Deep Learning for Hardware-Based Real-Time Fault Detection and Localization of All Electric Ship MVDC Power System

Many investigators are interested in improving the control strategies of hand prosthesis to make it functional and more convenient to use. The most used control approach is based on the forearm muscles activities, named ‘ElectroMyoGraphic’ (EMG) signal. However, these biological signals are very sensitive to many disturbances and are generally unpredictable in time, type, and level. This leads to inaccurate identification of user intent and threatens the prosthesis control reliability. This paper proposed a real-time fault detection and localization approach applied to handwriting device on the plane. This approach allows connecting inputs (IEMG signals)/outputs (pen tip coordinates) data as a parametric model for Multi-Inputs Multi-Outputs (MIMO) system. The proposed approach is considered as a model-independent abrupt or intermittent fault detection method and as an alternative solution to the unpredictable input observer based techniques, without any observability requirements. This approach allows detecting, in real time, several types of faults in one or two inputs signals and in the same or different instants. Our study is appropriate for many rapidly expanding fields and practices, including biomedical engineering, robotics, and biofeedback therapy or even military applications.

Download Full-text

Real-Time 3D Multi-Object Detection and Localization Based on Deep Learning for Road and Railway Smart Mobility

Journal of Imaging ◽

10.3390/jimaging7080145 ◽

2021 ◽

Vol 7 (8) ◽

pp. 145

Author(s):

Antoine Mauri ◽

Redouane Khemmar ◽

Benoit Decoux ◽

Madjid Haddad ◽

Rémi Boutteau

Keyword(s):

Deep Learning ◽

Object Detection ◽

Real Time ◽

Video Game ◽

Autonomous Vehicles ◽

Object Localization ◽

Driver Assistance Systems ◽

Smart Mobility ◽

Bounding Boxes ◽

Detection And Localization

For smart mobility, autonomous vehicles, and advanced driver-assistance systems (ADASs), perception of the environment is an important task in scene analysis and understanding. Better perception of the environment allows for enhanced decision making, which, in turn, enables very high-precision actions. To this end, we introduce in this work a new real-time deep learning approach for 3D multi-object detection for smart mobility not only on roads, but also on railways. To obtain the 3D bounding boxes of the objects, we modified a proven real-time 2D detector, YOLOv3, to predict 3D object localization, object dimensions, and object orientation. Our method has been evaluated on KITTI’s road dataset as well as on our own hybrid virtual road/rail dataset acquired from the video game Grand Theft Auto (GTA) V. The evaluation of our method on these two datasets shows good accuracy, but more importantly that it can be used in real-time conditions, in road and rail traffic environments. Through our experimental results, we also show the importance of the accuracy of prediction of the regions of interest (RoIs) used in the estimation of 3D bounding box parameters.

Download Full-text