Real-Time Safe Landing Zone Identification Based on Airborne LiDAR.

Over the past two decades, there has been a growing demand for generating digital surface models (DSMs) in real-time, particularly for aircraft landing in degraded visual environments. Challenging landing environments can hinder a pilot's ability to accurately navigate, see the ground, and avoid obstacles that may lead to equipment damage or loss of life. While many accurate and robust filtering algorithms for airborne laser scanning (ALS) data have been developed, they are typically computationally expensive. Moreover, these filtering algorithms require high execution times, making them unsuitable for real-time applications. This research aims to design and implement an efficient algorithm that can be used in real-time on limited-resource embedded processors without the need for a supercomputer. The proposed algorithm effectively identifies the best safe landing zone (SLZ) for an aircraft/helicopter based on processing 3D LiDAR point cloud data collected from a LiDAR mounted on the aircraft/helicopter. The algorithm was successfully implemented in C++ in real-time and validated using professional software for flight simulation. By comparing the results with maps, this research demonstrates the ability of the developed method to assist pilots in identifying the safest landing zone for helicopters.

Convolutional Neural Networks as Asymmetric Volterra Models Based on Generalized Orthonormal Basis Functions.

This paper introduces a convolutional neural network (CNN) approach to derive Volterra models of dynamical systems based on generalized orthonormal basis function (GOBF)-Volterra. The approach derives the parameters of the model through a CNN and the neural network's learned weights represent the poles of a system. Simulation results show that the parameters of the system can be exactly recovered when no noise is applied. Furthermore, when noise is present, the errors in the parameters are very small for both the linear and nonlinear cases. Finally, the approach is used to identify the model of a quadcopter using data from actual flight tests. Comparisons with previous works demonstrate that CNNs can be satisfactorily used for the identification of dynamical systems.

A Q-Learning Approach to Flocking With UAVs in a Stochastic Environment.

In the past two decades, unmanned aerial vehicles (UAVs) have demonstrated their efficacy in supporting both military and civilian applications, where tasks can be dull, dirty, dangerous, or simply too costly with conventional methods. Many of the applications contain tasks that can be executed in parallel, hence the natural progression is to deploy multiple UAVs working together as a force multiplier. However, to do so requires autonomous coordination among the UAVs, similar to swarming behaviors seen in animals and insects. This paper looks at flocking with small fixed-wing UAVs in the context of a model-free reinforcement learning problem. In particular, Peng's Q(λ) with a variable learning rate is employed by the followers to learn a control policy that facilitates flocking in a leader-follower topology. The problem is structured as a Markov decision process, where the agents are modeled as small fixed-wing UAVs that experience stochasticity due to disturbances such as winds and control noises, as well as weight and balance issues. Learned policies are compared to ones solved using stochastic optimal control (i.e., dynamic programming) by evaluating the average cost incurred during flight according to a cost function. Simulation results demonstrate the feasibility of the proposed learning approach at enabling agents to learn how to flock in a leader-follower topology, while operating in a nonstationary stochastic environment.