Research Intern
Are you applying to the internship?
Job Description
About the Company:
Mitsubishi Electric Research Labs, Inc. “MERL” is committed to equal employment opportunities for all employees and applicants, adhering to federal, state, and local laws regarding nondiscrimination. They expressly prohibit any form of workplace harassment. Working at MERL requires full authorization to work in the U.S. and access to technology subject to governmental access control restrictions.
Job Description:
MERL is seeking a Research Intern to contribute to the development of efficient transformer-informed stochastic MPC (Model Predictive Control) for the control of net-zero energy buildings. This internship offers the chance to work on cutting-edge research at the intersection of deep learning and predictive control, applied to a real-world system. The goal is to create and implement novel control strategies for energy-efficient buildings. The successful candidate is expected to publish the results of their work during the internship.
The expected internship duration is 3-6 months, with a flexible start date.
Key Responsibilities (inferred):
• Research and develop transformer-based models for probabilistic time-series prediction of building energy consumption and environmental factors.
• Design and implement stochastic MPC algorithms that utilize the transformer model predictions to optimize building energy usage while accounting for uncertainties.
• Formulate MPC/SMPC problems as convex optimization problems.
• Implement and test the developed algorithms in simulation or potentially on a real building system.
• Document research findings and prepare manuscripts for publication.
Ideal Candidate Requirements:
• Significant hands-on experience with stochastic MPC (SMPC)
• Publications in SMPC are a strong plus
• Fluency in Python and PyTorch
• Understanding of probabilistic time-series prediction
• Experience with convex programming
• Convex formulations of MPC/SMPC problems are a strong plus
• Completion of an MS degree or completion of >50% of a PhD program
