Micro Aerial Vehicles (MAVs) have gained a significant amount of research lately, with a number of universities and industry sponsors paving the way with micro flying robots to perform Intelligence, Surveillance and Reconnaissance (ISR) Missions. However, much of the work done in flapping wing MAVs till date has not shown performance improvements over their Vertical Take Off and Landing (VTOL), rotary-wing counterparts. Research and development over the years has shown that insects and birds have unmatched flying capabilities in the Low Reynolds Number Regime. Their phenomenal flight performance is attributed to among others the highly energy efficient actuation systems and their power to weight ratios. The paper proposes and illustrates new and energy-efficient actuation mechanisms and incorporating them into MAVs. The paper begins by elaborating on the background of our research and the need for energy efficiency in MAVs; followed by showcasing resonant, elastic/restorative actuation mechanisms and their control methodology to reduce flight-actuation energy requirements. In conclusion, the paper contrasts and compares all the actuation systems for power intake, scalability and mechanical complexity.

Multi-agent prognostic health and usage monitoring (Multi-PHUM) is proposed in the paper for use in fault diagnosis and prognosis. Georgia Tech has developed a Matlab simulation model of monopropellant propulsion for studying fault diagnosis. Here we apply Multi-PHUM to the fault diagnosis of the monopropellant propulsion system. In particular, regulator failure is discussed here.

This paper describes an application of one-class support vector machine (SVM) novelty detection for detecting seizures in humans. Our technique maps intracranial electroencephalogram (EEG) time series into corresponding novelty sequences by classifying short-time, energy-based statistics computed from one-second windows of data. We train a classifier on epochs of interictal (normal) EEG. During ictal (seizure) epochs of EEG, seizure activity induces distributional changes in feature space that increase the empirical outlier fraction. A hypothesis test determines when the parameter change differs significantly from its nominal value, signaling a seizure detection event. Outputs are gated in a .one-shot. manner using persistence to reduce the false alarm rate of the system. The detector was validated using leave-one-out cross-validation (LOO-CV) on a sample of 41 interictal and 29 ictal epochs, and achieved 97.1% sensitivity, a mean detection latency of -7.58 seconds, and an asymptotic false positive rate (FPR) of 1.56 false positives per hour (Fp/hr). These results are better than those obtained from a novelty detection technique based on Mahalanobis distance outlier detection, and comparable to the performance of a supervised learning technique used in experimental implantable devices (Echauz et al., 2001). The novelty detection paradigm overcomes three significant limitations of competing methods: the need to collect seizure data, precisely mark seizure onset and offset times, and perform patient-specific parameter tuning for detector training.

Rolling element bearings are widely used components in industrial equipments and their failure may result in severe damage of critical processes. Bearing plants are required to provide estimated performance data to their customers. These statistically based data are usually got through various testing experiments with product samples via bearing test rig. However, it is usually hard to get convincible results from traditional method due to the limitation of monitoring device and analysis methods. This paper designs an intelligent system for a bearing testing and inspection center of a bearing plant to monitor the health status and assess the performance of bearings being tested. The system has a distributed infrastructure and can continually collect testing data, analyze vibration signals, extract features of bearing fault, diagnose bearing faults and further assess the quality of the bearings. Fuzzy logic, wavelet neural network and dynamic wavelet neural network are employed as diagnostic/prognostic algorithm.

A partial differential equation (PDE)-based feature-level image fusion approach is proposed for multisensory image segmentation. The energy functional of the proposed fusion model is a weighted sum of several functionals, each constructed based on the characteristics of the sensor image. The weight selection decides the way that the model handles redundant, conflicting, or complementary information involved in the multisensory data. The method is implemented using level sets and is fast enough for real-time segmentation tasks. Finally the algorithm is applied to the segmentation of X-ray and visual images, and the results show that the fusion algorithm is efficient, accurate, and robust.

A partial differential equation (PDE)-based feature-level image fusion approach is proposed for multisensory image segmentation. The energy functional of the proposed fusion model is a weighted sum of several functionals, each constructed based on the characteristics of the sensor image. The weight selection decides the way that the model handles redundant, conflicting, or complementary information involved in the multisensory data. The method is implemented using level sets and is fast enough for real-time segmentation tasks. Finally the algorithm is applied to the segmentation of x-ray and visual images, and the results show that the fusion algorithm is efficient, accurate, and robust.

