R-3 Repository :: Browsing by Author "Cheatham, John B., Jr."

Browsing by Author "Cheatham, John B., Jr."

Now showing 1 - 20 of 24

A connectionist approach to autonomous robotic navigation
(1991) Weiland, Peter Lawrence; Cheatham, John B., Jr.
Robotic navigation has been an area of intense research since the onset of mobile robot development. The usefulness of mobile robots ultimately reside in their ability to move and interact with the environment. Current approaches to robotic navigation are primarily based on simulating intelligent, human-like behavior through the intelligent system model processing cycle; sense, perceive, reason, act. Unlike these methods, this thesis presents a navigation system based on biological and behavioral principles which functions in a stimulus-response manner. Using connectionist architectures, a relationship between stimulus and response is required through the learning of conceptual information pertaining to navigation. In this research, the mammalian visual system provides a guide for the processing of environmental stimulus. Simulated laser range data are processed in retinal patch size elements by a cellular neural network. This network is designed to detect obstacle existence for each patch segment based on an invariant feature of range discontinuity. Obstacle information is then mapped in binary format, indicating the traversable state of the patch, to the system's visual cortex. Response to this mapping is derived from a hierarchical structure of back error propagation neural networks in which each network has learned a particular navigational behavior; obstacle avoidance, wander, and goal seeking. Output from these networks indicate an appropriate motor response for the environmental stimulus. A simulation was developed to evaluate the performance of this system by having a robot traverse an environment. The connectionist approach was verified through system display of human-like navigational behavior for the simulation's environment. Advantages of the neural network approach were also demonstrated by its processing speed and adaptability. Procedures are discussed for actual system implementation in which cycle times of under one second are completely feasible. Proposals for unsupervised learning of navigational responses for environmental stimulus are also made. From the research presented in this thesis, a foundation is established for continuing the study of the connectionist approach to the problem of autonomous navigation.
A modular fingertip design for normal force feedback control of a 3-fingered dextrous robotic hand
(1992) Grisham, Joe Paul; Cheatham, John B., Jr.
Despite the advances in dextrous hand mechanics over the past decade, fingertip force remains predominantly controlled by flexor cable tension and joint position. A modular fingertip normal force sensor is developed here from basic semiconducting elastomer devices. Construction techniques are defined to shape standard Interlink Force Sensitive Resistors within a conformal mold of the human thumb and provide standard plug-in capability. Current-to-voltage electronic circuits are incorporated to linearize output and minimize the total sensor package. The sensor system is installed on each of three fingers and demonstrated to provide full control in elementary grasping strategies of a robotic hand.
A neural network approach to the redundant robot inverse kinematic problem in the presence of obstacles
(1991) Norwood, John David; Cheatham, John B., Jr.
Redundancy in robots is very much an open research area in the field of robotics. As the tasks required of robots become more and more complex, the ability of robots to perform satisfactorily in these applications must increase accordingly. Redundant manipulators have a greater ability to perform difficult tasks, such as obstacle avoidance, than non-redundant ones. In order to make use of this extra ability of redundant robots, more effective control schemes must continue to be developed and to this end, more and more researchers are looking to expand the body of knowledge in this area. This thesis addresses the problem of moving a redundant robot within a defined workspace in the presence of obstacles. Additionally, criteria are developed that may be applied to the robot to constrain the redundant equations. Finally, a neural network solution to the redundant inverse kinematic problem is presented. It will be shown that the inverse kinematics can be developed through a network architecture which provides accurate and fast solutions to a problem that is computationally and structurally complex. This effort will be kept within the context of already accepted methods currently in use for redundant robots, the overall goal of this research being to firmly establish the usefulness and applicability of neural network architectures to difficult robotic problems.
A new method for solving the kinematics of multifingered grasping and general redundant manipulators: A task oriented approach
(1991) Chen, Yu-Che; Cheatham, John B., Jr.; Walker, Ian D.
