Research On Robot Chemical Plume Tracking Based On Multi-source Information Fusion

Researchers have looked at organisms in large nature and found that many of them use chemicals in the air to sniff for prey,food and spouses.In addition,with the rapid development of the industrial economy,the human face the environmental problems caused by the rapid industrial development,such as industrial toxic gas leakage,emissions-induced air pollution,ecological environmental damage.At the same time,countries around the world are also facing terrorist attacks,inflammable and explosive materials,major fires and other accidents.Therefore,it has important scientific significance and practical application value to study the special robot which can help human to carry out fire search and rescue,remove explosive,earthquake rescue and detect contraband in complex and arduous environment.At present,the study of chemical plume tracking using mobile robot has become a hotspot in this field,and some research results have been obtained.However,at present,researchers use a single olfactory sensor or wind direction sensor to design the plume source search strategy,because of sensor performance,gas flow law and other factors affect the plume tracking effect.In order to better eliminate these factors,improve the accuracy of chemical plume tracking.Therefore,the method of chemical plume tracking based on multi-source information Fusion is studied in this paper.This paper is based on the Natural Science Foundation of Xinjiang Uygur Autonomous Region(serial number 2016d01c032),with the help of the research on the robot chemical plume tracking of multi-source information Fusion,its main work is as follows:1 In order to improve the effect of the chemical plume tracking,this paper uses a variety of sensors,and uses the multi information fusion technology to fuse the data collected by the sensor,and proposes a method of chemical plume tracking based on multi-source information Fusion.A multi information fusion model of robot CPT is established,and the fusion of multi-sensor information with fuzzy integration algorithm is adopted to realize the chemical plume tracking of robot.The data of the chemical concentration sensor,the vision sensor and the infraredobstacle-avoidance sensor are processed and extracted,the fuzzy membership degree is determined and the fuzzy confidence degree of the target is determined by the fusion rule,which provides the support for the robot to locate the chemical plume source.(2)design a rule-based robot CPT Intelligent Decision System,through the design of the rules,knowledge and inference mechanism of chemical plume tracking,the mobile robot can independently make decision to carry out corresponding tracking behavior,including chemical plume search,Tracking and plume source confirmation process.(3)The fusion of visual information to the chemical plume tracking process,the establishment of a chemical plume source based on the color model of the knowledge base,and through the chemical plume source for color feature extraction,based on the HSV color model of the 9 A color eigenvector and the minimum threshold for image matching.At the same time,we designed a chemical plume Source Model library to provide support for the determination of chemical plume source.(4)The overall design of the two-dimensional space CPT robot is designed,and the hardware circuit and software program of the CPT Robot are developed,and the hardware diagram of the double H Bridge Motor Drive control is given.Each function module is studied and the driving flowchart is given,and the driver is programmed.In addition,the three-dimensional space CPT robot is preliminarily explored,the autonomous tracking behavior principle model of three-dimensional CPT robot is studied,and the structure and hardware composition of the three-dimensional CPT robot are discussed.(5)based on the experimental verification of the robot chemical plume tracking strategy of multi-source information Fusion,the experimental scheme is designed according to the requirement of the robot chemical plume tracking strategy of multi-source information,and the experimental platform is built,and the Arduino IDE For the development of the environment to compile debugging,through experimental verification,the chemical plume source location error ?e is 0.16,The verifies the accuracy,feasibility and effectiveness of a robot chemical plume tracking strategy based on multi-source information Fusion.