Torque estimation in marine propulsion systems. (1st June 2022)

• Record Type: Journal Article
• Authors: Manngård, Mikael; Koene, Ivar; Lund, Wictor; Haikonen, Sampo; Fagerholm, Fredrik A.; Wilczek, Michał; Mnich, Konrad; Keski-Rahkonen, Joni; Viitala, Raine; Björkqvist, Jerker; Toivonen, Hannu T.
• Is Part Of: Mechanical systems and signal processing. Volume 172(2022)
• Publication Date: 2022-06-01
• Subjects: Simultaneous input and state estimation -- Optimal filtering -- Maritime systems; Structural dynamics -- Periodicals; Vibration -- Periodicals; Constructions -- Dynamique -- Périodiques; Vibration -- Périodiques; Structural dynamics; Vibration; Periodicals; 621
• DOI: 10.1016/j.ymssp.2022.108969
• Languages: English
• ISSNs: 0888-3270

Abstract

Abstract: An augmented Kalman filter for torque estimation in marine propulsion-system drive trains is presented. Propeller and motor excitations and torque responses are estimated based on a dynamical model of the system and inboard shaft measurements. Input excitations affecting marine propulsion systems are signals whose statistical properties vary between finite time intervals. Hence, in this paper, excitations are characterized as quasi-stationary signals with bounded power spectral density. Given that upper bounds on the spectral densities are known prior to estimation, it is shown that a linear time-invariant input-and-state observer, minimizing the worst-case power of the estimation errors, can be synthesized by conventional Kalman-filtering techniques. Experiments have been conducted on a laboratory-scale test bench to assess the applicability of the proposed observer for use in marine propulsion systems. The test bench was built to emulate the behavior of a full-scale propulsion system operated in ice and other high load conditions. Estimation results from a full-size underwater mountable azimuthing thruster are also presented. Experiment results show that torque excitations and torque responses at all locations of interest on the engine-propeller drivetrain can be estimated with high accuracy based on a few indirect measurements at convenient locations on the motor shaft.