Beside our arms, it is our hands that help us to perform tasks. At first glance, we seems to move our arms and use our hands naturally. On closer inspection, however, we find that both our arms and hands are redundant mechanisms which explains why our arms can approach an object using different postures and our hands can grasp the object with many feasible grasp configurations. Duplicating this phenomenon has led to many innovative designs of redundant arms as well as multifingered hands, and has sparked many useful analyses of these topics. This thesis presents a new approach to intelligent multifingered grasping and redundant arm manipulation. Methods proposed in this research yield a computationally efficient and physically meaningful model for the planning of grasp positions, the determination of squeezing finger forces and the visualization of the motions of redundant arms. Principles and concepts embedded in this analysis will help researchers to gain new insights toward better designs for hands and arms. The model developed in this thesis also provides mathematical justification for some of the motions of human arms and fingers. The main thrust of this analysis of multifingered grasping is the use of mechanical equivalent force/moment systems. These systems allow us to decompose each finger force into a normal and a tangential component. Using this decomposition of finger forces, we can visualize the contribution of each finger force by the resultant force and moment required to manipulate a grasped object. Additionally, the nature of this decomposition, which is one of the unique feature of our method, will allow intelligent utilization of touch sensors. The frictional constraints and the squeezing internal finger forces are elegantly taken into account through a computationally efficient algorithm for choosing grasp points. Our methods have also been extended into the grasp of solid objects. The second part of this thesis provides an efficient method for analyzing the inverse kinematic problem of redundant manipulators. Our method is based on fully using the directional properties of the columns of the Jacobian matrix which relates the joint velocities of the arm and the end effector velocities. Motions of a redundant manipulator are analyzed through its upper arm's motion and wrist's motion and the Jacobian matrix is partitioned accordingly. It is shown that the motion of the upper arm and the wrist can be evaluated separately and in parallel though this vectorized approach.
A path planning and obstacle avoidance hybrid system using a connectionist network
(1991) Schuster, Christopher Emmet; Cheatham, John B., Jr.
Automated path planning and obstacle avoidance has been the subject of intensive research in recent times. Most efforts in the field of semiautonomous mobile-robotic navigation involve using Artificial Intelligence search algorithms on a structured environment to achieve either good or optimal paths. Other approaches, such as incorporating Artificial Neural Networks, have also been explored. By implementing a hybrid system using the parallel-processing features of connectionist networks and simple localized search techniques, good paths can be generated using only low-level environmental sensory data. This system can negotiate structured two- and three-dimensional grid environments, from a start position to a goal, while avoiding all obstacles. Major advantages of this method are that solution paths are good in a global sense and path planning can be accomplished in real time if the system is implemented in customized parallel-processing hardware. This system has been proven effective in solving two- and three-dimensional maze-type environments.
A practical application of a force feedback control system for a mechanical, 3-fingered, dextrous hand
(1992) Hanusa, Paul Jeffrey; Cheatham, John B., Jr.
The trajectory planning and control system for a mechanical, 3-fingered, dextrous hand are analyzed and implemented. The forward and inverse kinematics of the fingers are derived and used to analyze the capabilities of the hand. The simple and effective grasping strategy employs minimal trajectory planning and relies solely on force feedback to define finger position and orientation. A force feedback controller is implemented on the hand which allows independent control of each finger. The hand is used in conjunction with a robotic arm to execute various tasks which require hand dexterity.
A prosthetic hand with tactile sensing
(1994) Gunawardana, Ruvinda Vipul; Cheatham, John B., Jr.
With recent advances in myoelectric signal processing, powered prosthetic hands with myoelectric control have become more accepted among users of prosthetic hands. Inspite of this progress, the hands that are currently available have limited functionality and provide little or no sensory feedback to the user. Thus the user cannot control the forces exerted by the hand on an object being grasped. The objective of the myoelectric hand project at Rice University is to create a hand that has greater functionality and also the ability to sense and control the forces. In the research described in this thesis, a prosthetic hand that is capable of controlling the grasp strength is built. Force Sensitive Resistors manufactured by Interlink electronics are used to sense the forces. These sensors are installed at eleven likely locations of contact. The system demonstrated the ability to control the grasp forces.
