Online state and parameter estimation of an electrohydraulic valve for intelligent monitoring

Summary form only given. A novel nonlinear state-space model for a two-stage proportional directional servovalve is proposed, that has all physical states and coefficients necessary for an online model-based condition monitoring and fault diagnosis algorithm which tracks abrupt or gradual changes in a number of key parameters and states. Extended Kalman filtering has enabled us to simultaneously reconstruct the system "hard-to-measure" states and estimate physical coefficients using only a few basic measurements. One of those crucial states is the instantaneous amount of fluid passing through a variable orifice valve. Using a simulation example, the flow rate has been accurately predicted through a novel model that lets the orifice effective area be an initially unknown function of the spool displacement. The area and also the size of spool deadband, both embedded in the function, are then automatically revealed during the online estimation process, while the switching between the open-to-supply and open-to-tank port models is internally made.