Note: This content is accessible to all versions of every browser. However, this browser does not seem to support current Web standards, preventing the display of our site's design details.


Approximation of Hybrid MPC for Building Control by Simple Decision Rules


A. Domahidi

Talk given at the OptiControl meeting.

While model predictive control offers a superior performance when compared to existing rule based solutions for building control, its complexity in terms of implementation and maintenance makes MPC difficult to apply in practice. In this talk, we describe how to extract simple decision rules for binary control inputs using machine learning methods such as support vector machines (SVMs) or AdaBoost. This allows to recover most of the performance that an MPC controller offers with simple decision rules based on majority voting. The extracted rules are based on classification and regression trees (CARTs), can be implemented on very simple microprocessors and are readable by humans, allowing for further manual tuning in the field.


Type of Publication:


File Download:

Request a copy of this publication.
(Uses JavaScript)
% No recipe for automatically generating a BibTex entry for (06)Talk
Permanent link