A Blob Method for Mean Field Control With Terminal Constraints
Authors:
Katy Craig,
Karthik Elamvazhuthi,
Harlin Lee
Abstract:
In the present work, we develop a novel particle method for a general class of mean field control problems, with source and terminal constraints. Specific examples of the problems we consider include the dynamic formulation of the p-Wasserstein metric, optimal transport around an obstacle, and measure transport subject to acceleration controls. Unlike existing numerical approaches, our particle me…
▽ More
In the present work, we develop a novel particle method for a general class of mean field control problems, with source and terminal constraints. Specific examples of the problems we consider include the dynamic formulation of the p-Wasserstein metric, optimal transport around an obstacle, and measure transport subject to acceleration controls. Unlike existing numerical approaches, our particle method is meshfree and does not require global knowledge of an underlying cost function or of the terminal constraint. A key feature of our approach is a novel way of enforcing the terminal constraint via a soft, nonlocal approximation, inspired by recent work on blob methods for diffusion equations. We prove convergence of our particle approximation to solutions of the continuum mean-field control problem in the sense of Gamma-convergence. A byproduct of our result is an extension of existing discrete-to-continuum convergence results for mean field control problems to more general state and measure costs, as arise when modeling transport around obstacles, and more general constraint sets, including controllable linear time invariant systems. Finally, we conclude by implementing our method numerically and using it to compute solutions the example problems discussed above. We conduct a detailed numerical investigation of the convergence properties of our method, as well as its behavior in sampling applications and for approximation of optimal transport maps.
△ Less
Submitted 15 February, 2024;
originally announced February 2024.
Competing Advanced Process Control via an Industrial Automation Cloud Platform
Authors:
L. L. Rokebrand,
J. J. Burchell,
L. E. Olivier,
I. K. Craig
Abstract:
This paper proposes an innovative approach for the advanced control of an industrial process via an automation cloud platform. Increased digital transformation and advances in Industrial Internet of Things (IIoT) technologies make it possible for multiple vendors to compete to control an industrial process. An industrial automation cloud platform facilitates the interaction between advanced proces…
▽ More
This paper proposes an innovative approach for the advanced control of an industrial process via an automation cloud platform. Increased digital transformation and advances in Industrial Internet of Things (IIoT) technologies make it possible for multiple vendors to compete to control an industrial process. An industrial automation cloud platform facilitates the interaction between advanced process control (APC) vendors and the process. A selector, which forms part of the platform, is used to determine the best controller for a process for any given time period. The article starts with a general overview of platform businesses, platforms aimed at industry, and the steps required to build such platforms. Issues that need to be addressed to make APC via an automation platform practically viable are discussed including what process information to provide to APC vendors, continuous evaluation of controllers even when not in control of the process, bumpless transfer, closed-loop stability, constraint handling, and platform security and trust. A case study is given of competing APCs via an industrial automation cloud platform. The process used in the study is a surge tank from a bulk tailings treatment plant, the aim of which is to keep the density of the tank out flow constant while maintaining a steady tank level. A platform facilitates the competition of three vendors for control of this process. It is shown that the cloud platform approach can provide the plant access to a superior controller without the need for directly procuring the services of an exclusive vendor.
△ Less
Submitted 26 November, 2020;
originally announced November 2020.
