An Empirical Vehicle Speed Model for Tuning Throttle Controller Parameters
Abstract
Controlling the speed of an autonomous ground vehicle is a necessity for autonomous driving. Proportional-integral-differential (PID) controllers are a common solution for controlling the throttle setting of a vehicle in order to achieve a desired speed. PID controllers feature three free parameters which must be "tuned" to achieve the desired controller behavior. While many theoretical approaches for automatic tuning of PID parameters have been considered, practical applications still require extensive field testing because even detailed physical models of ground vehicles often fail to adequately capture all the effects that influence vehicle speed. In order to facilitate tuning of PID parameters for a real-world vehicle, in this work a fully empirical model of vehicle longitudinal dynamics is proposed. With a short series of measurements, a predictive model of the vehicle speed can be developed by fitting the model to the measured data. The empirical model presented in this work has the advantages that it is simple - it does not require any detailed measurements of the vehicle properties but is rather easily fit to real measurements - and flexible - it can be used for a variety of vehicles and conditions. In this work, the development of the model is outlined, and an application of the model is shown for two different vehicles, the Polaris MRZR4 and the Clearpath Warthog. The applicability of the empirical model is demonstrated by tuning and testing a real PID controller for the MRZR4.
Document Details
- Document Type
- Technical Report
- Publication Date
- Jan 01, 2022
- Accession Number
- AD1166735
Entities
People
- Christopher Goodin
- Christopher H. Hudson
- Daniel W. Carruth
- Lucas D. Cagle
- Marc N. Moore
- Paramsothy Jayakumar
Organizations
- Mississippi State University