Showing papers by "Rajnikant Sharma published in 2019"

Learning-based adversarial agent detection and identification in cyber physical systems applied to autonomous vehicular platoon

Abstract: The security of cyber physical systems such as autonomous vehicle platoons plays a vital role in ensuring passenger safety. An adversary in control of a single vehicle can degrade platoon efficiency or even cause collisions. In this paper, we focus on detecting an attack meant to destabilize a platoon, thereby causing collisions, and identifying the source of the attack (i.e., the vehicle under control of the adversary) using Fully Connected Deep Neural Networks (FCDNN) and Convolutional Neural Networks (CNN). The vehicles in the platoon are assumed to be equipped with sensors (LIDAR and RADAR) that measure the range and relative speed of their immediate neighbors. These sensor data, modelled with noise following a Gaussian distribution, are used to train a FCDNN and CNN for attack detection and identification. The effectiveness of these networks are tested for different scenarios based on local and global sensor information availability. The initial study show that for a ten vehicle platoon, CNN detects and identifies an attack with highest accuracy of 97.5%, just by using the own local sensor information. We also show that the range measurements provide better accuracy in comparison to the velocity measurements. The platoon is modelled and simulated in MATLAB and neural networks are generated and tested using Tensorflow and Keras deep learning libraries.

Nonlinear Model Predictive Control to Aid Cooperative Localization

Abstract: This paper proposes a nonlinear model predictive control (NMPC) scheme to tackle the problem of localization and path planning of a group of unmanned aerial vehicles (UAVs) in global positioning system (GPS) denied environments. It is assumed that the UAVs can cooperate by sharing information among themselves. It is also assumed that the area under consideration contains some landmarks with known locations. The NMPC computes the optimal control inputs for the vehicles such that the vehicles cooperate to transit from a source location to a destination while choosing a path that will cover enough landmarks for localization. An Extended Kalman Filter (EKF) is used to estimate the vehicle positions using only relative bearing measurements. The efficacy of the proposed method was evaluated through numerical simulations, and the results are discussed.

Path planning for unmanned vehicles with localization constraints

Abstract: This article addresses a path planning problem for an unmanned vehicle in the presence of localization constraints. Landmarks are used for localizing the position of the vehicle at any time. The localization constraints require that at least two landmarks must be present in the sensing range of the vehicle at any time instant. This problem is formulated as a Mixed Integer Linear Program (MILP) which aims to compute an optimal path for the vehicle and the optimal locations where landmarks must be placed. The facial structure of the polytope of feasible solutions to the MILP is then analyzed, and a branch-and-cut algorithm is developed to find an optimal solution. Extensive computational results that corroborate the effectiveness of the proposed approach are also presented.

Cooperative Relative Pose Estimation to Aid Landing of an Unmanned Aerial Vehicle on a Moving Platform

Abstract: In this paper, we develop a method to estimate the relative position and relative heading of an Unmanned Aerial Vehicle (UAV) attempting to land on a moving platform using range-only measurements under GPS-denied conditions. Landing problems typically require precise estimation of relative position and heading, the proposed solution estimates relative pose between the platform (in this case a ship deck) and the landing multirotor. Vision-based landing techniques although accurate will fail under hostile weather conditions with the lack of Global Positioning System (GPS). This formulation investigates the effect on estimation quality when a single vehicle attempts to land using measurements from the ship only versus a team of multirotors assisting a vehicle to land on the ship. A multi-vehicle simulator has been developed in MATLAB/Simulink that facilitates testing our approaches. The simulation results clearly demonstrate the advantage of using a team for assisted landing rather than alone vehicle performing relative pose estimation using a limited range measurements.

Cooperative Search Area optimization using Multiple Unmanned Aerial Vehicles in a GPS-denied Environment

Abstract: This paper addresses the problem of maximizing surveillance area coverage using multiple Unmanned Aerial Vehicles (UAVs) in an obstacle-laden and Global Positioning System (GPS)-denied environment. The UAVs should achieve this objective using the shortest possible routes while staying inside the designated search space and avoiding the obstacles. To attain a desired area coverage, we divide the NP-hard multi-objective optimization problem of planning optimal routes for all UAVs into 3 parts: (a) optimizing search area coverage, (b) performing obstacle avoidance, and (c) using Cooperative Localization (CL) for state estimation. We demonstrate the efficiency of our algorithm through extensive simulations.

Development of Sliding-Mode Landing Controller using Cooperative Relative Localization with Pattern Generation

Abstract: In this paper, we develop a landing controller based on sliding-mode control law for landing a UAV on a moving ship using relative estimates in GPS-denied environments with range-only measurements Precise knowledge of relative position, orientation and velocity is required to accurately land an Unmanned Aerial Vehicle (UAV) onto the surface of a moving platform Although vision-based techniques have been used to previously solve such problems, they fail in dark or hostile weather conditions and also a line of sight is required with the landing platform at all times Cooperation among different Unmanned Vehicles (UVs) have been introduced to aid the estimation process We also investigate how different trajectories followed by supporting vehicles effect the localization accuracy A Matlab/Simulink simulator has been created and exhaustive simulations have been performed that demonstrates the effect of these trajectories on localization accuracy

Landmark Placement for Cooperative Localization and Routing of Unmanned Vehicles

Abstract: A majority of the routing algorithms for unmanned vehicles rely on Global Positioning System (GPS) information for localization. However, disruption of GPS signals, by intention or otherwise, can render these algorithms ineffective. Known landmark information can aid localization in GPS-denied environments and cooperative localization can add more benefits to multiple vehicles. As resource constraints usually exist in practical applications, this article provides a way to optimally place additional landmarks to aid cooperative localization. Specifically, given a fleet of vehicles, a set of targets and a set of pre-planned paths for each of the vehicles, we aim to optimally place landmarks such that each of the vehicles can estimate its position and orientation by obtaining relative measurement information from at least two landmarks, and the landmark placement cost is minimized. Based on the relative position measurement graph, an integer linear program and its relaxations are presented. The performance of the proposed methodology is evaluated and compared through extensive simulation results.

Cooperative Localization: Challenges and Future Directions

Abstract: This chapter provides a brief introduction to the cooperative localization problem, and some of the widely used techniques were discussed along with the mathematical analysis for convergence. The problem of cooperative localization is used to facilitate navigation in global positioning system-denied environments using the vehicle's motion information from its internal sensors and the information about its surroundings from its external sensors. Several different algorithms with various modifications have been employed to approach the problem of cooperative localization in both the centralized and distributed manner. Since maintaining system observability is crucial to ensuring filter consistency and in turn reliable estimates, quite a lot of attention has been devoted by several authors to develop controllers that can increase the information available to the system or maintain paths that will improve system observability. Cooperative localization has been employed by authors to solve complex problems such as landing Unmanned aerial vehicle on moving platforms such as cars or ships.