Dynamic Modeling of an Unmanned Helicopter From Flight Test Data

Accurately modeling the dynamic characteristics of a helicopter is difficult and time-consuming. This paper presents a new identification approach which applies the modes partition method and structure traversal (MPM/ST) algorithm. The dynamic modes, instead of model parameters of each model structure, are sequentially identified through MPM. The model with the minimum cost function (CF) is chosen from the best model set and is defined as the final model. Real flight tests of an unmanned helicopter are carried out to verify the identification approach. Time- and frequency-domain results of the identified models clearly demonstrate the potential of MPM/ST in modeling such complex systems.

Mobile Agent-Based Computing Resource and Usage Monitoring at Large Scale Computer Centers

Mobile agent technology has become an important approach for the design and development of distributed systems. However, there is little research regarding the monitoring of computer resources and usage at large scale distributed computer centers. This paper presents a mobile agent-based system called the Mobile Agent Based Computer Monitoring System (MABCMS) that supports the dynamic sending and executing of control command, dynamic data exchange, and dynamic deployment of mobile code in C/C++. Based on the Mobile-C library, agents can call low level functions in binary dynamic or static libraries, and thus can monitor computer resources and usage conveniently and efficiently. Two experimental applications have been designed using the MABCMS. The experiments were conducted in a university computer center with hundreds of computer workstations and 15 server machines. The first experiment uses the MABCMS to detect improper usage of the computer workstations, such as playing computer games. The second experimental application uses the MABCMS to detect system resources such as available hard disk space. The experiments show that the mobile agent based monitoring system is an effective method for detecting and interacting with students playing computer games and a practical way to monitor computer resources in large scale distributed computer centers.

Controlling Modular Reconfigurable Robots With Handheld Smart Devices

Highly reconfigurable modular robots face unique teleoperation challenges due to their geometry, configurability, high number of degrees of freedom and complexity. Current methodology for controlling reconfigurable modular robots typically use gait tables to control the modules. Gait tables are static data structures and do not readily support realtime teleoperation. Teleoperation techniques for traditional wheeled, flying, or submerged robots typically use a set of joysticks to control the robots. However, these traditional methods of robot teleoperation are not suitable for reconfigurable modular robotic systems which may have dozens of controllable degrees of freedom. This research shows that modern cell phones serve as highly effective control platforms for modular robots because of their programmability, flexibility, wireless communication capabilities, and increased processing power. As a result of this research, a versatile Graphical User Interface, a set of libraries and tools have been developed which even a novice robotics enthusiast can use to easily program their mobile phones to control their hobby project. These libraries will be beneficial in any situation where it is effective for the operator to use an off-the-shelf, relatively inexpensive, hand-held mobile phone as a remote controller rather than a considerably heavy and bulky remote controllers which are popular today. Several usage examples and experiments are presented which demonstrate the controller’s ability to effectively control a modular robot to perform a series of complex gaits and poses, as well as navigating a module through an obstacle course.

A Plug and Play Middleware for Sensory Modules, Actuation Platforms and Task Descriptions in Robotic Manipulation Platforms

We are developing a framework (RISCWare) for the modular design and integration of sensory modules, actuation platforms, and task descriptions that will be implemented as a tool to reduce efforts in designing and utilizing robotic platforms. The framework is used to customize robotic platforms by simply defining the available sensing devices, actuation platforms, and required tasks. The main purpose for designing this framework is to reduce the time and complexity of the development of robotic software and maintenance costs, and to improve code and component reusability. Usage of the proposed framework prevents the need to redesign or rewrite algorithms or applications due to changes in the robot’s platform, operating systems, or the introduction of new functionalities. In this paper, the RISCWare framework is developed and described. RISCWare is a robotic middleware used for the integration of heterogeneous robotic components. RISCWare consists of three modules. The first module is the sensory module, which represents sensors that collect information about the remote or local environment. The platform module defines the robotic platforms and actuation methods. The last module is the task-description module, which defines the tasks and applications that the platforms will perform such as teleoperation, navigation, obstacle avoidance, manipulation, 3-D reconstruction, and map building. The plug-and-play approach is one of the key features of RISCWare, which allows auto-detection and auto-reconfiguration of the attached standardized components (hardware and software) according to current system configurations. These components can be dynamically available or unavailable. Dynamic reconfiguration provides the facility to modify a system during its execution and can be used to apply patches and updates, to implement adaptive systems, or to support third-party modules. This automatic detection and reconfiguration of devices and driver software makes it easier and more efficient for end users to add and use new devices and software applications. In addition, the software components should be written in a flexible way to get better usage of the hardware resource and also they should be easy to install/uninstall. Several experiments, performed on the RISCbot II mobile manipulation platform, are described and implemented to evaluate the RISCWare framework with respect to applicability and resource utilization.

Prediction Models for the Estimation of Soil Moisture Content

This paper provides a survey on different kinds of prediction models developed for the estimation of soil moisture content of an area, using empirical information including meteorological and remotely sensed data. The different models employed extend over a wide range of machine learning techniques starting from Basic Linear Regression models through models based on Bayesian framework, Decision tree learning and Recursive partitioning, to the modern non-linear statistical data modeling tools like Artificial Neural Networks. The fundamental mathematical backgrounds, pros and cons, prediction results and efficiencies of all the models are discussed.

