|Title||Optimization and Controlled Systems: A Case Study on Thermal Aware Workload Dispatching|
|Publication Type||Conference Paper|
|Year of Publication||2012|
|Authors||Bartolini, A., M. Lombardi, M. Milano, and L. Benini|
|Conference Name||Proceedings of AAAI|
Despite successfully employed on many industrial problems, Combinatorial Optimization methods still have limited applicability on many real-world domains, often due to modeling difficulties. This is typically the case for systems under the control of an on-line policy: even when the basic controller rules are well known, capturing their behavior in a declarative model is often impossible by conventional means. Such a diffi- culty is at the root of the classical, sharp separation between off-line and on-line approaches. In this paper, we investigate a general method to model controlled systems, based on the integration of Machine Learning and Constraint Programming (CP) technology. Specifically, we use an Artificial Neural Network (ANN) to learn the behavior of a controlled system (a multicore CPU with thermal controllers) and plug it into a CP model by means of Neuron Constraints. Optimization is performed via Large Neighborhood Search and obtains significantly better results compared to an approach with no ANN guidance. Neuron Constraints were first introduced in [Bartolini et al., 2011b] as a mean to model complex systems: providing evidence of their applicability to controlled systems is a significant step forward, greatly broadening the applicability of combinatorial methods and disclosing unmatched opportunities for hybrid off-line/on-line optimization.