Research On Agricultural Machinery Automatic Navigation Based On Wireless Sensor Networks

With the development of artificial intelligence,the agriculture of our country has stepped into intelligentized stage,and they are two vital elements for the artificial intelligence that the perceptual technology and the automatic drive technology.The agricultural machinery path planning not only accords with the norms and requirements,but also has higher efficiency.At present,farmland wireless sensor network is used only for the acquisition of data and location information of data source.The position of unknown node in farmland is usually obtained by GPS or other sensor devices.Although various localization technologies have emerged in the area of agricultural machinery automatic navigation,they all have defects and shortcomings and they are costly.Accordingly,based on the researches,the thesis carried out the following researches on agricultural machinery automatic navigation :Firstly,agricultural machinery common operation models are analyzed.Advantages and disadvantages of various operation modes and their suitable farmland environment are summarized.The time consumed of various operation modes is analyzed.The basic positioning principles and error models of GPS and SINS and the common wireless sensor network localization algorithm are analyzed.The advantages and disadvantages of common localization algorithms are summarized.With the analysis of the cost of the equipment,the complexity of the method and the positioning performance,the location algorithm based on RSSI ranging is chosen to obtain the location of agricultural machinery.Secondly,through researching on the path planning of agricultural machinery,a path planning method based on adaptive genetic algorithm is designed.The model of farmland environment is established by grid method,and the barrier-free space is divided into several sub-areas.The performance of different operation modes of agricultural machinery is compared and it is verified that the nested operation mode has better performance.Finally,the path planning of farmland is carried out by using sub-area interlacing method based on adaptive genetic algorithm.Thirdly,the RSSI values of different antenna heights and diffierent carrier frequencies are obtained by the actual experiments in wheat field and corn field.According to the least square principle,the relation between RSSI value and transmission distance is obtained.A hybrid location algorithm of maximum likelihood and weighted centroid is designed in this thesis and the simulation results are compared under the condition of path loss model in wheat field.Simulation results show that the positioning performance of MLE-WCL is betterthan WCL and MLE.The SINS/GPS/MLE-WCL integrated navigation system is designed.The simulation results show that the performance of SINS/GPS/WCL-MLE integrated navigation is superior to SINS/GPS and SINS/MLE-WCL.Finally,the position real-time tracking system based on VC++ contains MapX control is designed.Location tracking is realized by the registration of farmland grid image and the agricultural machinery location is imported as point element.Track playback is realized by the ODBC binding between agricultural machinery location database and MapX.