Novel Non-Contacting Position Estimation System for Long-Stroke Actuators

Project summary:

This project will develop a prototype position measurement sensor useful for a wide variety of piston-cylinder actuators. The project will build on previous National Science Foundation research in which a position sensor suitable for small-stroke non-ferrous actuators was developed. While the previous invention has been successfully licensed to an industrial partner, it works well only for small stroke (up to 15 cm) applications. The fundamental research described in this project will enable enhancement of the sensing technology so as to provide accurate measurements over long stroke lengths. This will open up a large market for the sensor, including the long-stroke ferrous hydraulic actuators utilized on agricultural and construction vehicles.

The project will conduct research to develop technology that enables the non-contacting sensor to work for long-stroke applications in ferrous actuators. This requires addressing several fundamental challenges, including the lack of accuracy and lack of robustness of previous estimation algorithms based on the extended Kalman filter which relied on linearization of nonlinear magnetic field models. Other technical challenges include the presence of significant hysteresis in the magnetic field model due to the ferrous cylinder being made of soft steel which gets magnetized and demagnetized in real time from the oscillatory motion of the internal magnet, and the requirement to daisy-chain multiple sensors for a multi-output nonlinear system in which only subsets of the sensors can be utilized in piecewise regions of operation of the sensor system.

Solutions to address these challenges will be developed in this project. The proposed solutions include rigorous development of nonlinear-observer-based estimation algorithms that provide both global asymptotic stability as well as rejection of disturbances and noise, development of a new analytical model for hysteresis that has significant advantages over the traditional Preisach model, and a systematic design methodology to address sensor switching and daisy chaining of sensors.