Advancements in Data-Enabled Predictive Control and Model Predictive Path Integral Control

The field of control systems is moving towards more efficient and adaptive methods, particularly in the areas of Data-Enabled Predictive Control (DeePC) and Model Predictive Path Integral (MPPI) control. Researchers are exploring new approaches to improve the performance and scalability of these methods, such as incorporating datamodels, gain-scheduling, and Bayesian optimization. Noteworthy papers in this area include:

  • Datamodel-Based Data Selection for Nonlinear Data-Enabled Predictive Control, which proposes a novel approach to selecting relevant data columns based on the control objective.
  • DM-MPPI: Datamodel for Efficient and Safe Model Path Integral Control, which extends the Datamodels framework to MPPI control, enabling real-time estimation and efficient sample pruning.
  • Bayesian Optimization for Automatic Tuning of Torque-Level Nonlinear Model Predictive Control, which presents an auto-tuning framework for torque-based Nonlinear Model Predictive Control using high-dimensional Bayesian Optimization techniques.

Sources

Datamodel-Based Data Selection for Nonlinear Data-Enabled Predictive Control

DM-MPPI: Datamodel for Efficient and Safe Model Path Integral Control

How to Capture Human Preference: Commissioning of a Robotic Use-Case via Preferential Bayesian Optimisation

Gain-Scheduling Data-Enabled Predictive Control for Nonlinear Systems with Linearized Operating Regions

Bayesian Optimization for Automatic Tuning of Torque-Level Nonlinear Model Predictive Control

Gauss-Newton accelerated MPPI Control

Built with on top of