Path Planning Of Ships Based On Differential Evolution Algorithm

Path planning for ship navigation or the design of scheming route is an importantcomponent of the ship navigation process. With the great progress of technology andintelligent algorithms, path planning which uses the intelligent algorithm with high degree ofintelligent and adaptability is the major method for ship navigation.According to the technical characteristics of the ship path planning and with thebackground information of ship navigation, ship path panning method is studied in this paper.A novel improved differential evolution algorithm called DE With a ModifiedNeighborhood-Based Mutation Operator (MNDE) is proposed in the paper. The article putforwards a path planning method based on differential evolution algorithm for ship navigationwhich uses the known waypoint library of the area to be planned. Then this paperdemonstrates the superiority of improved algorithm and effectiveness of using it in ship pathplanningFirst, this paper analyses ship navigation and path planning problem. The type of shiproutes and the differences between them are elaborated. Then this article explains therequirements, general steps and main types of the ship path planning, also describes theplanning method to be used.Second, the basic principles and execution processes of differential evolution algorithmare elaborated. Then the article presents a novel improved differential evolution algorithm inaccordance with the characteristics of path planning for ship navigation. Then the controlparameters in the new algorithm are determined according to the situation of ship pathplanning.Next, the thesis expounds the procedure of ship path planning using the differentialevolution algorithm. The following work are included: completing the environmentalmodeling of the region where the ship will sail; the selection of the appropriate encodingmethod to express the route; the population initialization; making a detailed description aboutthe mutation operators in the modified differential evolution method and the selectingmethods of key parameters and operating vector; determining the cost function according tothe requirements of the planning mission and conducting the selection operation; defining theterminating condition of evolution and finishing the path planning progress.Finally, this paper verifies the performance of modified differential evolution algorithm and uses the algorithm proposed above in path planning for ship navigation by severalexperiments in MATLAB programming environment. The validity of the proposed novelalgorithm is proved by experiments results. Furthermore, the modified algorithm proved to besuperior via the simulation results of comparison with the two traditional differentialevolution algorithm strategies.