This article focuses on the sensor and actuator system of an autonomous indoor quadrotor helicopter. The sensor system has two parts: an inertial measurement unit (IMU) and a vision system. The fusion between the two systems is solved by extended Kalman filters. The calibration of the inertial measurement unit is described for various types of errors. The variance analysis is performed for the noise sources of sensors. The actuators of the helicopter are the four rotors. The identification and low level control of the brushless DC motor based rotor system is also presented. The embedded control system integrates a lot of processors and communication lines. The verification of the system's parts were performed under real-time conditions. The main results of the paper are the new calibration algorithm for the different sensors and the real-time realization of the complex sensory system that may be part of a formation control system of UAVs.

The paper deals with the quick prototype design and hardware-in-the-loop real-

No abstract

Power requirements and profitability of suckerrod pumping are basically determined by the torque load on the pumping unit's gearbox. Gearbox torques include the torque required to drive the polished rod and the torque used to rotate the counterweights. In addition to these, inertial torques arise in those parts of the pumping unit that turn at varying speeds. As shown in the paper, all torque components are functions of the crank angle; consequently, their exact calculation necessitates the knowledge of the crank angle versus time function. This circumstance, however, complicates torque calculations because contemporary dynamometers, used to acquire the necessary operating data, do not provide any information on the variation of the crank angle during the pumping cycle. The paper introduces a solution of the problem and presents an iterative calculation of the crank angle versus time function from dynamometer data. Based on this function crank velocity, crank acceleration, as well as beam acceleration can be calculated and all necessary gearbox torques can be evaluated. For calculating articulating inertial torque, the acceleration pattern of the walking beam during the pumping cycle is evaluated according to three different models and the accuracy of those is compared. The paper gives the details of the developed calculation models and presents a typical sample case.

This paper presents a method for GPS based orientation determination. The carrier phase GPS concept is used to achieve precise results. The typical problem of the integer ambiguity resolution is solved using additional magnetometer and accelerometer. For proper operation the magnetometer should be calibrated together with the moving vehicle. Two kinds of calibration methods are presented. The first can be used with aerial vehicles, the second with ground vehicles. The calibration methods are able to run online on embedded computers on the board of the vehicle. The calibration methods and the orientation determination are tested with real movements performed with a car and a sail-plane.