Development of ANVEL HIL/SIL Simulation Environment for Rapid Prototyping of Navigation Algorithms

Abstract

This thesis presents the development of a Hardware In the Loop/Software In the Loop (HIL/SIL) simulation environment with the purpose of testing Global Positioning System/Inertial Navigation System (GPS/INS) navigation units in ANVEL. HIL/SIL test beds are widely popular research tools that allow researchers to combine high fidelity vehicle simulations with errors inherent to hardware implementation, providing more realistic simulations. Researchers may therefore use these test beds further along in the design process with confidence the simulation is behaving as in the real-world, removing operators from the test environment until final stages of development when the technology is more fully proven. The work presented in this thesis focuses on the development of a modular HIL/SIL test bed using software and hardware add-ons to the Autonomous Navigation Virtual Environment Laboratory (ANVEL), a high fidelity vehicle simulation environment. The test bed is designed with the goal of testing GPS/INS navigation units, specifically the unit developed by Auburn's GAVLab.

The capabilities of ANVEL are extended through the use of a plugin to relay vehicle state information out of the simulation environment. A second plugin is developed for software GPS simulation in which satellites are simulated using broadcast ephemeris data. Ray tracing is made possible through ANVEL's physics engine, allowing the simulation to detect satellite obstructions. Vehicle state information from ANVEL is used to generate Inertial Measurement Unit (IMU) and Wheel Speed Sensor (WSS) data in software modules which corrupt the information according to error models derived from real-world sensors. Hardware implementations for the GPS, IMU, and WSS modules are then developed to add realism to the test environment. A Spectracom GPS simulator is used to provide Radio Frequency (RF) signal in real-time to the navigation unit using position and satellite availability information from ANVEL. A serial interface is then developed such that the IMU module outputs a serial signal to emulate the real sensor. Finally, quadrature signals are generated using Pulse Width Modulators (PWMs) to represent encoder pulses from wheel speed sensors. Each of the developed software and hardware modules are then validated in both static and dynamic test scenarios using error characteristics from sensors in the GAVLab navigation unit as benchmarks for comparison. Results from validation demonstrate the test bed is capable of outputting realistic sensor measurements which may be used interchangeably with sensors in a navigation unit for the purpose of algorithm development.