The mode-to-mode transition problem involves taking initial states in the start mode to the equilibrium point of the goal mode, where each mode of operation corresponds to an operating regime about an equilibrium point. Like the problem of dynamic transitions between various equilibria, there is no consistent theory that deals with the mode-to-mode transition problem. This paper presents a method of designing mode-to-mode controllers by blending the start and goal mode controllers. The blending gains are determined by the phase portrait assignment algorithm. The phase portrait assignment algorithm is a systematic technique that uses dynamic programming and center-point cell mapping to design fuzzy logic controllers. A hover mode to forward flight mode controller for a small-scale helicopter is synthesized to illustrate the design methodology. Simulation results show that the controller is able to transition stably from hover to forward flight. Finally, sensitivity analysis of the hover to forward flight controller is performed for small parameter perturbations. © 2000 John Wiley & Sons, Inc.

We present a quantitative method for identifying the onset of epileptic seizures in the intracranial electroencephalogram (IEEG), a process which is usually done by expert visual inspection, often with variable results. We performed a fractal dimension (FD) analysis on IEEG recordings obtained from implanted depth and strip electrodes in patients with refractory mesial temporal lobe epilepsy (MTLE) during evaluation for epilepsy surgery. Results demonstrate a reproducible and quantifiable pattern that clearly discriminates the ictal (seizure) period from the pre-ictal (pre-seizure) period. This technique provides an efficient method for IEEG complexity characterization, which may be implemented in real time. Additionally, large volumes of IEEG data can be analyzed through compact records of FD values, achieving data compression on the order of one hundred fold. This technique is promising as a computational tool for determination of electrographic seizure onset in clinical applications.

A recurrent fuzzy neural network with external feedback, called the Dynamical-Adaptive Fuzzy Neural Network (D-AFNN), is proposed in this article for adapting discrete-time dynamical systems. The fuzzy model is based on the Takagi-Sugeno inference method with polynomial consequent functions. The D-AFNN model is trained to learn the system dynamics using a recursive training algorithm. Both the epochwise and the on-line version of the recursive algorithm are considered. The standard representation of a dynamical system, using the fuzzy recurrent model, is discussed and the computational procedure for the derivation of the output gradients is analytically described. The performance qualities of the D-AFNN model are illustrated with two examples. In the first example, the model is trained to identify a nonlinear plant, while in the second the fuzzy recurrent model implements an IIR adaptive filter for noise attenuation. © 1996 John Wiley & Sons, Inc.

In this paper, three analytical tasks, stability analysis, robust control, and fuzzy intelligent organizer for a moving base robot are introduced. These tools are necessary for a mobile robot to perform complex assembly or manufacturing tasks in an unstructured environment. A stability envelope based on the relations between the system's mechanical energy and the stability relation is derived. A robust control algorithm is applied to the moving base robot that accounts for the dynamics of the composite system. A fuzzy intelligent organizing control strategy, based on a fuzzy rulebase and derived from measured force/moment data, for a robotic device performing quasi-static assembly tasks is presented. A modified distance metric is employed to measure the uncertainty (fuzziness) as it is related to specific control actions. A learning algorithm based on the probability of a fuzzy event is introduced. The top organizing level determines the most appropriate n-tuple control values with minimum fuzziness and downlinks them to the lower level of the system which carries out the specified task. Experimental results show the effectiveness of the proposed approach. Finally a hierarchical controller which consists of a higher level fuzzy expert organizer coordinating with a lower level robust controller is proposed. Experimental and simulation results are used to demonstrate the effectiveness of the algorithms.>

The design concept is presented of a novel spherical stepper robotic wrist motor capable of three-degree-of-freedom (DOF) motion in a single joint for applications where high-speed manipulation of end-effector orientation is required continuously in all directions. Typical applications are plasma and laser cutting and microassembly. The spherical motor is developed based on the principle of the variable-reluctance stepper motor. The differences between the operation principle of the three-DOF spherical motor and that of the conventional stepper motor are pointed out. The authors seek to establish a theoretical basis for design, prototype development, and performance prediction. An analysis of torque prediction is discussed, along with the presentation of kinematic and dynamic relationships. A laboratory prototype hybrid digital/analog control circuitry has been developed to demonstrate proof of concept feasibility and to assist in achieving an optimum design.>