Person:

Ueda, Jun

Permanent Link

https://hdl.handle.net/1853/71794

Associated Organization(s)

Organizational Unit

George W. Woodruff School of Mechanical Engineering

Full item page

Publication Search Results

Now showing 1 - 10 of 31

Encrypted Model Reference Adaptive Control with False Data Injection Attack Resilience via Somewhat Homomorphic Encryption-Based Overflow Trap

(Georgia Institute of Technology, 2024-08) Ueda, Jun ; Blevins, Jacob

Cloud-based control is prevalent in many modern control applications. Such applications require security for the sake of data secrecy and system safety. The presented research proposes an encrypted adaptive control framework that can be secured for cloud computing with encryption and without issues caused by encryption overflow and large execution delays. This objective is accomplished by implementing a somewhat homomorphic encryption (SHE) scheme on a modified model reference adaptive controller with accompanying encryption parameter tuning rules. Additionally, this paper proposes a virtual false data injection attack (FDIA) trap based on the SHE scheme. The trap guarantees a probability of attack detection by the adjustment of encryption parameters, thus protecting the system from malicious third parties. The formulated algorithm is then simulated, verifying that after tuning encryption parameters, the encrypted controller produces desired plant outputs while guaranteeing detection or compensation of FDIAs. This is a preprint version of the manuscript.
Secure Teleoperation Control Using Somewhat Homomorphic Encryption

( 2022-10) Kosieradzki, Shane ; Zhao, Xiaofeng ; Kawase, Hiroki ; Qiu, Yingxin ; Kogiso, Kiminao ; Ueda, Jun

The goal of this research is to establish control theoretic methods to enhance cyber security of networked motion control systems by utilizing somewhat homomorphic encryption. The proposed approach will encrypt the entire motion control schemes including: sensor signals, model parameters, feedback gains, and performs computation in the ciphertext space to generate motion commands to servo systems without a security hole. The paper will discuss implementation of encrypted bilateral teleoperation control schemes with nonlinear friction compensation. The paper will present (1) encrypted teleoperation control realization with somewhat homomorphic encryption and (2) simulation results.
Safe, Secure, and Stable Motion Control of Telemanipulators

(Georgia Institute of Technology, 2021-08-25) Ueda, Jun
Dynamics-based motion de-blurring for a PZT-driven, compliant camera orientation mechanism

(Georgia Institute of Technology, 2015-02) Kim, Michael ; Ueda, Jun

This paper proposes a method for removing motion blur from images captured by a fast-moving robot eye. Existing image techniques focused on recovering blurry images due to camera shake with long exposure time. In addition, previous studies relied solely on properties of the images or used external sensors to estimate a blur kernel, or point spread function (PSF). This paper focuses on estimating a latent image from the blur images taken by the robotic camera orientation system. A PZT-driven, compliant camera orientation system was employed to demonstrate the effectiveness of this approach. Discrete switching commands were given to the robotic system to create a rapid point-to-point motion while suppressing the vibration with a faster response. The blurry images were obtained when the robotic system created a rapid point-to-point motion, like human saccadic motion. This paper proposes a method for estimating the PSF in knowledge of system dynamics and input commands, resulting in a faster estimation. The proposed method was investigated under various motion conditions using the single-degree-of-freedom camera orientation system to verify the effectiveness and was compared with other approaches quantitatively and qualitatively. The experiment results show that overall the performance metric of the proposed method was 27.77% better than conventional methods. The computation time of the proposed method was 50 times faster than that of conventional methods.
Pneumatically-Powered Robotic Exercise Device to Induce a Specific Activation Pattern in Target Lower Extremity Muscles

(Georgia Institute of Technology, 2014-12) Henderson, Gregory ; Ueda, Jun

The goal of this research is to establish a methodology to actively control a pneumatically driven robotic device that can induce specific muscle force patterns in target muscles during a subject’s voluntary movement. In this paper, the generation of constant forces in the rectus femoris muscle throughout the knee extension, i.e., isotonic contractions, was studied. Due to a highly nonlinear nature of mapping the joint torque to muscle force, a simple application of constant torques to the knee joint would not realize isotonic contractions. The proposed robotic exercise accounted for nonlinear moment arms of muscles as functions of joint angles and nonlinear coordination of multiple muscles in the neuromuscular system to accomplish individual muscle control. A pneumatically powered one degree of freedom (DOF) device that can impose active force feedback control has been designed and built. An exercise-planning algorithm has been developed that involved a musculoskeletal model of the lower-body, and the dynamics of a pneumatic actuator. Five constant force profiles were tested for twenty healthy volunteers and electromyographic (EMG) signals were collected while the device was applying calculated force profiles.
Dynamic Cellular Actuator Arrays and Expanded Fingerprint Method for Dynamic Modeling

