Research On Multi-objective Real-time Avoidance Decision Based On Trial Manoeuvring In ECDIS

Marine transportation has many advantages such as strong carrying capacity and low average transportation cost,and has an important economic position in the transportation field.At the same time,the maritime navigation conditions are complex and variable,and once a sea accident occurs,it will bring huge losses.According to the statistics of the Maritime Organization,about 80%of marine accidents are caused by humans.In order to reduce the human error caused by the driver's mental,environmental and other unfavorable factors,the comprehensive decision-making by computer and the provision of auxiliary decision-making information have gradually become the application hotspots in the shipbuilding industry.At the same time,with the continuous development of computer simulation technology,the accuracy of ship position prediction using ship numerical calculation is also increasing;and the cost of computing unit is also decreasing,making parallel computing simulation prediction technology gradually meet the needs of engineering applications..Therefore,this paper combines the trial-and-error parallel simulation with the path planning algorithm to study the real-time object avoidance algorithm.Based on the electronic chart,the evasion test platform is developed to provide the driver with visualized ship avoidance assistance.Decision making tools.In this paper,we pioneered the combination of the ship motion mathematical model and the path planning algorithm through the similarity measure and which is applied to the ship collision avoidance engineering field.The collision avoidance algorithm based on the ship trial manoeuvring idea is explored.We first use numerical calculation to solve the ship motion model,and realize the trajectory parallel prediction function of the ship trial manoeuvring ship position.Then,the path planning algorithm of path smoothing is studied.The flow field modeling of real-time collision avoidance problem is completed.;then study the curve similarity measurement algorithm,compare the similarity between multiple trial trajectories and planned trajectories,and select the trial trajectory with the highest similarity,and the trial manoeuvring command corresponding to the highest similarity trajectory is used as the recommended instruction under the current conditions.Finally,this paper builds a ship trial manoeuvring simulation and avoidance decision verification platform.The main work done in this paper is as follows:(1)The ship real-time path planning algorithm is designed to automatically generate the recommended path.The algorithm models the navigation obstacles in the path planning problem as obstacles in the pipeline flow,and models the starting point and the end point of the path as the entrance and exit of the pipeline flow respectively,and solves the two-dimensional flow field by the lattice Boltzmann method.Using C++programming,the final recommended path is graphed using GDI+,and the A*algorithm and ant colony algorithm are used for comparison experiments.The results show that the curve path generated by the flow field method is more continuous and smooth.(2)Using the Douglas-Peucker method to simplify the trajectory curve and calculate the similarity between the curves,a planning-decision algorithm based on curve similarity is designed.The weighted similarity measure is used to comprehensively consider the distance between curves and the shape of the curve.The proposed path algorithm and the collision detection mechanism in graphics show the recommended optimal ship maneuvering recommendations.(3)Using C++to complete the construction of the trial manoeuvring simulation and avoidance decision-making verification platform,the ship trial control module was realized.The simulation platform built the basic ship motion Nomoto model to calculate the ship's trajectory.In addition,the simulation platform can be connected to other ship motion simulation models.Based on the similarity measurement theory,this paper studies the real-time collision avoidance algorithm by parallel simulation technology.The experimental results show that it is feasible to apply the curve similarity measurement theory and parallel simulation technology to the ship collision avoidance problem.