A Multiple UAV Simulation for Researchers

Abstract

MultiUAV is capable of simulating 8 vehicles and many targets. There are 10 targets in the released version of MultiUAV. Simulated vehicles include embedded flight software (EFS) and vehicle dynamics. EFS is the software that implements cooperative control algorithms. The vehicle dynamics are simulated with six-degree-of-freedom (SDOF) equations of motion. The vehicle model includes an autopilot that makes the vehicles capable of waypoint navigation. Recently, a new communication simulation has been added to MultiUAV that makes it possible to track individual messages sent between the vehicles. Also uncertainty in target identification and weapon lethality has been added to MultiUAV to enable new research 3 . This paper describes MultiUAV, a simulation that is capable of simulating multiple unmanned aerospace vehicles which cooperate to accomplish a predefined mission. The simulation was constructed using the Mathwork’s Simulink simulation software. Construction of the simulation satisfied the need for a simulation environment that researchers can use to implement and analyze cooperative control algorithms. The simulation is implemented in a hierarchical manner with inter-vehicle communication explicitly modeled. During construction of MultiUAV, issues concerning memory usage and functional encapsulation were addressed. MultiUAV includes plotting tools and links to an external program for post-simulation analysis. Each of the vehicle simulations include six-degree-of-freedom dynamics and embedded flight software. The embedded flight software consists of a collection of managers (agents) that control situational awareness and responses of the vehicles. Managers included in the simulation are: Tactical Maneuvering, Sensor, Target, Cooperation, Route and Weapons. 2 MultiUAV IMPLEMENTATION MultiUAV is organized hierarchically with two major toplevel blocks, Vehicles and Targets, see Figure 1. The other two blocks at the top level, Initialization and DataForPlotting, call functions to initialize the simulation and save simulation data for plotting. The top-level blocks contain the sub-blocks and connections required to implement simulation of the 8 vehicles and 10 targets, see Figures 2 and 3. 1 BACKGROUND There are three types of functions used in MultiUAV, script, compiled and Simulink built-in. The majority of functions in MultiUAV are written in MATLAB’s script language and accessed from Simulink using S-function blocks. This makes it convenient for researchers since many know MATLAB script language. The compiled functions include the classes used from CATA and other functions that are available in other languages. The comIn order to implement and evaluate cooperative control strategies for unmanned aerospace vehicles (UAVs), a simulation environment was needed. During an earlier project, Control Automation and Task Allocation (CATA), a multiple-vehicle/multi-agent simulation was developed 1. The CATA simulation was constructed in C++ and was modified for use in cooperative control research. This simulation was found to be very useful in cooperative control studies 2 . Since the CATA simulation was relatively large and written in C++ it proved to be difficult for other researchers to use. This prompted the development of a Simulink-based multi-vehicle/multi-agent simulation (MultiUAV). Simulink is a symbolic programming environment which makes the simulation relatively easy for researchers to understand and use 6 .