An In-Pipe Leak Detection Sensor: Sensing Capabilities and Evaluation

In most cases the deleterious effects associated with the occurrence of leak may present serious problems and therefore leaks must be quickly detected, located and repaired. The problem of leakage becomes even more serious when it is concerned with the vital supply of fresh water to the community. In addition to waste of resources, contaminants may infiltrate into the water supply. The possibility of environmental health disasters due to delay in detection of water pipeline leaks has spurred research into the development of methods for pipeline leak and contamination detection. Leaks in water pipes create acoustic emissions, which can be sensed to identify and localize leaks. Leak noise correlators and listening devices have been reported in the literature as successful approaches to leak detection but they have practical limitations in terms of cost, sensitivity, reliability and scalability. To overcome those limitations the development of an in-pipe traveling leak detection system is proposed. The development of such a system requires a clear understanding of acoustic signals generated from leaks and the study of the variation of those signals with different pipe loading conditions, leak sizes and surrounding media. This paper discusses those signals and evaluates the merits of an in-pipe-floating sensor.

Digital Fractional Order Savitzky-Golay Differentiator and Its Application

Firstly the one-dimension digital fractional order Savitzky-Golay differentiator (1-D DFOSGD), which generalizes the Savitzky-Golay filter from the integer order to the fractional order, is proposed to estimate the fractional order derivative of the noisy signal. The polynomial least square fitting technology and the Riemann-Liouville fractional order derivative definition are used to ensure robust and accuracy. Experiments demonstrate that 1-D DFOSGD can estimate the fractional order derivatives of both ideal signal and noisy signal accurately. Secondly, the two-dimension DFOSGD is obtained from 1-D DFOSGD by defining a group of direction operators, and a new image enhancing method based on 2-D DFOSGD is presented. Experiments demonstrate that 2-D DFOSGD has very good performance on image enhancement.

Optimal Error-Clamping Design for Position-Based Visual Servoing of a 2-DOF Model Helicopter

A vision-based servoing technique is proposed for a 2 degrees-of-freedom (dof) model helicopter equipped with a monocular vision system. In general, these techniques can be categorized as image- and position-based, where the task error is defined in the image plane in the former and in the physical space in the latter. The 2-dof model helicopter requires a configuration-dependent feed-forward control to compensate for gravitational forces when servoing on a ground target. Therefore, a position-based visual servoing deems more appropriate for precision control. Image information collected from a ground object, with known geometry a priori, is used to calculate the desired pose of the camera and correspondingly the desired joint angles of the model helicopter. To assure a smooth servoing, the task error is parameterized, using the information obtained from the linearaized image Jacobian, and time scaled to form a moving reference trajectory. At the higher level, a Linear Quadratic Regulator (LQR), augmented with a feed-forward term and an integrator, is used to track this trajectory. The discretization of the reference trajectory is achieved by an error-clamping strategy for optimal performance. The proposed technique was tested on a 2-dof model helicopter capable of pitch and yaw maneuvers carrying a light-weight off-the-shelf video camera. The test results show that the optimized controller can servo the model helicopter to a hovering pose for an image acquisition rate of as low as 2 frames per second.

A Method for Simultaneously Satisfying Important Constraints and Dependencies for Many Different Types of Processes in Embedded Real-Time Systems

In most embedded, real-time applications, processes need to satisfy various important constraints and dependencies, such as release times, offsets, precedence relations, and exclusion relations. Embedded, real-time systems with high assurance requirements often must execute many different types of processes with such constraints and dependencies. Some of the processes may be periodic and some of them may be asynchronous. Some of the processes may have hard deadlines and some of them may have soft deadlines. For some of the processes, especially the hard real-time processes, complete knowledge about their characteristics can and must be acquired before run-time. For other processes, prior knowledge of their worst case computation time and their data requirements may not be available. It is important for many embedded real-time systems to be able to simultaneously satisfy as many important constraints and dependencies as possible for as many different types of processes as possible. In this paper, we discuss what types of important constraints and dependencies can be satisfied among what types of processes. We also present a method which guarantees that, for every process, no matter whether it is periodic or asynchronous, and no matter whether it has a hard deadline or a soft deadline, as long as the characteristics of that process are known before run-time, then that process will be guaranteed to be completed before predetermined time limits, while simultaneously satisfying many important constraints and dependencies with other processes.

Ground Antenna Scheduling Algorithm for Multi-Satellite Tracking

In this paper, we address the optimal ground antenna scheduling problem for multiple satellites when multiple satellites have visibility conflicts at a ground station. Visibility conflict occurs when multiple satellites have either overlapping visibilities at a ground station or difference with time of loss of signal (LOS) of a satellite and time of acquisition of signal (AOS) of another satellite is less than reconfiguration time of ground station. Each satellite has a priority value that is a weight function with various factors. Multi-antenna scheduling (MAS) algorithm 1 and Multi-antenna scheduling (MAS) algorithm 2 are proposed to find the optimal schedule of multi-antenna at a ground station using pre-assigned priority values of satellites. We use the depth first search (DFS) method to search the optimal schedule in MAS algorithm 1 and MAS algorithm 2. Through the simulations, we confirm the efficiency of these algorithms by comparing with greedy algorithm.