Method for planning route of unmanned ship based on SVM algorithm

Abstract

The invention belongs to the field of unmanned ship route planning, and particularly relates to a method for planning the route of an unmanned ship based on the SVM algorithm. The method comprises the following steps: constructing a standard unmanned ship, and putting the standard unmanned ship to sail in a typical sea area; searching standard related data when the standard unmanned ship encounters a barrier in the typical sea area, and saving the standard related data into a database; calculating by utilizing a least square SVM algorithm according to the database to obtain a complete sample bank; searching actual related data when the standard unmanned ship encounters a barrier in the typical sea area when the unmanned ship sails in the actual sea area; comparing the actual related data with the data in the complete sample bank to obtain a corresponding actual sample model; and planning the route of the standard unmanned ship according to the actual sample model. The method enables the planning result to be high in precision and small in error under the condition of limited sample data, and has good independence and strong stability.

Description

technical field [0001] The invention belongs to the field of unmanned ship path planning, in particular to an unmanned ship path planning method based on SVM algorithm. Background technique [0002] With the continuous application of artificial intelligence, unmanned ship-related technologies have developed rapidly, but the development of domestic unmanned ships is still in its infancy, mainly used in meteorological monitoring, environmental protection and other fields, which have a certain impact on people's life and work influences. Due to the weak environmental awareness and autonomous navigation control capabilities of domestic unmanned ships, and there are many interference factors, they are far from meeting the practical requirements, so that it is necessary to use multi-objective autonomous path planning methods to break through the bottleneck of unmanned ship research . [0003] Most of the current unmanned ships use human remote control or use GPS navigation syste...

AI Technical Summary

Problems solved by technology

However, if the cost is controlled and the number of samples is reduced, the path error planned by the above algorithm will be large, and the motion accuracy cannot be guaranteed. Moreover, there are many interference factors at sea, which will greatly affect the motion path accuracy.

Therefore, there is no path planning method for unmanned ships with high path optimization accuracy, small error, good autonomy, strong stability, and less sample data.

Images