4,130 research outputs found
Stochastic Control of Population Dynamics Using Kalman Filtering with Applications to Artificial Muscle Recruitment
No full textThis paper addresses a problem in distributed control: given a large number of identical hybrid-state agents, control the ensemble behavior of the agents assuming that only limited information is available about the agents' states. This process has relevance to a number of biologically-inspired control problems, such as motor recruitment. In this paper, we describe a stochastic control policy capable of achieving convergent control of the distribution of an ensemble of finite state agents in this way. Using techniques developed for the observation of biological population dynamics, we show that it is possible to observe the state distribution of agents under our control policy using a Kalman filter. Look-ahead control laws based on the Kalman filter estimates are used to achieve a high degree of stability and robustness in systems exhibiting large time delays. An example of control over a hybrid-state, recruitment-like controller for an artificial muscle is presented
Stable control of distributed hysteretic systems using cellular broadcast stochastic feedback
No full textThis work develops a provably stable broadcast feedback approach to regulating the aggregate output of a collection of stochastically behaving cellular units with pronounced hysteresis. Similar to skeletal muscle, comprised of a number of muscle fibers, the cellular units are independent and each contributes to an aggregate system output. The control architecture consists of a central controller that only observes the aggregate system output and cell-level controllers that manage cell state on the hysteresis loop. The central controller stochastically recruits cells to reduce expected squared error. Stability analysis based on supermartingale theory guarantees that the control framework is stable.National Science Foundation (U.S.) (Grant no. NSF CMSI-0728162
Harry Pepper Fonds
No full textThe fonds consists of newspaper articles written by Harry Pepper from the Rossland Miner and the Trail Times.Born in 1913, Harry Pepper spent most of his youth in Suffolk, England, before immigrating to Canada in 1929. He spent his early years in Canada working on farms during the summer months, and in the bush during the winter. In 1940 he enlisted for the war, and joined the RoyaL Canadian Artillery, fighting in the 8th Canadian Field Regiment. During the war, Harry served in England, North Africa, Italy, France, Belgium and Holland. While overseas, he married, and his new family returned to Canada in 1945. He moved to Trail as a foreman of the 4X Bakery in 1946, and was later employed by Buchan’s Bakery. In 1949 he began working at Cominco, and stayed there until his retirement. Harry Pepper was incredibly active in the sports community in Rossland, and volunteered with soccer, baseball, softball, lacrosse, and hockey. He was also an avid Curler and Golfer. He wrote a sports column for the Rossland Miner called “As I See It” and stayed with the newspaper until the building burned down and it subsequently went out of business. On his 60th birthday, he was asked to do a sports column in the Trail Daily Times called “Pepp Talk”
Scaling up shape memory alloy actuators using a recruitment control architecture
No full textThis paper presents new experimental results from a human-size robotic forearm, created to demonstrate the effectiveness of recruitment-based control architectures for large actuators made from shape memory alloys (SMA) and other active materials. The robot arm is actuated antagonistically by two actuators made up of 60 small SMA springs arranged in parallel, which are activated in an on/off fashion using Joule heating. The force and stiffness of each actuator is controlled by recruiting a desired number of springs to contract. The joint position is then controlled using equilibrium point servo control. The results presented in this paper show that the combination of equilibrium point control of the arm joint and recruitment-based control of each actuator's stiffness solve some of the major problems of scalability and response speed often associated with active material actuators.National Science Foundation (U.S.) (grant number 0143242
Nonlinear Feedback Control of a Gravity-Assisted Underactuated Manipulator With Application to Aircraft Assembly
No full textA nonlinear feedback scheme for a gravity-assisted underactuated manipulator with second-order nonholonomic constraints is presented in this paper. The joints of the hyper articulated arm have no dedicated actuators but are activated by gravity. By tilting the base link appropriately, the gravitational torque drives the unactuated links to a desired angular position. With simple locking mechanisms, the hyperarticulated arm can change its configuration using only one actuator at the base. This underactuated arm design was motivated by the need for a compact snake-like robot that can go into aircraft wings and perform assembly operations using heavy end-effectors. The dynamics of the unactuated links are essentially second-order nonholonomic constraints for which there are no general methods to design closed-loop control. We propose a nonlinear closed-loop control law that is guaranteed to be stable in positioning one unactuated joint at a time. We synthesize a Lyapunov function to prove the convergence of this control scheme. The Lyapunov function also generates estimates of the domain of convergence of the control law for various control gains. The control algorithm is implemented on a prototype three-link system. Finally, we provide some experimental results to demonstrate the efficacy of the control scheme.Boeing Compan
Mag-Foot: a steel bridge inspection robot
No full textA legged robot that moves across a steel structure is developed for steel bridge inspection. Powerful permanent magnets imbedded in each foot allow the robot to hang from a steel ceiling powerlessly. Although the magnets are passive, the attractive force is modulated by tilting the foot against the steel surface. This allows the robot to slide its feet along the surface using "Moonwalk" and "Shuffle" gait patterns. The robot can also detach its feet and swing them over small obstacles. These diverse walking patterns are created with a single servoed joint and 2 sets of simple locking mechanisms. Kinematic and static conditions are obtained for the under-actuated legged robot to perform each gait pattern safely and stably. A proof-of-concept prototype robot is designed, built, and tested. Experiments demonstrate the feasibility of the design concept and verify the analytical results
