Nonlinear Model Predictive Coordinated Standoff Tracking of Moving Ground Vehicle

Abstract

This paper proposes a nonlinear model predictive control framework for coordinated standoff tracking by a pair of unmanned aerial vehicles. The benefit of this approach is to get optimal performance compared to using a decoupled controller structure: heading control for standoff-distance keeping and speed control for phase-keeping. The overall controller structure is fully decentralized in a fact that each unmanned aerial vehicle optimizes its controller based solely on the future propagation of the pair vehicle states and the target estimates received via communication. This paper uses an acceleration model for a sophisticated and realistic target dynamics, which can consider more reasonable system noise covariance matrix reflecting the target's motion characteristics. To simplify optimization formulation and decrease computation burden, a new manipulation using inner product of position vectors of the unmanned aerial vehicles with respect to the target position is proposed for antipodal tracking instead of using the relative phase angle difference. To consider a more realistic situation, inequality constraints are considered for collision avoidance between unmanned aerial vehicles and control input saturations using penalty functions in the model predictive control scheme. Simulations with a pair of unmanned aerial vehicles are done using a realistic car trajectory data in a urban environment in the United Kingdom to verify the feasibility and benefit of the proposed approach with comparing to a Lyapunov vector field guidance.

Decentralised Standoff Tracking of Moving Targets Using Adaptive Sliding Mode Control for UAVs

“…This study considers acceleration dynamics [19], [7] for a ground moving vehicle, which irregularly performs stop-and-go manoeuvres. UAVs are assumed to be equipped with a GMTI sensor to localise the position of target.…”

“…If the frequencies of the tracking guidance and autopilot are not too close, it is common to initially design and verify the guidance law and control algorithm separately. Therefore, like in many literatures considering similar guidance problems [5], [7], [10], above simple kinematics is used for the UAV model. However, the final validation needs to be made with higher complexity simulation models and flight tests, and these remain as future work.…”

“…The vehicle trajectory data acquired at 2Hz in a S-Paramics [21] traffic model of Devizes, United Kingdom, are used to generate the GMTI measurements composed of relative range and azimuth angle with respect to a position of UAV, as shown in Fig. 6 [7]. The ground vehicle departs at the western side of Devizes and traverses a part of the town center and then turns back until the journey ends at the northwestern side of Devizes.…”

“…Summers et al [6] extended this phase keeping idea with a variable airspeed controller to multiple UAVs using information architectures in vehicle formations. Unlike decoupled control structures, Kim et al [7] proposed a nonlinear model predictive approach to find the heading and velocity control inputs simultaneously using a coupled structure, providing optimal performance in terms of tracking as well as control efforts. Oh et al [8] used the solution of differential geometry between the UAV and a target, which brings several advantages along with its inherent simplicity: rigorous stability, explicit use of a target velocity, and tuning parameter reduction.…”

Airborne behaviour monitoring using Gaussian processes with map information

“…The model regards the target acceleration as a process correlated and exponentially decreasing in time, which means if there is a certain acceleration rate at a time t, then it is likely to be the same jerk also at a time instant t + τ as [15], [16]:…”