Dr. Vladimir Mahalec

Professor and Associate Director (Grad studies - W Booth School of Engineering Practice and Technology)

Dynamic Real-Time Optimization | Hybrid First-Principles/Data-Driven Models | Integrated Community Energy Systems | Machine Learning | Optimal Planning and Scheduling | Supply Chain Optimization

MACC Researcher since May 2010

Department of Chemical Engineering

Research Summary

Software engineering and application design, development, and implementation (advanced process control, process engineering and design, planning, scheduling, plant optimization, feedstock selection, trading and logistics). Expertise in value chain optimization in petroleum, petrochemicals, and other process industries.

Employment History

1991-2006 Aspen Technology Inc. Cambridge Massachusetts, Senior Vice President
for Technology Management

1978-1991 Esso Petroleum, Manager, Process control Group

1977-1978 MIT Cambridge Massachusetts, Research Associate

Research Interests

Civic and industrial energy systems have traditionally been designed and operated as separate, stand alone entities. Individually, they both offer ample opportunities to improve their performance and part of our research is devoted to improving designs of CCHP systems. Additional, possibly even greater opportunities lie in integrated design and operation of civic and industrial energy systems.
Design and operation of process plants enables improvements in plant performance via designs which minimize energy usage and optimal operation. Besides working on more efficient plant designs, our research includes globally optimal nonlinear control and dynamic real-time optimization, process scheduling, and production planning. We have introduced inventory-pinch concept as a basis for aggregation and decomposition of large scale nonlinear planning and scheduling problems and are currently exploring its use in planning under uncertainty.
Structure design and operation of supply chains in real life is carried out in a competitive environment. Part of our research is to define game-theoretic planning framework which accounts for actions of various players in the market and enables best decision making in light of the competitive decisions by other players.
Process modelling efforts in our team focus on combining models derived from actual plant data with the first principles (hybrid models) in order to develop modeling methodology leading to models which are very accurate, small in size, and mostly linear. This enables the same process unit model to be used in scheduling, in planning, and in real-time optimization.
While we do not focus on inventing new algorithms, in case that the existing methods are not adequate for solving problems at hand, we do not shy away from pursuing development of algorithms as needed to solve problems at hand.

Profile

Dr. Mahalec spent many years in industry, including 15 at Aspen Technology, where he was Senior VP of Technology. His expertise includes process modeling, simulation, design, RTO, and supply chain optimization.

Education

Dipl. Ing., U. of Zagreb, Croatia
Ph.D. Chemical Engineering, University of Houston