Optimal sensor placement using machine learning. (15th December 2017)

• Record Type: Journal Article
• Authors: Semaan, R.
• Is Part Of: Computers & fluids. Volume 159(2017)
• Page Start: 167
• Page End: 176
• Publication Date: 2017-12-15
• Subjects: Machine learning -- Optimal sensor placement -- Flow control; Fluid dynamics -- Data processing -- Periodicals; 532.050285
• DOI: 10.1016/j.compfluid.2017.10.002
• Languages: English
• ISSNs: 0045-7930

Abstract

Highlights: A machine learning-based approach for optimal sensor placement is proposed. The approach relies on input variable importance ranking. Method validated against POD mode extrema, and against a brute force approach. Flow conditions and sensor type have an effect on optimal sensor placement. Choice of response function has limited influence on optimal sensor placement. Abstract: A new method for optimal sensor placement based on input variables importance of machine learned models is proposed. With its simplicity, adaptivity, and low computational cost, the method offers many advantages over existing approaches. The method is implemented on flow over an airfoil equipped with a Coanda actuator. The analysis is based on flow field data obtained from two-dimensional unsteady Reynolds averaged Navier–Stokes (URANS) simulations with different actuation conditions. The optimal sensor locations are compared against the current de-facto standard of maximum POD modal amplitude location, and against a brute force approach that scans all possible sensor combinations. The results show that both the flow conditions and the type of sensor have an effect on the optimal sensor placement, whereas the choice of the response function appears to have limited influence.