A semi-autonomous mobile robot/teleoperator with applications as an aid for severely handicapped people
(1990) Regalbuto, Michael Anthony; Cheatham, John B., Jr.
A framework for a practical, easy-to-use system which can allow severely handicapped people to control and interact with their home and work environments is presented. Software written for a personal computer allows the user to command a robot to move about a structured indoor environment as well as control household appliances. Underlying these functions is a user-generated computer model of the environment which maintains spatial and relational information about household items. The model represents a data base which is accessed by a variety of procedures in commanding the robot to perform specific tasks, such as localization and navigation. Developed incrementally, a model can take on as much detail as necessary to command the robot to perform a given task. Beyond its use as a domestic aid, the system can serve as an inexpensive testbed for research in intelligent mobile robotics and can ultimately be targeted for use by the general population. The system is composed of relatively inexpensive equipment designed for in-home use. Software is written in object-oriented Allegro Common LISP and runs on a Macintosh SE having 4Mb of RAM. A HERO 2000 robot was assembled from a kit and modified to extend its workspace to cover desks and tabletops. Control of lamps and household appliances is via an X-10 Powerhouse computer interface module, which is connected to the SE. An Esteem 9600 baud wireless modem augments the 600 baud radio frequency communications link provided with the robot, allowing untethered operation. A custom Machine Control Interface has been developed for interfacing severely physically handicapped users to the system.
A vision-based fuzzy logic and neural network approach to the control of hyper-redundant robot manipulators
(1994) Magee, Kevin Nowell; Cheatham, John B., Jr.
Hyper-redundant robot manipulators possess a very large degree of kinematic redundancy and are capable of motion similar to that of snakes and elephant trunks. Because of the computational burden required to calculate the pseudo-inverse of the manipulator Jacobian matrix for high degree of freedom robots, hyper-redundant manipulators have proven challenging to control by traditional methods. Additionally, control can be further complicated because the large number of joints and links in a hyper-redundant arm can be a source of measurement error in real-world systems. A fuzzy logic and neural network based control system for hyper-redundant arms is presented which operates on data from real-time vision. The neural network maps goal position and orientation to desired arm configuration. A modified region fill algorithm is used to provide an estimate of the current configuration as seen in two camera views. A fuzzy logic rule base constructed from human intuition specifies motor signals which servo the arm from the current position to the goal position according to velocity profiles modeled after human goal-directed movement strategies. As a test case, the control system is applied to a thirty-two degree of freedom robot arm designed and built at Rice University. The controller is demonstrated to provide accuracy similar to that of humans on certain tasks. Although application is made specifically to a hyper-redundant arm, the control system developed in this research also could be applied to many lower degree of freedom manipulators, provided that their motion is heuristically easy to describe.
An experimental investigation of indexed indentation of rock
(1967) Musselman, John Anthony; Cheatham, John B., Jr.
An experimental investigation is made of the idealized drilling problem of a flat punch loading a horizontal rock surface near a previous crater. Tests are run on a limestone to determine the force required to form a chip and the chip shape as functions of the confining pressure, the inclination of the free surface from the horizontal loaded surface, and the placement or "indexing" distance of the load from the free surface. The results are compared with the theoretical predictions of plastic limit analysis.
Design, analysis, implementation, and control of a mobile robotic testbed for telepresence
(1992) Adnan, Sarmad; Cheatham, John B., Jr.
