Search

Recent human-in-the-loop (HIL) optimization studies using wearable devices have shown an improved average metabolic reduction by optimizing a small number of control parameters during short-duration walking experiments. However, the slow metabolic dynamics, high measurement noise, and experimental time constraints create challenges for increasing the number of control parameters to be optimized. Prior work applying gradient descent and Bayesian optimization to this problem have decoupled metabolic estimation and control parameter selection using fixed estimation intervals, which imposes a hard limit on the number of parameter evaluations possible in a given time budget. In this work, we take a different approach that couples estimation and parameter selection, allowing the algorithm to spend less time on refining the metabolic estimates for parameters that are unlikely to improve performance over the best observed values. Our approach uses a Kalman filter-based metabolic estimator to formulate an optimal stopping problem during the data acquisition step of standard Bayesian optimization. Performance was analyzed in numerical simulations and in pilot human subject testing with two subjects that involved optimizing six control parameters of a single-joint exosuit and four parameters of a multi-joint exosuit.

In recent years, many studies examined if EEG signals from traumatic brain injury (TBI) patients can be used for new rehabilitation technologies, such as BCI systems. However, extraction of the high-gamma band related to movement remains challenging due to the presence of surface electromyogram (sEMG) caused by unconscious facial and head movement of patients. In this paper, we proposed a modified independent component analysis (ICA) model for EMG artifact removal in the EEG data from TBI patients with a hemicraniectomy. Here, simulated EMG was generated and added to the raw EEG data as the extra channels for independent components calculation. After running ICA, the independent components (ICs) related to artifacts were identified and rejected automatically through several criteria. EEG data underlying hand movement from one healthy subject and one TBI patient with a hemicraniectomy were conducted to verify the efficacy of this algorithm. Results showed that the proposed algorithm removed sEMG artifacts from the EEG data by up to 86.72% while preserving the associated brain features. In particular, the high-gamma band (80 to 160 Hz) was found to arise principally from the hemicraniectomy area after this technique was applied. Meanwhile, we found that the magnitude of gamma power during movement improved after removal of sEMG artifacts.

This paper provides an experimental analysis on the effect of various ground configurations on circuit current return paths. It provides an in-depth investigation of corrupted voltage readings in analog-to-digital converters (ADCs), which leads to inaccurate sensor readings and may cause system instability in embedded applications. In the ever-expanding field of robotics the electric motor is a fundamental component. Wheeled, legged, flying, swimming, stationary, swarming, and other types of robots require some sort of actuator to provide motion to the specific system part, or of the entire system. This analysis examines DC and servo motors noise produced in readings obtained from selected voltage, temperature, and distance analog sensors. PSoC-based CY8CKIT-001 Development Board and PSoC-based EagleSoC Development Board demonstrate how the separation of ground lines in the mixed-signal PCB circuit ensures accurate analog sensor readings.

This paper introduces a semi-autonomous robotic platform for future planetary exploration. The Microcontroller Unit (MCU) of this platform is reconfigurable to support varied demands of scientific applications using Programmable System on Chip (PSoC) technology. The PSoC acquires and processes sensor data, then relays the data to a remote control station. The platform supports three modes of operations: (1) anytime user remote-control mode, (2) limited time autonomous control mode, and (3) mixed-control mode, which assigns certain tasks to run in automated execution mode, while reserving critical or complex tasks for user-controlled execution. The real-time task scheduling for precise control is supported by the Micrium μC/Ο8-ΙΠ realtime operating system (RTOS). The use of such an RTOS also allows interleaving of data acquisition for enhancing effective bandwidth.

Previously, Hybrid-Routing Algorithm Model was proposed by the Structures, Pointing, and Control Engineering (SPACE) University Research Center 1 at California State University of Los Angeles (CSULA). The Hybrid-Routing Algorithm Model utilizes semi-autonomous control during robotic operations in unknown, remote, hazardous, or otherwise inaccessible terrains. In order to observe and monitor mobile robot's surroundings, an onboard virtual reality system has been designed and implemented. It operates by remotely following the commanded movement of the robotic mission operator's headset. The proposed design features the Inertial Measurement Unit (IMU) chip secured to the center of the iTV-goggles unit., which is worn by the robotic mission operator. Direction cosine matrix algorithm is used to update the body coordinates of the IMU chip to the global reference system. Euler angles are computed based on the angular orientation of the mission operator's headset and are transmitted to the mobile robot. The observation system contains multiple servo motors that rotate onboard camera based on the received pitch roll, and yaw information. The observation system transmits compressed visual information of robot's surroundings to the Host Computing Station. This video is rendered via iTV-goggles display to the mission operator.

This paper focuses on the implementation and quantitative analysis of a high-performance parallel processing aerospace information server. An innovative model of software architecture is provided to effectively utilize the computational power of a parallel server platform for efficient, on-demand aerospace information exchange through the Internet. This is a representative application for servers whose features are common to the classical client-server model. The server architecture supports thread, core, and/or processor-level parallel processing for high performance computing. Memory devices (i.e. cache memory, main memory, and secondary memory) are either shared or distributed among the computational units. Such features facilitate our study of identifying and overcoming the architectural bottlenecks of current commercial server configurations.

Control of complex, flexible structures requires substantial amounts of computational power to achieve precision performance in both space and time. This is due to the fact that such structures are inherently multiple input, multiple output systems whose complexities increase significantly with each additional input or output parameter. The other design difficulty is due to the requirement of realtime computation and data communication since such systems have to be controlled on the fly. Thus, general purpose computer architectures are simply insufficient in this scenario. This paper introduces the development of an embedded computing system that supports the implementation of control algorithms on a segmented reflector telescope testbed. Decentralized algorithms have been recognized and proven to provide promising performance of control, and are being used for primary mirror shaping and precision pointing control of this testbed. The system architecture of the testbed is featured with a suite of interconnected signal processors and high performance I/O devices to meet the computational requirements, and hence, allows real-time performance of the control algorithms. The system also supports fault-tolerance during processor failure or recovery by leveraging the technologies of decentralized control with its associative pipelined task mapping mechanism

This paper outlines the ongoing research at the Structure, Pointing and Control Engineering (SPACE) Laboratory to achieve a decentralized control of a segmented reflector testbed. The main effort is concentrated on achieving the system's top-level requirements including figure maintenance of the primary mirror to within 1 micron RMS distortion with respect to a nominal shape of the primary mirror, and precision pointing with accuracy of 2 arc seconds using the decentralized control technique. Having the time realization of the system at hand permits the implementation of such strict control requirements. Therefore, this entails system identification methods to achieve such a realization. In addition, by capitalizing on the characteristics of decentralized control, a fault-tolerant pipelined parallel processing design is developed. This approach features both improved load balancing for any number of processors and tasks and allows recovery from one or more processor failures.

The next generation space telescope (NGST) is expected to produce a vast amount of raw images, which need to be stored and disseminated efficiently for astronomical research and education. Therefore, effective compression and transmission methods are very important. This paper proposes a joint compression and transmission scheme for efficient streaming of astronomical images based on their special characteristics. The objective of the joint approach is to maximize the recognition of the astronomical entities presented in the image under the same rate constraint and channel loss behavior. In the compression stage, each astronomical entities is segmented and coded independently using zero-tree based wavelet codec, and robust packetization scheme is employed to prevent error propagation; in the transmission stage, content information of the image is utilized to maximize the perceptual quality of the entities at the presence of packet loss. Experiment results are presented to demonstrate the effectiveness of the proposed approach.