Experiment setup design for non-dimensional vehicle dynamics studie

The aim of this work is to design and develop a complete setup for experimenting non-dimensional vehicle dynamics and validations over various control strategies and manoeuvres for road cars. The target setup consists of an RC (Radio-Control) vehicle equipped with sensors capable of measuring real-time vehicle responses such as accelerations and yaw rate, and also applying autonomous steer-by-wire commands to the vehicle. Hardware network is built upon Arduino platform including an IMU sensor, Arduino microcontroller, and Wi-Fi modules. The key idea to design the experiment setup is derived from the dynamic similarity relationships between a real system (the prototype) and a scaled version (the model) using the science of dimensional analysis. The setup connects the dynamic responses of scaled vehicles to a 1:1 vehicle with a mathematical theorem called Buckingham Pi-theorem. In this thesis, the step-by-step development of each part of the setup is clearly documented including formulation of Pi parameters, Arduino program codes, Steering control by wire, PID controller implementation, response logging, response postprocessing, setup installation in the vehicle, validation of the setup etc., The setup thus developed has a scaled vehicle equipped with 16-bit 3-axes wireless acceleration and angular velocity sensor and 100-200Hz data logger, PID controller feature, Programmable data logger console with in-built noise filtering, wirelessly controllable steering motor and throttle along with a treadmill track. The correct functioning of the setup is validated against the known bicycle model by steady-state responses such as beta response, lateral velocity response, curvature gain response, yaw rate response and its values are verified against the sensor data under steady-state test conditions(3DOF model). Transient conditions were validated using sensor data and Matlab State Space vehicle dynamics models. The results thus obtained were found to be satisfactory to continue further development process. Specific tests were performed to estimate cornering stiffness values of the tires and yaw moment of inertia of the vehicle. The setup can retrieve the parameters such as angular velocities and accelerations etc., required for non-dimensional vehicle dynamics analysis along with the known physical parameters of the vehicles such as mass, moment of inertia and cornering stiffness of the tires to generate Pi parameters. Thus, the Pi parameters generated can be matched with dynamics of a 1:1 car. The dynamics of the car of interest can be simulated in this setup through its flexible physical variables and testing conditions.