Intelligent Sensor Validation And Sensor Fusion For Reliability And Safety Enhancement In Vehicle Control
Alice Merner Agogino, Kai Goebel, Satnam Alag · RePEc: Research Papers in Economics · 1995
In this report we present an evaluation of methods for validation and fusion of sensor readings obtained from multiple sensors, to be used in tracking automated vehicles and avoidance of obstacle in its path. The validation and fusion is performed in two modules which are part of a larger five-module hierarchical supervisory control architecture. This supervisory control architecture operates at two levels of the Automated Vehicle Control Systems (AVCS): the regulation and the coordination level. Supervisory control activities at the regulation layer deal with validation and fusion of the sensor data, as well as fault diagnosis of the actuators, sensors, and the vehicle itself. Supervisory control activities at the coordination layer deal with detecting potential hazards, recommending the feasibility of potential maneuvers and making recommendations to avert accidents in emergency situations. In this grant we formulated the need for an hierarchical approach and then focussed in depth on the two modules sensor validation and sensor fusion. Tracking models were introduced for the various operating states of the automated vehicle, namely vehicle following, maneuvering, i.e. split, merge, lane change, emergencies, and for the lead vehicle in a platoon. The Probabilistic Data Association Filter (based on Kalman filtering) is proposed for the formation of real time validation gates and for fusing the validated readings. A topology for an influence