A bond graph model of a Skystream wind turbine is proposed. This system is composed by a permanent magnet synchronous generator (PMSG), an 3-phase recti er, a Boost converter and an inverter. In order to verify the PMSG bond graph model, a junction structure to get the mathematical representation of this particular non-linear system is presented. Simulation results of each bond graph models using 20-SIM software are described. The performance of the complete system considering the wind model and the aerodynamic part connected to the Skystream wind turbine is shown. KEY WORDS Bond graph, Skystream, PMSG, Recti er, Boost converter, Inverter.
Artificial intelligence based prediction models provide good air pollution forecasters, proper to real time forecasting systems. Among them, artificial neural networks are the most used ones, being universal approximators. Usually, the identification of the best neural model (i.e. most accurate one) is based on experiments and results of time series analysis. The paper focuses on time series analysis for particulate matter (PM) air pollutant prediction with artificial neural networks in the Ploiesti city. Two types of neural models were used: feed forward and radial basis function. For each model we have experimented several architectures in order to identify the most accurate one in terms of root mean square error and average square error. The experimental datasets include five time series with concentration measurements of five air pollutants, PM10, CO, NO2, NOx, and SO2 in the period 2008-2012 at PH-6 air quality monitoring station from the Ploiesti city.
A static model of cerebral blood circulation is constructed for developing an automatic control system of physiological brain state, which will be used in brain resuscitation. In this paper, the constructed model was partially verified in the simulation using the actual data of a 70 years old patient with stenosis at left internal carotid artery. In the simulation, each arterial blood flow in cerebral blood circulation were calculated and were compared with the actual data measured by MRI and SPECT. In addition, the distribution of blood pressure in the Circle of Willis and the change of blood pressure along blood flow were also calculated for verifying against physiological knowledge. These results indicated that this model can represent the arterial part of cerebral blood circulation of this patient. And, the results also indicate the possibility that a static model such as our model represents cerebral blood circulation sufficiently. Thus, the construction of our model is enough significant.
Faculty of Engineering Department of Electrical and Mechanical Engineering UNESP State University, Av. Luiz Edmundo Carrijo Coube, 14-01
In this paper, we propose an approach for abnormal event detection, using the object-level spatio-temporal representation. Our approach detects an abnormal event in complex scenes which contain objects classified in various categories. We compute the object-level 3D Region-of-interest (3D RoI) and extract object-level 3D volume. Then, the object-level 3D volume is inputted to 3D deep convolutional neural network (3D-DCNN) for detecting the abnormal event. In the experiments, we compare our method with several methods on our experimental dataset.
Causally dynamic hybrid bond graphs are generally considered unsuitable for simulation, and causality is therefore often constrained in hybrid bond graph models.This paper demonstrates how a causally dynamic model can be simulated, using a buck converter as a case study.A causally dynamic hybrid bond graph (utilizing controlled junctions) is used to derive a mixed-Boolean state equation.This state equation is transferred to MATLAB R , where a simple routine assigns values to the Boolean parameters and then solves the model.Where storage elements are in dynamic causality, the model takes descriptor state form and an implicit solver is used.Solver choice and event detection are discussed.MATLAB R was selected as an accessible environment which allows this type of model to be coded and solved, but the technique could be used in an environment of the practitioners choice.The power converter is successfully modeled with a fast simulation time, demonstrating that simulating causally dynamic hybrid bond graphs is possible and merits further refinement.
Recently, interest in a passive coherent location (PCL) system using commercial broadcasting signals has been increased since it exploits the existing broadcasting facilities without revealing the receiver’s location.
The paper presents the automatic control of aircraft in longitudinal plane, during landing, by using the linearized dynamics of aircraft, taking into consideration the sensor errors and other external disturbances.