(Georgia Institute of Technology, 2014-07) MacNair, David ; Ueda, Jun

A key step to understanding and producing natural motion is creating a physical, well understood actuator with a dynamic model resembling biological muscle. This actuator can then serve as the basis for building viable, full-strength, and safe muscles for disabled patients, rehabilitation, human force amplification, telerobotics, and humanoid robotic systems. This paper presents a cell-based flexible actuator modeling methodology and the General Fingerprint Method for systematically and efficiently calculating the actuators’ respective dynamic equations of motion. The cellular actuator arrays combine many flexible ‘cells’ in complex and varied topologies for combined large-scale motion. The cells can have varied internal dynamic models and common actuators such as piezoelectric, SMA, linear motor, and pneumatic technologies can fit the model by adding a flexible element in series with the actuator. The topology of the cellular actuator array lends it many of its properties allowing the final muscle to be catered to particular applications. The General Fingerprint Method allows for fast recalculation for different and/or changing structures and internal dynamics, and provides an intuitive base for future controls work. This paper also presents two physical SMA based cellular actuator arrays which validate the presented theory and give a basis for future development.
Improved Stability of Haptic Human-Robot Interfaces using Measurement of Human Arm Stiffness

(Georgia Institute of Technology, 2014-06) Gallagher, William ; Dalong, Gao ; Ueda, Jun

Necessary physical contact between an operator and a force feedback haptic device creates a coupled system consisting of human and machine. This contact, combined with the natural human tendency to increase arm stiffness to attempt to stabilize its motion, can reduce the stability of the system. This paper proposes a method to increase stability on demand while maintaining speed and performance. Operator arm stiffness is not directly measurable, so controllers cannot typically account for this issue. The causes of arm end-point stiffness are examined as related to system stability, and a method for estimating changes in arm stiffness based on arm muscle activity was designed to provide a robotic controller with additional information about the operator. This was accomplished using electromyograms (EMGs) to measure muscle activities and estimating the level of arm stiffness, which was used to adjust the dynamic characteristics of an impedance controller. To support this design, the correlation between EMGs and arm stiffness was validated experimentally. Further experiments characterized the effects of the designed system on operator performance. This showed increased stability and faster, more accurate movements using the compensating system. Such a system could be used in many applications, including force assisting devices in industrial facilities.
A Force and Displacement Self-Sensing Piezoelectric MRI-Compatible Tweezer End Effector with an On Site Calibration Procedure

(Georgia Institute of Technology, 2014-04) McPherson, Timothy ; Ueda, Jun

This paper describes a self-sensing technique for a piezoelectrically driven magnetic resonance imaging (MRI)-compatible tweezer style end effector, suitable for robot assisted MRI guided surgery. Nested strain amplification mechanisms are used to amplify the displacement of the piezo actuators to practical levels for robotics. By using a hysteretic piezoelectric model and a two port network model for the compliant nested strain amplifiers, it is shown that force and displacement at the tweezer tip can be estimated if the input voltage and charge are measured. One piezo unit is used simultaneously as a sensor and an actuator, preserving the full actuation capability of the device. An on-site calibration procedure is proposed that calibrates the combined electromechanical model without requiring specific loading conditions on the inner piezoelectric actuators. Experimental validation shows an average of 12% error between the self-sensed and true values.
Nested piezoelectric cellular actuators for a biologically inspired camera positioning mechanism

(Georgia Institute of Technology, 2013-10) Schultz, Joshua ; Ueda, Jun

Using successive stages, or nesting compliant amplification mechanisms, soft actuators with performance suitable for robotic applications can be constructed with piezoelectric ceramic as the active material. This paper presents a mathematical framework that describes the interactions among the various amplification mechanisms in a hierarchical nested structure. A formal treatment of nested amplification mechanisms results in two theorems that describe the stiffness properties of the whole actuator in terms of the properties of each mechanism in the hierarchy. These theorems show that the stiffness properties of the actuator can be computed by considering only the outermost few layers in the nested configuration. By virtue of this hierarchical structure, the actuator also assumes a cellular structure; it functions by summing the effects of on-off inputs coupled by a flexible connective medium. This requires a paradigm shift when selecting control strategies. A multilayer strain amplification mechanism is designed to meet the required range of travel for a biologically inspired camera positioning mechanism, and a switching control method for the actuator's 16 on-off inputs is discussed.
Relaxed Individual Control of Skeletal Muscle Forces via Physical Human-robot Interaction

(Georgia Institute of Technology, 2013-06) Gallagher, William ; Ding, Ming ; Ueda, Jun

This study develops a relaxed formulation of a method for controlling individual muscle forces using exoskeleton robots. Past studies have developed a muscle-force control method with very strict limitations on the conditions. These conditions will be removed, and the problem will be reformulated as a constrained optimization of several parameters. The optimization algorithm recognizes when a solution to the muscle control problem cannot be exactly realized, and finds the solution that minimizes the mean errors of the individual muscles between expected and desired muscle activation. This is demonstrated in a computer simulation of human arm dynamics and compared against the prior method to demonstrate its wider applicability. In addition, the control method is extended to resolve issues associated with a nonideal exoskeleton with incomplete torque application to the joints. A quasi-optimized motor-task that minimizes the errors in target muscles and nontarget muscles can be obtained. This paper presents theoretical analysis, simulation, and experimental results on the performance of the relaxed individual muscle control.