MPPI-Generic

MPPI-Generic is a C++/CUDA header-only library for conducting stochastic optimal control in real-time on a NVIDA GPU.

Main Features

• Performance-focused code optimizations that makes near-full utilization of NVIDIA hardware
• Dynamics and Cost Function-agnostic implementations of Model Predictive Path Integral (MPPI) [1], Tube-MPPI [2], and Robust MPPI (RMPPI) [3] algorithms.
• Largest variety of provided Dynamics and Cost Functions publicly available to use with MPPI, Tube-MPPI, and RMPPI
• Different sampling distributions such as Gaussian noise, Colored noise [4], NLN noise [5], and Smooth-MPPI’s sampling scheme [6]
• APIs to allow for researchers to write custom Dynamcis, Cost Functions, stochastic Controllers, and Sampling Distributions that allow plug-and-play swapping with existing components.

Table of Contents

• The MPPI algorithm
• Installation
• Getting started using the library
• How do we achieve high performance?
• Comparisons to other popular implementations of MPPI

Citation

If you use this work, please cite the following paper:

@misc{vlahov2024mppi,
      title={MPPI-Generic: A CUDA Library for Stochastic Optimization},
      author={Bogdan Vlahov and Jason Gibson and Manan Gandhi and Evangelos A. Theodorou},
      year={2024},
      eprint={2409.07563},
      archivePrefix={arXiv},
      primaryClass={cs.MS},
      url={https://arxiv.org/abs/2409.07563},
}

Publications using MPPI-Generic

We have been developing MPPI-Generic for years to ensure that it is useful on real hardware in a variety of situations. Below are papers that have already started using the MPPI-Generic library.

B. Vlahov, J. Gibson, D. D. Fan, P. Spieler, A.-a. Agha-mohammadi, and E. A. Theodorou, “Low Frequency Sampling in Model Predictive Path Integral Control,” IEEE Robotics and Automation Letters, pp. 1–8, 2024. [Online]. Available: https://ieeexplore.ieee.org/document/10480553

A. M. Patel, M. J. Bays, E. N. Evans, J. R. Eastridge, and E. A. Theodorou, “Model-Predictive Path-Integral Control of an Unmanned Surface Vessel with Wave Disturbance,” in OCEANS 2023 - MTS/IEEE U.S. Gulf Coast, Sep. 2023, pp. 1–7. [Online]. Available: https://ieeexplore.ieee.org/document/10336978

J. Gibson, B. Vlahov, D. Fan, P. Spieler, D. Pastor, A.-a. Agha-mohammadi, and E. A. Theodorou, “A Multi-step Dynamics Modeling Framework For Autonomous Driving In Multiple Environments,” in 2023 IEEE International Conference on Robotics and Automation (ICRA). IEEE, May 2023, pp. 7959–7965. [Online]. Available: https://ieeexplore.ieee.org/document/10161330

M. Gandhi, H. Almubarak, Y. Aoyama, and E. Theodorou, “Safety in Augmented Importance Sampling: Performance Bounds for Robust MPPI,” Apr. 2022. [Online]. Available: http://arxiv.org/abs/2204.05963

M. Gandhi, B. Vlahov, J. Gibson, G. Williams, and E. A. Theodorou, “Robust Model Predictive Path Integral Control: Analysis and Performance Guarantees,” IEEE Robotics and Automation Letters, vol. 6, no. 2, pp. 1423–1430, Feb. 2021. [Online]. Available: https://arxiv.org/abs/2102.09027v1

References

[1] G. Williams, P. Drews, B. Goldfain, J. M. Rehg, and E. A. Theodorou, “Aggressive Driving with Model Predictive Path Integral Control,” in 2016 IEEE International Conference on Robotics and Automation (ICRA). IEEE, 2016, pp. 1433–1440. [Online]. Available: https://ieeexplore.ieee.org/document/7487277/

[2] G. Williams, B. Goldfain, P. Drews, K. Saigol, J. Rehg, and E. Theodorou, “Robust Sampling Based Model Predictive Control with Sparse Objective Information,” in Robotics: Science and Systems XIV. Robotics: Science and Systems Foundation, Jun. 2018. [Online]. Available: http://www.roboticsproceedings.org/rss14/p42.pdf

[3] M. Gandhi, B. Vlahov, J. Gibson, G. Williams, and E. A. Theodorou, “Robust Model Predictive Path Integral Control: Analysis and Performance Guarantees,” IEEE Robotics and Automation Letters, vol. 6, no. 2, pp. 1423–1430, Feb. 2021. [Online]. Available: https://arxiv.org/abs/2102.09027v1

[4] B. Vlahov, J. Gibson, D. D. Fan, P. Spieler, A.-a. Agha-mohammadi, and E. A. Theodorou, “Low Frequency Sampling in Model Predictive Path Integral Control,” IEEE Robotics and Automation Letters, pp. 1–8, 2024. [Online]. Available: https://ieeexplore.ieee.org/document/10480

[5] I. S. Mohamed, K. Yin, and L. Liu, “Autonomous Navigation of AGVs in Unknown Cluttered Environments: Log-MPPI Control Strategy,” IEEE Robotics and Automation Letters, vol. 7, no. 4, pp. 10 240–10 247, 2022. [Online]. Available: https://ieeexplore.ieee.org/document/9834098

[6] T. Kim, G. Park, K. Kwak, J. Bae, and W. Lee, “Smooth Model Predictive Path Integral Control Without Smoothing,” IEEE Robotics and Automation Letters, vol. 7, no. 4, pp. 10 406–10 413, 2021. [Online]. Available: https://ieeexplore.ieee.org/document/9835021