The aviation industry is motivated to develop and validate new aircraft models for the prediction of engine performance. These models are used in the preliminary aircraft design in order to predict its engines performance. The purpose of this study is to design an accurate model of the fan and compressor engine components. This model will then be integrated in a full engine model based on a component modeling approach. Several methods already exist to model compressing components. Among them, the stage-stacking method is used in this paper. This method can be used to predict the compressor performance but also its deterioration (ex: “fouling”). In both cases, the principle is the same. Each stage is separated, and the first stage outputs are used for the next stage as inputs until the last stage is reached, when the final outputs are acquired. A Cessna Citation X Level D Research Flight Simulator designed by CAE Inc. is used to sample the data needed to identify and validate the engine models elaborated for the whole flight envelope. Level D is the highest level of certification given by the FAA for the flight dynamics. Thus the simulator is used as real aircraft flight dynamics data. Different flight tests were performed to mesh the flight envelope for different flight conditions (Mach numbers from 0 to 0.92; altitude from 0 to 50000ft and Throttle Lever Angle,TLA, from “idle” to “max” in degrees). Nonetheless, the stage stacking method needs information which is not always available, such as the blade angle. A “grey box” approach was chosen. The unmeasurable parameters were identified in order to tune the model, and thus to reduce the global error between the model and the simulation data. A “black box” approach was also with an optimisation algorithm to approximate the outputs as polynomial functions of the inputs. In both cases the results were found to be accurate.
Normally industrial painting robots have an articulated structure and have six degrees of freedom (DOF). This project investigated a 3 DOF robot for painting vertical objects of different colors. The proposed project utilized a SCARA type robot which has stiffness in the vertical direction and the largest workspace in industrial applications. The gripper of the robot was replaced with an end-effector which was designed to hold a color sensor and an automatic spray gun. The color sensor detects the color to be painted from a sticker pasted on the corner of the object and gives a signal to the Programmable Logic Controller (PLC) for selecting the proper container in the tank reservoir. A painting reservoir, consists of different painting tanks, was attached to the robot as a mobile unit and an electronic switch was utilized to direct the cable to the selected color based on the signal from the color sensor. The equatıons of motıon for the SCARA robot is presented based on the Lagrange-Euler technique. The joints trajectory are designed as fifth-order polynomials to estimate the required torques and linear force for actuators. Performance analysis for accurate and homogeneous painting is provided.
A geographical name is a conventional language sign, which is not only used for a geographic entity, but also widely applied in the naming of other entities, thus significantly affecting the accuracy of news-specific geographical name extractions.
This paper presents a procedure to transform a chaotic logistic map into a continuous-time delay chaotic system by using sampled-data representation of continuous-time models. Because of this, the chaotic behavior of the resultant scheme is easy to proofread. A numerical illustration is also realized by utilizing Matlab/Simulink, where the new resultant chaotic attractor is shown
The condition number of a Sylvester matrix composed of the denominator and numerator polynomial of a model is investigated numerically for an LLC resonant dc-dc converter. The order of the model is changed from six to one in a certain manner and they are discretized using the step-invariant and matched-pole-zero methods expressed in the delta and shift operators. Their condition numbers show that in both operator forms, the higher the order is, the larger the condition number becomes, implying poorer numerical conditioning. It was also found that the stepinvariant model is more ill-conditioned than the matchedpole-zero models in general. While the number becomes constant as the sampling frequency increases for the delta cases, it keeps increasing for the shift cases, which is well known. However, in the range of sampling frequencies expected in recent practices, the models expressed in shiftform have smaller condition numbers and are betterconditioned than those in delta-form of respective orders.
A vibration suppression control problem by using on-off valves for a pneumatic isolation table is considered. In this system, positioning control is difficult because the system possesses a non-linear quantizer essentially. As a result, positioning error is remained in general. To overcome this problem, we have proposed a method considering the common multiple constraint, which is the appropriate relationship between the number of inhalation times and the number of exhalation times. By using this method, high convergence performance can be obtained without any offset. However, only vibration in the vertical direction is considered. Thus, we propose a method of suppressing the vibration of the two-axis direction by generalizing the previous method. The effectiveness of the control method is verified by numerical simulation.