Letter from Harry and Yaso Ueno to Michi and Walter Weglyn, June 03, 1987
No full textA letter from Harry and Yaso Ueno to Michi and Walter Weglyn in which the authors lament a United States Supreme Court and criticize Japanese American Citizens' League (JACL) leader Mike Masaoka.These materials are from box 73 and 74 of the Frank Chin Papers. The Frank Chin Papers contain personal and professional correspondence between Frank Chin and Michi Weglyn relating to particular projects on which either author was working as well as files related to the Day of Remembrance Tribute to Michi Weglyn
Wearable Conductive Fiber Sensors for Multi-Axis Human Joint Angle Measurements
No full textBackground: The practice of continuous, long-term monitoring of human joint motion is one that finds many applications, especially in the medical and rehabilitation fields. There is a lack of acceptable devices available to perform such measurements in the field in a reliable and non-intrusive way over a long period of time. The purpose of this study was therefore to develop such a wearable joint monitoring sensor capable of continuous, day-to-day monitoring. Methods: A novel technique of incorporating conductive fibers into flexible, skin-tight fabrics surrounding a joint is developed. Resistance changes across these conductive fibers are measured, and directly related to specific single or multi-axis joint angles through the use of a non-linear predictor after an initial, one-time calibration. Because these sensors are intended for multiple uses, an automated registration algorithm has been devised using a sensitivity template matched to an array of sensors spanning the joints of interest. In this way, a sensor array can be taken off and put back on an individual for multiple uses, with the sensors automatically calibrating themselves each time. Results: The wearable sensors designed are comfortable, and acceptable for long-term wear in everyday settings. Results have shown the feasibility of this type of sensor, with accurate measurements of joint motion for both a single-axis knee joint and a double axis hip joint when compared to a standard goniometer used to measure joint angles. Self-registration of the sensors was found to be possible with only a few simple motions by the patient. Conclusion: After preliminary experiments involving a pants sensing garment for lower body monitoring, it has been seen that this methodology is effective for monitoring joint motion of the hip and knee. This design therefore produces a robust, comfortable, truly wearable joint monitoring device.National Science Foundation (Grant: NSF 0097700
Control of a pneumatically actuated joint for wearable supernumerary robotic limbs application
No full textThesis: S.B., Massachusetts Institute of Technology, Department of Mechanical Engineering, 2016.Cataloged from PDF version of thesis.Includes bibliographical references (page 31).Presented is work on the development of the Supernumerary Robotic Limbs project, headed by Federico Parietti in the d'Arbeloff Labs under Prof. Harry Asada. Specifically, this paper focuses on the integration of lightweight, pneumatic systems for prismatic joint actuation, and the various control schemes studied. This joint serves as the leg of the robot, and extends from the hip of the wearer to contact the ground. The design consists of a two-way pneumatic cylinder inside a load bearing carbon fiber sleeve, actuated with a nominally closed 5-3 way solenoid valve, and weighs in at <1kg per actuator. The positional control scheme is closed via tracking from a linear magnetopotentiometer, while the force control scheme utilizes both the positional tracking as well as a load cell at the foot of the leg. System modeling of the actuator dynamics allowed for development of a model based proportional control method. Optimization of the proportional gain and system delay time produced a rise time of 200ms given a step input command for a 250mm stroke. The developed scheme was implemented in the full wearable system to assist a human support weight in crouched positions and standing up from a sitting position. Initial testing has shown the effectiveness of the power, compactness and compliance of pneumatic systems in a wearable robotic device.by Roger Lo.S.B
Kalman filter for inhomogeneous population Markov chains with application to stochastic recruitment control of muscle actuators
No full textA population of stochastic agents, as seen in swarm robots and some biological systems, can be modeled as a population Markov chain where the transition probability matrix is time-varying, or inhomogeneous. This paper presents a Kalman filter approach to estimating the population state, i.e., the headcount of the number of agents in each possible agent-state. The probabilistic state transition formalism originated in Markov chain modeling is recast as a standard state transition equation perturbed by an additive random process with a multinomial distribution. An optimal linear filter is derived for the recast state equation; the resultant optimal filter is a type of Kalman filter with a modified covariance propagation law. Convergence properties are examined, and the state estimation error covariance is guaranteed to converge. The state estimation method is applied to stochastic control of muscle actuators, where individual artificial muscle fibers are stochastically recruited with probabilities broadcasted from a central controller. The system output is the resultant force generated by the population of muscle fibers, each of which takes a discrete level of output force. The linear optimal filter estimates the population state (the headcount of agents producing each level of force) from the aggregate output alone. Experimental results demonstrate that stochastic recruitment control works effectively with the linear optimal filter.National Science Foundation (U.S.) (Grant CMSI 0728162
- …
