A new combination method for multisensor conflict information

Information fusion is a very important technology which can use multisensor network data to get a better performance than single one; therefore, it is widely used in the filed of target recognition, target tracking, automatic control, decision making and so on. However, because of noise and interference, sometimes the sensors may obtain erroneous, inaccurate or heterogeneous data, which will produce the conflict information among different sensors and get the wrong result after information fusion. In this paper, based on the Dempster–Shafer (D–S) theory, we introduce how to set up the model of multisensor network information fusion. And then, we discuss the problem of conflict information fusion in the framework of evidence and several improved methods are introduced. Finally, based on Mahalanobis distance, an improved solution method is presented. The numerical simulation results prove that this new improved method can get the same result as traditional methods, beyond which it can make a reasonable decision with high conflict information. Therefore, this new improved method can be used in the filed of high noise and interference.