선박 실운항 데이터 기반 제어이득 자동 조정 알고리즘 개발

Abstract

In this study, an automatic gain tuning algorithm for ship was proposed. The proposed algorithm is designed to tune the control gain of the ship controller automatically rather than trial and error. The forward speed model and steering model were derived from linearization and simplification of the 3-DOF nonlinear equation of motion. System identification was performed based on the derived dynamic models and ship operation data. The control gain was derived from an error model constructed by combining the steering model and controller system model. It was automatically tuned by performing maneuvering simulations and sensitivity analysis on the control performance at various control gains. It was verified that the algorithm corrects the gain more accurately and faster than trial and error through outdoor free running tests. Based on this study, it was possible to confirm the accuracy of the simple dynamics model and reduce the time and cost required for tuning.