Novel human-centric force control methods of power assist robots for object manipulation

An AMB is not widely used because of its high cost and because its dimensions are significantly larger than those of mechanical bearings. Since a 6 salient-pole type AMB has fewer poles than the usual 8 salient-pole type AMB, it is easier and less costly to miniaturize. However, because the magnetic circuit is not independent in the x and y directions, bearing forces for coil currents become complicated. To solve this problem, two control methods have been proposed. One is a minimum energy control method, which minimizes copper loss in coils. The other is a maximum bearing force control method, which generates a larger bearing force than that of the minimum energy control method. In this study, the two control methods are introduced and rotational tests are carried out for comparison. The rotor could rotate up to 53,000 rpm with the minimum energy control and maximum bearing force control methods. Moreover, rotational tests were carried out with attaching a weight to the rotor. The rotor could rotate up to 42,000 rpm and 38,000 rpm with the minimum energy control method and maximum bearing force control method, respectively. Experimental results show that the minimum energy control method provides less power consumption than the maximum bearing force control method. On the other hand, the maximum force control method provides smaller maximum coil current than that of the minimum energy control method.

This paper deals with the design of a 1DOF power assist robot for lifting objects. Human's vertical lifting force, comprises of inertial force and gravitational force, was considered as the desired dynamics of the robot system. We hypothesized that human's perception of object weight due to inertial force might be different from the perceived weight due to gravitational force for lifting object with the power assist robot system. We then simulated the system using MATLAB/SIMULINK and lifted different sizes of objects with the system. We analyzed human's perceived weights as well as load force characteristics for lifting objects with the system. The analyses show that human's perception of object weight is related to load force rates instead of load force magnitudes. The analyses clarify that the magnitude of the peak load force is to be optimized in order to optimize the motion of the power assist robot. On the other hand, load force rate is to be optimized in order to optimize the feeling of heaviness of object manipulated with the power assist robot. The results also introduce the technique of simultaneous optimization of motion and perceived heaviness. Finally, we proposed using the findings to design human-friendly power assist robots for carrying heavy objects in various industries.

A Force Control Method Using the Relative Motion between the Manipulator and a Workpiece

Power assist robots are usually used for disabled and elderly people to augment their abilities and skills. This paper proposes to use these robots to handle heavy objects in industries, and thus brings a novelty in the applications of power assist robots. However, it is difficult to optimize perceived heaviness and motion either independently or simultaneously for lifting objects with power-assist. Hence, this paper investigates the techniques to optimize perceived heaviness and motion following bionic and psychophysical approaches. We developed two systems-one was used to lift objects manually, and another was a power assist system to lift objects with it. Several hypotheses and strategies related to weight perception and time constant were adopted. Humans lifted objects manually and with power-assist independently. Analyses showed that load force rate for power-assisted lifting were lower than that for manual lifting. We hypothesized that time constant of the assist system might be responsible for this. We changed time constant and found that increase in time constant reduced perceived heaviness and load force. Then, objects were lifted with power-assist in some selected conditions pertaining to time constant. Analyses showed that perceived heaviness was related to load force rate while object motion (acceleration) was related to load force magnitude. It was then demonstrated how to independently optimize perceived heaviness and motion by optimizing load force rate and its magnitude respectively. Techniques for simultaneous optimization of motion and perceived heaviness were also presented. Finally, we proposed to use the findings to develop power assist robots for manipulating heavy objects in industries that may enhance interactions with humans in terms of maneuverability, safety etc.

In this paper we provide a review of the state of the art for active vibration control of axially moving materials (string and beam, etc.) in the past ten years, with particular regard to the subjects of some important active control methods of axially moving strings. First, an introduction is given to explain the aims and scope of this paper. This is followed by a comprehensive discussion of the active wave control methods presented by many investigators in the past ten years, including feedback control methods (such as the modal control method, the direct velocity feedback control method and the control method via transfer function formulation), active wave control methods combining transfer function, boundary control methods using the Lyapunov function, the variable structure control method, the adaptive control method, the vibration control of an axially moving beam and the active control of a moving material with arbitrarily varying length, etc. Finally, we conclude with a discussion of several issues for future research in this area.

