Wei-Xu Cai

Title: Wei-Xu Cai and Data-Driven Measurement of Physical Properties.

Introduction

Wei-Xu Cai is an inventor associated with National Central University in Taoyuan, Taiwan. His patent work describes applying deep learning to estimate physical properties of a fluid from measurements of its mass flow through a vessel.

Latest Patents

Cai is credited with 1 patent: “Method of Measuring Physical Properties.” The method uses vessel shape information and training physical-property parameter sets to generate simulated mass-flow data. A deep-learning model is trained using the simulated data as input and the corresponding physical-property parameters as output. During measurement, fluid passes through an opening into a loader, and a scale records the loader’s accumulated weight over time. The resulting measured mass-flow data is then supplied to the model to estimate the fluid’s physical-property parameters.

Career Highlights

The patent highlights an approach that combines physical measurements, vessel geometry, and machine learning. By using weight accumulation over time to derive mass-flow data, the method connects an observable measurement to a model-based estimate of fluid properties.

Collaborations

The supplied information identifies Chun-Hung Weng as Cai’s coworker. It does not specify their roles or describe a particular collaboration on the patent.

Conclusion

Wei-Xu Cai’s patent presents a deep-learning method for estimating fluid properties from mass-flow measurements, reflecting the intersection of measurement technology and computational modeling.