Research On Approach And System For Mobil Robot Navigation Based On Wireless Sensor Networks

Mobil Robots are expected to do some routine or danger tasks automatically for human in many situations. Among the techniques related to the research of Mobil Robot, localization and navigation are the core techniques that are critical for Mobil Robot to realize true intelligence and complete autonomous moving.The navigation methods that are frequently used in Mobil Robots mainly include inertial navigation, visual navigation, sensor-based navigation and satellite navigation. The inertial navigation method can provide fully independence and complete information for the robot navigation, while its localization error will accumulatively increase as the cruising range increases. The visual method can provide better positioning accuracy, but the view field is quite limited. The sensor-based method is usually used for local navigation and collision avoidance. The Global Positioning System (GPS) is the most accurate among the satellite navigation systems that are in use. Very high positioning accuracy can be achieved if the GPS and the inertial navigation system were used together. The GPS is controlled by the American Department of Defense. Its unauthorized user only can obtain low positioning accuracy without safety. So, it is necessary to study new navigation methods that are independent on the GPS for many applications such as the outer space exploration and military use.In this dissertation, Wireless Sensor Networks (WSNs) was introduced into the research of Mobil Robot. A new navigation approach, as a supplement of current methods, was proposed based on the localization function of Wireless Sensor Networks. In cases that the satellite system cannot or should not be used, this method could instead the satellite navigation method in Mobil Robot to provide the localization and navigation. The various information about the environment such as the temperature, the humidity and the slope of the ground that is obtained by the WSN can be used to help the decision-making in navigation and path planning.The main achievements and innovations in this dissertation include:1. The restriction on energy is the primary challenge on the research and application of WSNs. To sufficiently use the limited energy of the network nodes and effectively extend the life of the network and the navigation system, a fair energy usage routing algorithm for WSNs based on the Ant Colony Optimization Algorithm was studied and proposed. The idea of swarm intelligence was used to realize the efficient transmission of messages and equalized consuming of energy. Simulation result showed that this algorithm could extend the life of the network and the navigation system effectively.2. To make full use of the performance of the fair energy usage routing algorithm, the voltage-based and current-based energy measuring methods were studied and analyzed. Then a voltage- and current-based node energy measuring method was designed according to the energy consuming model and the working characteristics of the nodes in WSNs. This method implemented the precise real-time measuring of the remained energy of the network nodes.3. To realize the localization of Mobil Robot and provide the position information for environment modeling and navigation, the localization method by WSNs was studied. Based on the Chipcon (TI) CC2431, some WSN nodes that with localization function were developed and made. Test on the nodes showed that the localization performance of these nodes were equivalent to that of GPS, which means they satisfied the requirement of robot navigation.4. A WSN-based method of environment modeling and map building was proposed. It utilized the distributed environment information obtained by the WSN to establish the environment model and the grid map with multiple attributes. Simulative analysis showed that the environment model established by this method achieved good match result on the map.5. The dynamic monitoring function of WSN was utilized to adapt the Mobil Robot to the requirement of navigation on the changing environment. An on-line path planning method based on WSNs was proposed. Simulation study showed that it could let the Mobil Robot make trade-off between safety and efficiency through adjusting the parameters used in the algorithm.6. Also in this dissertation, a WSN-based experimental Mobil Robot navigation system was designed and implemented to verifying the proposed navigation approach. The experimental result showed that the proposed approach could fit the application requirement of Mobil Robot navigation.