In Part I of this work, the dynamic model of drop-induced shock mitigation based on magnetorheological (MR) energy absorber (EA) was deduced. The shock mitigation control effectiveness of constant force control method and optimal Bi number control method were compared based on Bingham model and resistor-capacitor (RC) operator-based hysteresis model. In this part, to further explore the possibility of the application of MREA in the shock mitigation control system, a single-degree-of-freedom (SDOF) shock mitigation control system is built with MREA connected in parallel with a coil spring, which is then connected in series with a vehicle tire to build a one-fourth vehicle suspension shock mitigation control system. Next, the constant force control method and optimal Bi number control method are extended to the SDOF shock mitigation control system. Based on the concept of the optimal Bi number, a two-degree-of-freedom (2DOF) optimal Bi number control method applied to the one-fourth vehicle suspension system is proposed by decoupling and recoupling the damping and stiffness of the vehicle suspension system. Using the control methods conducted in Part I, the motion state of the drop-induced shock mitigation process is simulated and analyzed. The control performances of the control methods in the SDOF shock mitigation control system are compared. Besides, the average velocity change rate (AVCR) and velocity-acceleration conversion ratio (V-ACR) are used to evaluate the effectiveness and stability of the control methods. Finally, the conclusion that the 2DOF optimal Bi number control method can effectively realize the shock mitigation control of a one-fourth vehicle suspension system is verified.

This paper proposes a motion reproducing system for human pinching and rotational tasks in the two dimension. By using coordinate transformation to polar coordinates from Cartesian coordinates of grasping motion, internal force control and attitude control motions are able to reproduce independently. Proposed method makes it possible to reproduce human pinching and rotational motions even if the size of target objects are different compared to motion saving phase. In the experiments, results demonstrate that conventional method to deal with the difference about the size of objects cannot reproduce the motions concern pinching motions and validity of the proposed method is confirmed.

Similar to many complex systems, the operation process of electric balanced forklifts has characteristics such as time-varying model parameters and nonlinearity. Establishing an accurate mathematical model becomes challenging, making it difficult to apply model-based control methods in engineering practice. Aiming at the longitudinal control system of electric forklifts containing external disturbances, this paper proposes an improved full-format dynamic linearization model-free adaptive PID control (iFFDL-MFA-PID) method. Firstly, the full-format dynamic linearization (FFDL) method is employed to transform the operating system of the electric balanced forklift into a virtual equivalent linear data model. Secondly, the nonlinear residual term and pseudo-gradient (PG) of the data model are estimated using the difference estimation algorithm and the optimal criterion function, respectively. Furthermore, in order to enhance the robustness of the system, the idea of intelligent PID (iPID) is introduced and the principle of equivalent feedback is utilized to derive the iFFDL-MFA-PID control scheme. The design process of this scheme only requires the use of the input and output data of the system, without relying on the mathematical model of the system. Finally, the iFFDL-MFA-PID method proposed in this paper is simulated and tested with the EFG-BC/320 counterbalanced forklift equipped in the Special Equipment Testing Center and compared with the model-free adaptive control method (FFDL-MFAC) and the PID control method. Simulation results show that the speed-tracking error of the electric forklift truck under the action of the iFFDL-MFA-PID algorithm is maintained within ±0.132 m/s throughout the process, achieving higher tracking accuracy and better robustness compared to the MFAC and PID methods.

Comparison of myo-electrical control methods: From muscle energy consumption to mechanical speed output

Electromyography (EMG) can be used as an input signal for Human-Computer Interfaces, which are used in the disabled community. Typically, both the user-condition and the end-goal application dictate the number and locations of EMG sensors. We developed two cursor control methods intended for high-level spinal cord injury patients, where available muscle sites are limited to the head and neck areas. For patient comfort, it is desirable to minimize the number of sensors and their intrusiveness while maximizing the functionality of the control method. Both our control methods use a single, noninvasive, surface, differential EMG sensor at the temporalis muscle that controls a cursor in two degrees of freedom (DOF). Fourteen inexperienced able-bodied subjects completed a Fitts’s law task-based cursor-to-target paradigm using both control methods (auto-rotate and rotate). Subjects evaluated and compared performance between the two control methods, which both enabled the cursor to move in 2-DOF. They differed in the level of user control, where one allowed direction manipulation in 1-DOF and 2-DOF in the other. Subjects also completed pre- and post-session surveys and the National Aeronautics and Space Administration Task Load Index for workload assessment. In general, subjects’ performance improved with subsequent sessions within each control method. Subjects achieved a higher throughput (better performance) in the auto-rotate method, had lower workload scores, and tended to prefer this control method. However, about half the subjects felt the manual rotate method allowed them more control over cursor behavior. Our results suggest that a viable cursor control method can be achieved with only a single muscle site.

