Intelligent vehicle SLAM data association method based on improved artificial fish swarm algorithm

Abstract

The invention discloses an intelligent vehicle SLAM data association method based on an improved artificial fish swarm algorithm. The method is characterized by firstly, using an independent compatible criterion and a combined maximum likelihood criterion to determine an association hypothesis and converting a SLAM data association problem into a combined optimization problem; secondly, using an improved artificial fish swarm algorithm based on a jump behavior and a taboo strategy to solve the combined optimization problem, and solving an optimal data association set; introducing the jump behavior in the artificial fish swarm algorithm so that one part of artificial fishes jump out of a local extremum and global optimum is reached as far as possible; then using the improved artificial fish swarm algorithm based on the jump behavior to search a global suboptimal solution and taking the global suboptimal solution as an initial solution of a taboo search algorithm; and using the taboo search algorithm to search a local optimal solution so as to enhance a global optimization capability and optimization efficiency. By using the method of the invention, in a large outdoor range scene, an intelligent vehicle SLAM data association problem is effectively solved, a correct rate of data association and search efficiency of the optimal association set are increased, and operation time is reduced.

Description

technical field [0001] The invention belongs to the field of autonomous navigation and positioning of intelligent vehicles, and relates to a data association algorithm for simultaneous positioning and mapping of intelligent vehicles, and more specifically, relates to an intelligent vehicle SLAM data association method based on an improved artificial fish swarm algorithm. Background technique [0002] Simultaneous localization and mapping (SLAM) can be divided into two parts: state estimation and data association. As the premise and foundation of state estimation, data association provides correct input for state estimation, ensures the real-time and accuracy of SLAM, and plays a vital role in SLAM. [0003] At present, the data association algorithms in smart car SLAM include Independent Compatibility Nearest Neighbor (ICNN), Maximum Likelihood (ML), Multiple Hypothesis Tracking (MHT) and Joint Compatibility Branch and Bound Algorithm ( Joint Compatibility Branch and Bound,...

AI Technical Summary

Problems solved by technology

[0004] In order to overcome the technical problems in the SLAM data association algorithm that the accuracy of association is greatly affected by the environment, the overall performance of association matching cannot be effectively guaranteed, and the real-time application of large-scale association is limited, and aiming at reducing the complexity of the algorithm and improving the accuracy of association, the present invention A smart car SLAM data association method based on the improved artificial fish swarm algorithm is proposed. Using the advantages of the improved artificial fish swarm algorithm to solve the combinatorial optimization problem, combined with the determination criteria of the compound association assumption, the improved artificial fish swarm algorithm is applied to the search SLAM. In the optimal association solution, the search efficiency of the optimal solution is improved under the premise of ensuring the correct rate of association, that is, the running time of the algorithm is reduced

Images