Officially admitted by the Office of then Philippine Supreme Court Chief Justice Reynato Puno (summit convener) for use during the National Consultative Summit on Extrajudicial Killings and Enforced Disappearances - Searching for... more
Officially admitted by the Office of then Philippine Supreme Court Chief Justice Reynato Puno (summit convener) for use during the National Consultative Summit on Extrajudicial Killings and Enforced Disappearances - Searching for Solutions, Manila, Philippines, 16-17 July 2007.
The state of practice in developing embedded infrastructure software uses component-based architectures such as the AUTOSAR initiative (www.autosar.org), along with static code analysis tools to capture design flaws. However, these still... more
The state of practice in developing embedded infrastructure software uses component-based architectures such as the AUTOSAR initiative (www.autosar.org), along with static code analysis tools to capture design flaws. However, these still fail to address the dynamic and real-time aspect of the infrastructure software. Fortunately, current advances in MBD technology indicate that embedded-software developers can expect more tool support for the whole ECU software spectrum. We report in this article an overview on the X-in-the-loop V&V process and draw some lessons learned from related industrial experience.
Importance of data has increased dramatically in recent times. Companies are using the online data available to improve their products and data analysis. Traffic data of vehicles on road can be of great importance if they can be... more
Importance of data has increased dramatically in recent times. Companies are using the online data available to improve their products and data analysis. Traffic data of vehicles on road can be of great importance if they can be effectively retrieved from the videos. Contrary to normal data traffic data is the form of videos as a result immense amount of data is found in surveillance videos which goes unprocessed thus resulting in loss of information. Our project aims to provide an effective way to retrieve vehicle data from videos. This data includes the vehicle's license plate number, colour, manufacturer's name, and the time-stamp. This data can then be used for various purposes. This data will help to analyse the traffic on road at a much deeper level. This analysis will help to gain insights which will help improve the future of road transportation. Also, since the data is in a text form it will be much easier and faster to search or filter data. If required this filtered data can be correlated with video data. In the above process without the text data, one would have to scrub through the video files to reach a required point. Also, the data which is stored in the text form is of much less size than video files. Implementation of our project will help to create a smart traffic system. We retrieve the required information one at time. Firstly, the vehicle is detected with the help of YOLO (You Only Look Once). Then from the images of each vehicle required data is extracted. Firstly, the license plate of the vehicle is detected. With some pre-processing applied the license plate image is sent for OCR (Optical Character Recognition), which retrieves the text on the license plate. Then the vehicle's colour is extracted. Efforts are being made to remove environmental interference on the vehicle, so that the genuine colour of the vehicle is obtained. Also, the vehicles manufacturer is identified. As one can notice all the listed features (license plate, colour, manufacturer) are the primary attributes to identify a vehicle. Combinations of these attributes can be used to recover other required data about the vehicle. If speed detection of vehicles is done in an effective way, it will be possible to reduce speed violations. Thus, making the roads a much safer place.
Low speed estimation in DTC IMD is not accurate due to the presence of transient offset, drift and domination of ohmic voltage drop in the measured stator voltages and currents used for estimating the stator flux required for accurate... more
Low speed estimation in DTC IMD is not accurate due to the presence of transient offset, drift and domination of ohmic voltage drop in the measured stator voltages and currents used for estimating the stator flux required for accurate estimation of speed. EKF is a nonlinear, recursive adaptive algorithm capable of estimating speed ranging from very low speed to rated speed using equation of motion from noisy measured currents and voltages based on state space technique. In the previous work a new state space model of IM was developed for estimation in EKF by feeding load torque profile as an input variable instead of estimating it by considering load torque as constant, validated using MATLAB-Simulink software. In this paper real time validation of the EKF controller with load profile fed as input for speed estimation in DTC IMD is carried out using OPAL-RT simulator and real time results validates the simulation results and proves the effectives of the new EKF for low speed estimation in DTC IMD.