Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

RTI-NMPC for Control of Autonomous Vehicles Using Implicit Discretization Methods

Matheus Wagner, Julio E. Normey-Rico

TL;DR

This work presents a nonlinear model predictive control formulation based on real-time iteration using an implicit discretization of the system's dynamics, with the objective of achieving greater prediction accuracy and lower computational cost when dealing with stiff dynamical systems, as is the case for vehicle dynamics.

Abstract

Recent efforts in the development of autonomous driving technology have induced great advancements in perception, planning and control systems. Model predictive control is one of the most popular advanced control methods, but its application to nonlinear systems still depends on the development of computationally efficient methods. This work presents a nonlinear model predictive control formulation based on real-time iteration using an implicit discretization of the system's dynamics, with the objective of achieving greater prediction accuracy and lower computational cost when dealing with stiff dynamical systems, as is the case for vehicle dynamics. The proposed method is described and later evaluated on a simulation scenario considering modeling errors and external disturbances. The presented results demonstrate the effectiveness of the method when it comes to tracking a given trajectory and its low computational burden, measured in terms of execution time.

RTI-NMPC for Control of Autonomous Vehicles Using Implicit Discretization Methods

TL;DR

This work presents a nonlinear model predictive control formulation based on real-time iteration using an implicit discretization of the system's dynamics, with the objective of achieving greater prediction accuracy and lower computational cost when dealing with stiff dynamical systems, as is the case for vehicle dynamics.

Abstract

Recent efforts in the development of autonomous driving technology have induced great advancements in perception, planning and control systems. Model predictive control is one of the most popular advanced control methods, but its application to nonlinear systems still depends on the development of computationally efficient methods. This work presents a nonlinear model predictive control formulation based on real-time iteration using an implicit discretization of the system's dynamics, with the objective of achieving greater prediction accuracy and lower computational cost when dealing with stiff dynamical systems, as is the case for vehicle dynamics. The proposed method is described and later evaluated on a simulation scenario considering modeling errors and external disturbances. The presented results demonstrate the effectiveness of the method when it comes to tracking a given trajectory and its low computational burden, measured in terms of execution time.

Paper Structure

This paper contains 9 sections, 30 equations, 3 figures, 2 tables.

Table of Contents

  1. Introduction
  2. Implicit Real-Time Iteration Nonlinear Model Predictive Control
  3. Vehicle Dynamics Models
  4. Simulation Model
  5. Prediction Model
  6. Case Study
  7. Scenario Description
  8. Results
  9. Conclusion

Figures (3)

  • Figure 1: Reference and simulated trajectories for the NMPC based on RTI and direct NLP solution.
  • Figure 2: Simulation results for the RTI and direct NLP controllers: (a) Wind disturbance for the proposed evaluation scenario. (b) Wheel torque resulting. (c) Steering angle rate. (d) Tracking error.
  • Figure 3: Execution time observations for the NMPC based on the direct NLP solution and on the RTI scheme