Improved NSGA-II algorithm for constrained thrust allocation of dynamic positioning ships in rough sea conditions

Abstract

Thrust allocation is one of the key technologies in dynamic positioning (DP) systems. Accurately allocating the thrust and angle of each thruster will achieve the desired force and moment of the ships, which is crucial for improving the positioning accuracy and positioning performance of marine ships. This paper develops an improved Non-dominated Sorting Genetic Algorithm (NSGA-II) to handle thrust allocation problem of dynamic positioning (DP) system in rough sea conditions. Firstly, the multi-objective optimisation model is built considering the thrust prohibited area, output angle, output thrust, angle change rate and thrust change rate as the constraints, and considering the power consumption, thrust error and thruster wear, and singular structure penalty term as the optimisation objectives simultaneously. Then, an improved NSGA-II algorithm is proposed to optimise the selection process of Pareto-optimal solution. Finally, multiple simulation results show that improved NSGA-II has better optimisation performance compared with basic NSGA-II. All of the comparisons could fully be conducted to demonstrate the effectiveness and advantages of the proposed algorithm. Abbreviations: DP: Dynamic Positioning; NSGA-II: Non-dominated Sorting Genetic Algorithm; SQP: Sequential Quadratic Programming; GA: Genetic Algorithm; PSO: Particle Swarm Optimisation; ABC: Artificial Bee Colony; AHABC: Adaptive Hybrid Artificial Bee Colony.