State Estimate Based on Nonlinear Filter of ElectroSlag Remelting Process

State estimate strategy of electroSlag remelting process based on nonlinear filter algorithm was put forward in order to settle difficulty that electrode immersion depth can not directly measured, other state variable can not be unable directly measured and directly measure variable is with all kinds noise. The nonlinearfilter algorithms including unscented Kalman Filter UKF, Divided Difference filter (DDF) that includes first-order DDF (DD1) and second-order DDF (DD2) were established. Estimated performance of three filter estimators was simulated. Study results shows that nonlinear filter algorithm of electroslag remelting process is feasible. It can effectively estimate the state variable including slag temperature, electrode immersion depth, and electrode mass, ram position etc. The estimate precision of DD1 and DD2 is better than UKF for root-mean-square error (RMS). As The computing efficiency of DD1 higher than DD2, the DD1 is more feasible of electroSlag remelting process though the estimate precision of DD2 better than DD1.