Intelligent controls for electro-hydraulic poppet valves

This paper describes an intelligent controller that combines an auto-calibration state-trajectory-based control method with a simple algorithm that enables fault detection. The application of the controller to achieve intelligent electronic pressure control in a fluid power system is included. More specifically, the proposed control algorithm is designed to learn the inverse input-state map of an Electro-Hydraulic Poppet Valve (EHPV) with the aid of a Nodal Link Perceptron Network (NLPN). While in operation, the controller enforces the tracking of a desired hydraulic supply pressure profile at the same time that it learns the inverse input-state calibration map and monitors the latter for deviations from prescribed bounds. The tracking performance of the auto-calibration scheme is validated experimentally and the fault detection algorithm is explored via a computer simulation.