Research On Multi-Ship Collision Avoidance And Path Planning Based On Electronic Chart

The intelligent ship is of great significance in the aspects of energy conservation,emission reduction and ship safety.The International Maritime Organization has incorporated the Maritime Autonomous Surface Ships(MASSs)into her work plan and is actively promoting it recently.One key and difficult point in researching of intelligent ships or MASSs is automatic collision avoidance(CA)for ships in complex waters.For present researches,some prominent problems still exists:without consideration of the International Regulations for Preventing Collisions at Sea(COLREGS)and the uncoordinated CA actions for target ships,applying the assumption of COLREGS-constrained actions even constant course&speed status for all target ships;without using the electronic chart platform,and not applicable to the long path plan and complex waters with dynamic ships and the static obstacles.In view of this,the thesis focuses on the following aspects:First,this thesis proposes a real-time and deterministic automatic CA method for autonomous ships in open water,under considering ship's dynamic properties and the COLREGS.This method applies a modified Artificial Potential Field(APF)model,which contains a new modified repulsion potential field function with respect to the subdivided CA zones and definitions based on the ship's domain and the core rules of COLREGS.Also,a global gravitational potential field is constructed in this model,which is suitable for dynamic goal tracking and static goal guidance.By combining the above,a new repulsive force and gravitational force model is determined so as to satisfy the requirement of CA and path planning for MASSs.At last,according the navigation practices,constraints of dynamic properties of ships and virtual forces of subdivided CA zones,a real-time automatic CA algorithm is developed.Simulations in MATLAB show that the method is fast,effective,and deterministic for path planning in complex situations with multiple moving target ships and stationary obstacles and can account for the unpredictable strategies of other ships.Second,based on the point,line and face vector data and their corresponding characteristic attributes in electronic charts,a method about environment potential field map construction and its parameter setting are presented.It's noted that the modeling accuracy of the raster-based environment map is not high and difficult to distinguish different kinds of obstacles.The author provides a method to construct a visual potential field environment map which can conveniently control the influence range of the different kinds of obstacles according their attributes.The method can establish an environment potential field for static obstacles with irregular shapes(such as an implicit function curve,a convex polygon and even a concave polygon),and can control the safe distance between the MASS and these obstacles by simple parameter adjustment according to their attributes.This method can improve the modeling precision of the environment map,and is easy to be combined with the multiship CA algorithm.Additionally,it lays the foundation for the CA decision-making algorithm based on the electronic chart.Third,a path-guided hybrid artificial potential field(PGHAPF)method is first proposed to resolve the multiship CA problem in restricted waters including dynamic target ships and static obstacles.The proposed strategy,which is a fusion of the potential field and gradient methods,consists of potential-based path planning for arbitrary static obstacles(with high priority),gradient-based decision-making for dynamic target ships and goals,and their combination with consideration of dynamic model of a Mariner class vessel,the prior path and waypoint selection optimization.By detecting the danger of the high potential field value of the static obstacle in front of the own ship,to determine the direction of the minimum value of the nearby potential field,and to make the own ship to take actions in advance so as to keep the safe distance with obstacles.Through gradient-based calculation to find the CA decision of course change,which is constrained by COLREGS and goal guidance.Simulations show that the PGHAPF method can rapidly generate adaptive,collision-free and COLREGS-constrained trajectories in restricted waters by deterministic calculations.Furthermore,this method has the potential to perform path planning on real-ship system and to overcome some drawbacks of traditional APF methods,such as local minima and the "Goals Non-Reachable with Obstacles Nearby"(GNRON)problem.Fourth,the multiship CA decision-making module and the dynamic link library based on the PGHAPF are established and integrated in "full mission ship handling simulator test platform".In this platform,the multi-ship real-time automatic CA and path planning based on the electronic chart is realized.In addition,the coordinated or uncoordinated CA actions for multiple MASSs and non-autonomous ships are also realized in the restricted waters.A large number of simulation experiments show the adaptability of the algorithm for multi-scale ship types,and the robustness of the algorithm under the influence of winds and currents.The proposed methods overcome some limitations of the existing researches,and provide the basis for the multiship automatic CA algorithm testing and application in the real-ship system.