| Recently,wearable sensor network that is a kind of new wireless network comprisingseveral sensors to monitor attitude in real time and track the information on the target carrier suchas hunman body has turned into a hot research topic about information science and computertechnology. The algorithm of dissimilar sensors data fusion researched in this thesis is setted inthe wearable sensor network.The algorithm of dissimilar sensors data fusion based on attitude estimate reduces noise byfilter algorithms to improve the computational accuracy.Depengding on the different applicationenvironments,the carrier attitude can be represented by a variety of methods such as eulerangles,quaternion and rotation matrix,and these representations can be converted between eachother.Concerning the computational methods of the carrier attitude,the thesis researchrespectively for the acceleration sensor,gyroscope and magnetic sensor,and also analyse theapplication and methods of the three sensors in attitude estimate. About filter algorithms,thisthesis focus on the widely used linear dispersed Kalman filter algorithm, the extended Kalmanfilter algorithm and the unscented Kalman filter algorithm for the nonlinear system,while furtherresearch the data fusion algorithms based on adaptive filter,including Least Mean Square (LMS)algorithm and Recursive Mean Square (RLS) algorithm and compare the difference of all thealgorithms by the final simulation results.Based on LMS algorithm and the methods of attitude estimate of the accelerationsensor,gyroscope and magnetic sensor, this thesis proposed a adaptive complementary datafusion algorithms based on LMS. In order to prove the validity of the proposed algorithms,thethesis design an experimental scheme rationally, and simulated the experiment for the differentalgorithms.The experimental result shows that the algorithm achieves high accuracy,shortdelay,reducing computation complexity and little storage space compared to the exisingalgorithms. |
PDF Full Text Request