State estimator based on an indirect Kalman filter for a hydraulically actuated multibody system.

Area of Science:

• Multibody Dynamics
• Hydraulic Systems Engineering
• Control Theory

Background:

• Coupled multibody systems often integrate with hydraulic subsystems, necessitating accurate state estimation.
• Real-time inference of dynamical states requires fusing model data with sensor information.

Purpose of the Study:

• To propose a novel state estimator for hydraulically actuated multibody systems.
• To extend the indirect Kalman filter for hydraulically driven dynamics, incorporating exact Jacobians.

Main Methods:

• Developed a state estimator by coupling a semi-recursive multibody formulation with an indirect Kalman filter for hydraulics.
• Utilized lumped fluid theory for the hydraulic subsystem and a monolithic coupling approach.
• Investigated covariance matrix structures for plant and measurement noise.

Main Results:

• Applied the state estimator to a hydraulically actuated four-bar mechanism, considering modeling uncertainties.
• Simulated measurements incorporated white Gaussian noise to mimic real-world sensors.
• Evaluated performance across four sensor configurations with varying sampling rates.

Conclusions:

• The proposed state estimator accurately estimates the work cycle and hydraulic pressures in coupled multibody systems.
• Demonstrated the efficacy of the indirect Kalman filter extension for hydraulically actuated dynamics.
• The method effectively handles modeling errors and sensor noise in complex systems.