Research On Computationally Expensive Optimization Problems Based On Modified Pigeon-inspired Algorithms

It is often essential for a robot to sense and be familiar with the circumstance when entering an unknown environment with the task of intelligent planning.On this occasion,it is necessary to model and make a thorough comprehension of the environment.And this process can be implemented by producing a semantic map based on the semantic annotation to an environmental map.As is known,a detailed,accurate and deep-level semantic map is the key factor for the robot to complete its corresponding tasks and the crucial issue is to make the robot locate itself without the assistance of the map.In another word,it is important to note that the robot could form the semantic map with the combination of the estimation of its position and related sensors in an unacquainted circumstance.This paper mainly talks about how to generate the semantic map and position the robot..More specifically,there are three subtasks including the common map generation about the unknown environment,the semantic map generation and the mission planning based on the established semantic map.Main contributions are described as follows:Firstly,we analysis the sensors of the robot and the motion errors.Then we put forward a rapid constructive way to establish the semantic map based on the robot dead reckoning.Secondly,on the foundation of the established common map,we additionally utilize the robot's voice module to solve the problem of acquainting itself with the environment and making the semantic map simultaneously.To be specific,the robot is made to recognize the information of objects in the environment through man-machine dialogue.After that,we give some explanations and tags of the semantic information in order to build a semantic map of the environment.Finally,the robot can receive the users' instructions.By the way of word recognition and English segmentation,the robot is able to extract the type and location of subjects from the instructions.Furthermore,we perform a path planning algorithm on the established semantic map to complete the tasks,i.e.,program the ideal routine to the destination...