discussed and a methodology that implements deterministic coordinating filter control laws by nondeterministic automata is presented. The covering-based filter law construction methodology presented here is further demonstrated to provide less restrictive control than existing results on state-feedback supervisory control.

A methodology for modular model-building in discrete automation

Our objective is to develop a general and versatile approach for building structured formal models of complex automated systems in order to facilitate their control and diagnosis. For this purpose, we present a methodology that builds the complete model of a system by composing models of the individual hardware components, their physical coupling, and the associated control logic. We choose to employ a hierarchical decomposition that separates the control logic into a high level that manages the sequence of control actions and a low level that implements the control actions. The low level is composed of control logic and physical components (sensors and actuators) grouped into a device. In order to capture the physical constraints between the components in a device, we propose the notion of a physical constraint automaton, which is composed with the generic component automata to generate the complete model of the device. We also show how the methodology allows the introduction of component faults into the overall model. The effectiveness of the proposed approach is demonstrated on a micro flexible manufacturing system.

Hardware-based activities for flipping the system dynamics and control curriculum

The advent of new software tools and the rapidly declining cost of hardware have changed the nature of what is required of the modern engineer. These changes require modifications to the curriculum, but also provide a new set of tools to help students learn. The work of this paper outlines some hardware-based activities that have been developed and piloted at the University of Detroit Mercy (UDM) with the purpose of transforming the way system dynamics and control courses are taught to make them more effective and relevant to todays engineering students. These activities offer the promise to improve student understanding of the underlying theory, while exposing students to some practical considerations related to control system design and implementation. The activities that have been developed have been designed to be relatively inexpensive and to require minimal additional instruction for students to carry out. These goals were sought in order to allow the activities to be performed during a short class period (such as in an inverted course), or to be performed outside of class as part of a homework assignment, such that additional courses would not need to be added to an already crowded curriculum. The experiments that have been developed specifically rely on inexpensive Arduino hardware and MATLAB/Simulink software.

Development of an introductory course on the modeling and control of Advanced Electric Vehicles

In January of 2010 the University of Detroit Mercy began offering a Graduate Certificate in Advanced Electric Vehicles. This program was intended as a means to quickly and effectively educate practicing engineers in the skills that they would need as the number of electric and hybrid-electric vehicles being produced increases. The program was initiated with the support of Ford Motor Company who has committed to sending at least 125 engineers through the program in its first five years of existence. A central element of this program is a course on the modeling and control of advanced electric vehicles. This paper describes the structure and content of this course, as well as the lessons learned and what they mean for the introductory system dynamics and control curriculum in general.

Planning under abstraction within a supervisory control context

In this work we employ hierarchy in order to greatly reduce the computational complexity for generating the optimal control for a discrete event system. Specifically, we propose a formal notion of equivalence we term cost equivalence that we use to generate an abstraction of a weighted automaton representing a supervisory controller. We demonstrate that this abstraction can be employed for planning, even in the presence of events that cannot be controlled. Furthermore, we propose additional requirements that we show guarantee that the plan generated from the abstracted automaton achieves optimal behavior when applied to the original unabstracted automaton. The results we generate in this paper have application to a broad class of graph-search problems.

Scaling the formal synthesis of supervisory control software for multiple robot systems

This paper demonstrates the application of a range of theoretical tools to generate real-time control software for multiple ground robots working together cooperatively. Specifically, existing discrete event system theory is applied to synthesize high-level supervisory control logic that is guaranteed to maintain the behavior of multiple robots within requirements defined by a set of formal specifications. The modeling of the high-level behavior of the robots in their given environment, as well as the formal specifications, is described in detail. The resulting models are represented as finite-state automata. In this work we assume that some events cannot be controlled, though all events are assumed to be observable. In addition to generating control logic that is guaranteed to keep the robots safe, results are also presented for choosing from amongst a set of allowed robot behaviors in order to achieve behavior that is “good” in some sense. Specifically, a modified version of Dijkstras algorithm is employed to choose a path through the finite-state automaton representing the allowed robot behaviors. This modified algorithm is able to address multiple robots and the fact that some events cannot be controlled (commanded). The resulting high-level robot events are then connected to the continuous, time-driven behavior of the robots through a series of low-level algorithms. The result of this work is demonstrated in simulation for a simple, but demonstrative scenario.

Simulation-Based Design of the Geometry and Control System for an Omnidirectional Ground Vehicle

This paper describes the simulation-based design of a teleoperated, omnidirectional ground vehicle. The multi-body dynamic simulation employed is developed in the Simulink environment, specifically employing the Simscape extension of Simulink. The accuracy of the simulation is validated by comparison to data taken from one physical instantiation of the vehicle. The use of simulation allows controlled and rapid “testing” of various configurations of the vehicle without requiring any new construction and without endangering physical hardware. The simulation also provides estimates of quantities, like road force and inertial position, that are difficult to measure. The elements of the vehicle design that are specifically investigated in this paper are the vehicle’s physical geometry and its control system. The design of the vehicle’s control system is challenging because of nonlinearities and uncertainty in the model and because it is desired to control three tightly-coupled outputs (longitudinal, lateral, and angular velocity) via six different inputs (force generated at each of the vehicle’s six wheels).© 2014 ASME

Design, Simulation, and Control of a Teleoperated Omnidirectional Ground Vehicle

This paper describes the design, construction, and simulation of a prototype, teleoperated, omnidirectional robotic ground vehicle. The design of a dynamic control system to assist the human operator of the vehicle is also presented. This work sought to test the feasibility of a novel vehicle architecture and to develop a dynamic multi-body simulation tool to assist in the development of future iterations of such a vehicle. The vehicle design seeks to achieve high-speed, omnidirectional mobility, and modest off-road capability. This paper presents results from the physical operation and simulation of the vehicle as well as describing some future work to achieve improved performance of the vehicle system.Copyright