An Improved Thrust Allocation Method Based on Genetic Algorithm and Pseudo Inverse

Abstract

This paper addresses the problem of real-time thrust allocation for Dynamic Positioning (DP) marine vessels. An novel thrust allocation strategy based on pseudo inverse and genetic algorithm is proposed, which can eliminate thruster chattering and reduce the optimization time. First, a more realistic thrust allocation model is established by taking fuel consumption, singularity avoidance and actuator wear reducing into consideration. Then, through using pseudo inverse of the thruster configuration matrix, the two-variable optimization problem is simplified into two single-variable optimization problems. And genetic algorithm (GA) is utilized to find the optimal direction, while Quadratic Programming (QP) is used to find the suboptimal propeller force. Simulation examples demonstrate that the proposed thrust allocation strategy has better search performance and cost less time than traditional GA optimization.