This paper presents a GPU implementation of two foreground object segmentation algorithms: Gaussian Mixture Model (GMM) and Pixel Based Adaptive Segmenter (PBAS) modified for RGB-D data support. The simultaneous use of colour (RGB) and depth (D) data allows to improve segmentation accuracy, especially in case of colour camouflage, illumination changes and occurrence of shadows. Three GPUs were used to accelerate calculations: embedded NVIDIA Jetson TX2 (Maxwell architecture), mobile NVIDIA GeForce GTX 1050m (Pascal architecture) and efficient NVIDIA RTX 2070 (Turing architecture). Segmentation accuracy comparable to previously published works was obtained. Moreover, the use of a GPU platform allowed to get real-time image processing. In addition, the system has been adapted to work with two RGB-D sensors: RealSense D415 and D435 from Intel.
In this paper some parameterizations of controls are examined in a Lie algebraic method of motion planning for driftless nonholonomic systems. The purpose of the examination is to establish how numerous the parameterization should be and which items of a harmonic basis are to be included into the parameterization. An algorithm is presented to evaluate parameterizations without (or reduced) impact of a local, desired direction of motion.
This paper presents a decentralized control approach based on a neural network for a hydrostatic transmission. The bent-axis angle of the hydraulic motor is adjusted by a pure feedforward control law based on identified physical parameters, whereas the corresponding motor angular velocity is controlled using a combination of a generalized proportional-derivative (PD) controller and a multilayer perceptron with one hidden layer that is trained by an error-feedback learning approach and uses only measurable input variables. In this observer-free control structure, the neural network learns the inverse dynamics by minimizing the PD controller output and, as a consequence, an accurate tracking of the desired trajectory is achieved. As no physical modelling is required for the motor velocity control design, it can be considered as model-free. The tracking performance shows the robustness of the overall control structure for the hydrostatic transmission despite disturbances and uncertainties. The proposed control scheme is investigated by simulations first. Second, experimental results are presented taken from a dedicated test rig at the Chair of Mechatronics, University of Rostock. Finally, an experimental comparison with results from previous work is provided.
This paper discusses the process of balancing the satellite simulator mounted on the spherical air bearing table. In order to accurately simulate the satellite motion with such a test stand it is necessary to bring the Center Of Mass (COM) of the system as close as possible to the Center Of Rotation (COR) of the air bearing by moving the balancing masses. This calibration process reduces the gravity torque influencing the system movements. A new batch method of determining the COM of the balancing platform is proposed, allowing for its later adjustment. The novelty of the method comes from the idea that the freely rotating system model can be divided into rigid part for which the center of mass is constant, and the balancing masses constituting the variable influence on the COM. Advantage of this approach is the fact, that while gathering data for the batch calibration counterweights can be actuated in a known way, which turns out to greatly improve the estimation precision. Potential disadvantage is, that estimate of the masses and paths of movement of the counterweights is required, which in practice constitutes additional sources of error. Sensitivity analysis is performed to asses the viability of this trade-of considering the inaccuracies in the balance masses paths, as well as sensor noises and misalignment.
This work presents an original software and hardware system whose objective is to detect pressure leaks. Two methods for detection of leaks are considered: the first one is based on an industrial vision system, the second one on a proprietary ultrasonic sensor using Fast Fourier Transformation (FFT). Automation of the measuring process has been done by an industrial six axis robotic arm. For experiments three original laboratory stands have been used.
The investigated method enables precise shaping of acoustic radiation of a vibrating plate, i.e. it allows one to relocate or create resonances and anti-resonances for selected frequencies, simultaneously altering their acoustic radiation efficiency in a desired manner. The method can be very beneficial for plates used as noise barriers, both in passive and active applications. The acoustic radiation shaping method involves mounting several additional ribs and masses to the plate surface at locations followed from an optimization process (sensors and actuators can also be included, if active control is considered). The optimization process requires a model of the vibroacoustic system, a cost function corresponding to the considered objective, and an optimization algorithm. In this paper, an introduction of A-weighting to the cost functions, which reflects a human perception of the noise radiated by or transmitted through the plate, is investigated. It follows from the analysis of obtained results that the introduction of A-weighting can provide even 8 dBA better passive noise attenuation.
Self-Balancing Electric Motorcycle (SBEM) is a dynamic and nonlinear electromechanical system. In this paper, the process of mathematical modelling and linearization of SBEM is presented. The model of the control system in Matlab environment is implemented. The control system using the PID controller is designed. The operation of particular structures of the PID controller on the simulation model is compared. Due to simulation research, the most appropriate structure and parameters of the PID controller are chosen.
The paper assumes the composition of robotic systems out of embodied agents. It presents a utilitarian decomposition of an agent into subsystems. Both subsystem behaviours and their selection can be described in terms of Finite State Automatons (FSA), thus Hierarchic FSAs result. Mathematical formalisation of this description enables the verification of correctness of some aspects of system operation.
In this paper, an enhanced Artificial Potential Field (AFP) algorithm applied to autonomous mobile robot is presented. The proposed solution is extended by an additional BUG algorithm and a ground quality indicator. The modification allows to avoid local minima in path planning caused by complex terrain obstacles. The developed algorithm takes into account the substrate quality, classifying a poor ground as an obstacle. It was implemented and tested in Matlab software, utilizing the track planning algorithm as a state machine. The simulation environment enables graphical presentation of the chosen path and the arrangement of moving area. The article presents and discusses a basic issue in local path planning of autonomous mobile platforms, i.e. the ground quality, which is ignored in classical algorithms. This is a non-trivial problem, which impacts the success rate of getting the final destination. The developed algorithm is extended by a BUG rule and ground quality indicator which allows to avoid immobilization of the platform.
This paper presents an implementation and tests of the hardware in the loop control idea using Scilab-Xcos software – an open source alternative for Matlab-Simulink simulation environment and modular drive system dedicated to brushless motors. Communication infrastructure was based on USB interface and special protocol between hardware and Scilab-Xcos. For the test of the hardware in the loop control, the Field Oriented Control was implemented within Scilab-Xcos. Thanks to this approach, control algorithms do not have to be implemented directly on the device, but in an environment that ensures convenient operation, including quick tests with variable design structure. Software can be tested on both Windows and Linux. Using Linux OS makes this solution pure open source.
In this paper, a problem of autonomous ship utility model identification for control purposes is considered. In particular, the problem is formulated in terms of model parameter estimation (one-step-ahead prediction). This is a complex task due to lack of measurements of the parameter values, their time-variability and structural uncertainty introduced by the available models. In this work, authors consider and compare two utility models based on often utilised ship model structures with time-varying parameters identified recursively using the extended Kalman lter (EKF). The validation results have been obtained using simulation experiments in which the required information for the parameter estimation task had been generated using a cognitive model of B-481 ship. The results indicate the benefits and drawbacks, in terms of estimation accuracy and computational complexity, of using each of the investigated utility model structures.
