Nonlinear Algorithms for Combining Conflicting Identification Information in Multisensor Fusion
Jeffery D. Hurley, Clint Johnson, Joel Dunham, Jimmy Simmons · 2019
One of the most difficult problems in advanced systems today is handling all the identification information that is received from an array of sensors. Dealing with all this information becomes even more of an issue when there is conflicting information. Historically, evidence theory has been used to combine information from different sensors, but handling conflicting information in an intuitive manner continues to be a challenging problem. Often when conflicting information is detected the set of information is processed for outliers in order to remove the conflicting information before fusing the remaining set. This paper describes the use of a nonlinear algorithm in conjunction with the foundations of evidence theory to handle the combination of sensor data in an intuitive manner while also managing conflicting information. The results from a novel nonlinear algorithm is contrasted and compared with results from other modern day evidence theory algorithms.