A unique mobile telepresence testbed has been designed and implemented. This testbed is a tool for research in telepresence and tele-existence for control of remote robotic systems. An eight-degree-of-freedom, redundant manipulator has been designed and implemented for this system. Resolved acceleration control and impedance control have been demonstrated. An omnidirectional base has been built to provide human-like movement capabilities to the telepresence testbed. Control software written for the system allows easy control of the base and the arm. Hand-controllers are used to guide the system trajectories. Ethernet, serial links, or wireless radio modems can be used as the control medium. Use of individual motor control processors for each motor allows high servo update rates to be achieved. A high level, modular and extensible library of routines has been written to allow easy programming of the system by future researchers. A head-tracking platform with color stereo cameras provides video feedback to the operator with depth perception to allow fine manipulation tasks.
Development of a strain softening constitutive model for rock
(1988) Claborn, Forrest Jay; Cheatham, John B., Jr.
A plasticity based constitutive model with two yield surfaces is developed to model the nonlinear axial and volumetric behavior observed for rocks in triaxial compression. The failure surface is allowed to soften at low pressures to simulate fracturing. A method for treating the corner where the yield surfaces intersect is developed. In comparisons with experimental data, the model accurately predicts the axial response in triaxial tests. The model predicts too much dilation for low confining pressure triaxial tests. The constitutive model was used in conjunction with a finite element code to simulate axisymmetric indentation of rock. Because of the existence of the corner in the yield surface, the simulation was computationally too slow to be practical.
Development of sensors and algorithms for automating robotic grasping
(1989) Fisher, Paul Bruner; Cheatham, John B., Jr.
The difficulty of locating and grasping a target via a robot functioning as a remote teleoperator varies with the operator's perspective of the target area and the skill of the operator. To limit the amount of time and skill required to grasp a target, several methods were investigated to automate this process. These methods involved the development of sensor systems and control algorithms directed at the detection of the target in a pre-defined work area, location of the target's edges, approaching the target along its centerline, and ultimately grasping the object. The specific application involved in this research concerned a mobile robot locating and grasping random objects within a room. This thesis describes the results obtained with both the tactile and the non-tactile sensors. The non-tactile sensors were chosen for the final configuration because they could locate the target more quickly and they were safer.
Forces required to roll an idealized drill bit across a Coulomb plastic material
(1968) Biggs, Michael Don; Cheatham, John B., Jr.
This thesis presents a systematic method for determining the forces on a simplified roller bit. The kinematics of an actual roller bit drilling in rock is complex; nevertheless, it is desirable to obtain an accurate analytical model. As a preliminary study, the roller bit is replaced in this thesis by a simple, two-dimensional bit and the rock is replaced by a Coulomb plastic material. Results from previous work, plus additional assumptions, were incorporated into a computer program to determine the forces required to move the bit in a specified manner. Twelve problems were worked to test the program. By studying similar simplified drilling problems, accompanied by supporting laboratory work, one could eventually obtain an accurate analytical model for drilling of an actual roller bit. This could lead to the compilation of a catalog of results which, in turn, would allow selection of the optimum drill bit for any given situation.
Genetic algorithm tuning of a fuzzy logic controller for a dynamic system
(1995) Wang, Lui; Cheatham, John B., Jr.
This thesis describes the use of the genetic algorithm to facilitate the design process of a fuzzy logic based controller. The basic mechanics of both the fuzzy logic control system and the genetic algorithm are presented for establishing the foundation of this research. A detailed design process to integrate the fuzzy logic controller and the genetic algorithm is disclosed. Software is developed to simulate the dynamics of a two-link planar manipulator, and a multiple-link fuzzy logic control (MLFLC) system is developed to control a non-linear robotics system. In this work, the genetic algorithm technique is used to design both the Universe of Discourse of some of the control variables and also the shape and location of membership functions. As a result, the system is able to maintain control with moderate errors. The most important contribution of this work is that it demonstrates the effectiveness of using genetic algorithms to optimize fuzzy logic control systems.
Kinematic analysis and trajectory control of the Rice omni-directional mobile robot (Rice University, Texas)
(1989) Adnan, Sarmad; Cheatham, John B., Jr.
