NSF Award
2301535
This Engineering Research Initiation (ERI) grant will fund research that enables reliable operation and cost-effective maintenance of machinery in which hydraulic cylinder actuators play a central role—of critical importance to the transportation, manufacturing, construction, and agricultural industries—thereby promoting the progress of science and advancing the national prosperity. Given its high power density and large force capacity, the hydraulic cylinder is a trusted workhorse across many industrial applications. By its design, it is also vulnerable to a variety of critical faults, such as seal failures, leakage, fluid contamination, and reduced load-carrying capacity, that may negatively affect performance and efficiency, and result in high costs from downtime, permanent damage, or operator injury. To ensure that such faults can be accurately detected and identified, this project relies on a combination of physics-based modeling and innovative frequency domain analysis to develop a novel diagnostic methodology that accounts for system nonlinearities and closed-loop operation. Such a diagnostic paradigm is also anticipated to find application in a variety of energy and power transmission systems, including electric vehicles. Efforts to disseminate research outcomes to local industry in Southeast Michigan, undergraduate research opportunities, and integration of modeling and systems diagnostics principles in coursework will produce broader societal impact.<br/><br/>This research aims to develop the foundations for a comprehensive systems diagnostics methodology for hydraulic cylinders that relies on physics-based modeling rather than statistics and machine learning techniques, as has been common in recent years. Such foundational contributions will be made through the design of a new residual generator that extracts faulty features from sensor measurements, as well as of a residual evaluator that makes a diagnostic decision by comparing the residual with the prescribed threshold under a certain operating condition. Several tasks will be pursued through a combination of theoretical modeling, numerical simulations, and physical experiments, including comprehensive fault analysis using a nonlinear, inverse, frequency domain representation of the system dynamics, and formulation and testing of a parameter estimation algorithm for such a nonlinear inverse model under closed-loop operation using an adaptive digital twin. Hardware-in-the-loop simulations will be performed to validate the methodology and investigate its robustness also in the presence of multiple simultaneous faults.<br/><br/>This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.