At present, traffic signal systems almost exclusively use detection systems that provide input in the form of “contact closures,” or binary ON/OFF states representing vehicle presence in detection zones. This limited information constrains the potential performance of traffic signal control. Advanced control algorithms often make use of estimated vehicle arrival times but such data is challenging to estimate with existing detector data. Connected vehicles (CVs) are able to provide such data directly but it is unlikely that all or even most traffic will be connected within the foreseeable future. Vehicle trajectory data from sensors can fill this gap and facilitate a smooth transition from contact closure-based signal control to trajectory-based signal control. This dissertation seeks to improve this situation by exploring potential applications of vehicle trajectory data within existing traffic signal control infrastructure. One issue is the sensing range at which trajectory data can be acquired for intersection control. To test the sensitivity of advanced control methods to the sensing range, two advanced signal control methods, a self-organizing algorithm that adds to existing actuated control methods, and the Phase Allocation Algorithm proposed in recent research on CV-based signal control, were implemented and tested in simulation under varying volume configurations and sensing ranges. The results show that increasing sensing range is associated with improved performance, with the performance of the two advanced control methods converging at sensing ranges of 1000–1500 ft. Actuated control is an existing function in all modern traffic signal controllers and presents an opportunity for scalable improvement of signal control by replacing contact-closure detection with trajectory data. The second study described in this thesis explores the replacement of conventional passage timer logic, which seeks to measure gaps in traffic, with direct measurement of gaps with trajectory data. A highly customizable and robust signal controller was developed in a simulation framework to facilitate a proof-of-concept study. A trajectory-based control method called “gap seeking” was developed that uses trajectory data to identify gaps in the arriving traffic. A variation on this control method called “advanced gap seeking” searches for gaps at an upstream location. Results of a simulation study demonstrate that the methods are able to deliver significant improvements over conventional control methods. The third study of this dissertation proposes additional trajectory-based actuation methods for green extension and evaluates their performance individually and in combination when implemented under the previously developed gap seeking methods, under conventional control, and in a hybrid detection environment where trajectories are only available on the major street. The proposed methods include type II dilemma zone protection, a method called “queue clearance” that aims to eliminate split failures by extending green until vehicles that arrived in red have cleared the intersection, and two algorithms for accommodating platoons from previous studies, which carry a potential for self-organizing control. Simulation evaluations suggest that these methods individually meet their objectives, with some potential trade-offs in delay. Some combinations of the proposed methods show potential for substantial improvement over existing control methods.

An improved centralized/decentralized accurate reactive power sharing method in AC microgrids

This article makes a comparative analysis between the vector control, and the method of direct electromagnetic force control of linear induction motors. The dynamic models of linear induction motors applying the two control methods are presented. For verification of the effective control of linear induction motors comparative simulations are made.

The Universal Jamming Gripper, made up of a coffee grain-filled balloon membrane and a vacuum pump, operates by applying an unvarying or constant negative pressure set by the manipulator on the target object for gripping, not taking into account the suitable amount of negative pressure depending on the size and weight of the target object. With the constant or same amount of negative pressure applied on the target objects of different sizes and weight, the gripper can diminish the structural integrity of some of these target objects. To eradicate this problem, the researchers developed a system to vary the negative pressure of the Universal Jamming Gripper through an artificial neural network depending on the size of the object, obtained through a vision-based object classification scheme, and weight obtained from a load cell connected to the computer. An Artificial Neural Network (ANN), trained with three inputs, such as the pixel area of one side of the target object, the pixel area of another side of the object and its weight, is used to automatically determine the optimum negative pressure needed to successfully grip the target object. After testing and experimentation, the ANN is proven to output the optimum negative pressure needed to successfully conform to and grip the target object, as evident with the 99.131% result from testing, based on the regression plot from MatLab.

The study of control methods for the robotic testing system for human musculoskeletal joints