The forward/inverse kinematics, trajectory planning and control system for a mobile omni-directional robot are analyzed and implemented. For kinematic analysis each wheel of the robot has been modeled as a "higher pair" joint. Both the forward and inverse kinematic solutions of the robot are dealt with. The trajectory planning for the robot is done on two levels, a global robot trajectory and individual wheel trajectories. A simple, Linear Segment with Parabolic Blends, trajectory is used on the local, wheel level. On the global level the robot can be made to move along higher order trajectories. A PID control system for the robot is implemented for this purpose. Using a hierarchical control scheme the wheel encoder feedback is converted to global motion feedback. This control scheme does not account for wheel slippage.
Myoelectric signal recognition using genetic programming
(1995) Fernandez, Jaime Julio; Cheatham, John B., Jr.
This thesis presents a new method of myoelectric signal recognition. Myoelectric signals are electric signals generated by the motion of a person's muscle and can be used as control input for prosthetic hands. It uses genetic programming to create a set of equations capable of recognizing three different myoelectric signals. Three different approaches are presented. The first approach uses genetic programming to create three separate equations. Each equation is capable of recognizing a different pair of the three myoelectric signals. The solution is accomplished by the signal that exactly corresponds to two of the three equations. The second approach creates a single equation capable of distinguishing the three signals. The last approach is a hybrid solution. It uses a simple equation to distinguish 90% of the three signals. It then uses a more complicated equation to distinguish the remaining 10% of the signals.
Plastic limit analysis of a coulomb material loaded by a flat punch
(1966) Pittman, Robert Wayne; Cheatham, John B., Jr.
The critical load required for a rigid, flat punch to cause failure of a material obeying the linear Coulomb-Mohr yield criterion in plane strain is bracketed by upper and lower bound values determined by plastic limit analysis. These values are a function of the cohesion and angle of internal friction of the material, the inclination of the free surface from the horizontal loaded surface, and the placement or "indexing" distance of the load from the free surface. Theoretical failure paths indicating possible volume removal of the material under critical loading by the flat punch are postulated and compared with actual failure paths observed in damp sand model experiments at atmospheric conditions.
Post buckling behavior of a circular rod constrained within an inclined hole
(1988) Chen, Yu-Che; Cheatham, John B., Jr.
A three-dimensional model of the deflected configuration of the portion of a circular rod constrained to be in contact with a circular cylinder is developed in this thesis. This work involves the development of an analytical method to determine the postbuckling behavior of tubular members in inclined holes. The analytical method introduces the concept of the critical hole inclination angle, $\alpha\sb{\rm cr}$, and the buckling, $\alpha\sb{\rm cr*}$, to determine the configurations of the rod. For hole inclination angles less than $\alpha\sb{\rm cr}$, a rod is in a helical form. A formula for the wall contact force in this region is derived to explain the existence of hysteresis during a cycle of axial loading followed by a corresponding load decrease. Torque effects and gravity effects are found to affect the helical configuration of the rod slightly. When the hole inclination lies between these two angles, a rod falls into a sinusoidal configuration. (Abstract shortened with permission of author.)
Remote control of a robotic arm through speaker-dependent, isolated-word speech recognition
(1991) Atkinson, William Thomas; Cheatham, John B., Jr.
The focus of this thesis is the remote control of a six degree-of-freedom mid-range industrial robot arm through the use of speech recognition. The robot arm is controlled through speaker-dependent, isolated-word speech recognition. Template matching techniques are used to interpret verbal input and make the robot arm perform a desired task. The user inputs one- or two-word commands to interact verbally with the robot. The verbal input is matched against a vocabulary of known commands for recognition. Once the command is interpreted, the corresponding subroutine is run and the robot arm performs the task. The movement of the arm, to include the manipulation of its gripper, is accomplished in discrete increments. The program is designed to allow for multiple users and it can be modified to perform robotic control for other tasks.

Browsing by Author "Cheatham, John B., Jr."

Results Per Page

Sort Options