EKF Moisture Estimation in Slider-Crank Organic Waste Presses: Angle-Synchronized vs. Revolution-Averaged Motor Current Analysis

A prior simulation study demonstrated that the Extended Kalman Filter (EKF) applied to induction motor stator current can estimate the moisture content of pressed organic waste with a root-mean-square error (RMSE) of 3.18%, assuming a constant mechanical lever arm. In practice, the slider-crank mechanism selected for the AENG-OWP-15 household organic waste processor imposes a crank-angle-dependent lever arm $r_{\text{eff}}(\theta)$, introducing rapid sinusoidal oscillations into the motor torque and current signal. This paper extends the baseline EKF design to handle this compound signal, evaluating three architectural approaches: (1) a Naive application of the EKF to the raw angle-modulated signal; (2) Angle-Synchronized sampling restricted to a $\pm15^\circ$ window around $\theta = 90^\circ$ where the lever arm is maximal; and (3) Revolution-Averaged current processing, which filters out angle dependence entirely over a complete crank revolution. Simulation results for a 60-second press cycle indicate that while the Naive EKF fails