Research On Safe Navigation Of Mobile Intelligent Agents Based On Internet Of Things

Wireless sensor network(WSN)is a distributed network system consisting of many self-organized sensor nodes deployed in monitoring areas,which is an important supporting technology of the Internet of Things.Safe emergency navigation is a new application of wireless sensor network,which can guide the mobile robot and other agents on the site to reach the target safely in case of emergency.It has a broad application prospect in large intelligent unmanned warehousing systems.Aiming at the problems of low dynamic adaptability and navigation efficiency of traditional emergency navigation algorithms,this thesis is devoted to the research of a dynamic emergency navigation method for mobile agents based on wireless sensor network.The main research contents are as follows:(1)Aiming at the shortcomings of the traditional emergency navigation algorithms,a global emergency navigation algorithm for mobile agents based on wireless sensor network is proposed.In this algorithm,a prediction model based on time series is introduced to predict the dynamic changes of the environment sensed by WSN,and then the potential field of each node is established by comprehensively considering its predictive danger value and the distance to the exit.Based on the dynamically updated potential field,the mobile agents exploit the gradient descent method to efficiently approach the target areas node by node.The simulation results demonstrate the superiority of the proposed algorithm in terms of the navigation efficiency and safety in dynamic environments.(2)Considering that the emergency navigation algorithm based on single-hop prediction which only uses the environmental information within the single-hop communication range of the node has certain limitations,this thesis further proposes a dynamic global emergency navigation algorithm for mobile agents based on multi-hop prediction via wireless sensor network.In the process of navigation decision,this algorithm considers the changing trend of the environment around the nodes in the range of multi-hop communication.In order to verify the feasibility and performance of the algorithm,the performance of the emergency navigation algorithms based on 1-hop prediction,2-hop prediction and 3-hop prediction are compared through simulation experiments.(3)Aiming at the problems of unknown location information and the failure to consider obstacle avoidance in local navigation in wireless sensor network,this thesis also proposes a local navigation algorithm based on wireless sensor network,including path node localization and obstacle avoidance algorithms.Firstly,a location algorithm based on iterative trilateration localization method based on RSSI ranging is proposed to obtain the location information of the path node.Then based on the known location information,the dynamic window approach is exploited to realize the local navigation of mobile agents avoiding obstacles and dangerous areas.(4)Finally,the experimental platform for safe navigation of a mobile agent with ROS based on WSN is built,and the proposed theoretical method is verified by real experiments.Experiments show that the dynamic emergency navigation method proposed in this thesis has high accuracy and adaptability.