Multi-sensor integration system based on fuzzy inference and neural network for industrial application

Abstract

The authors deal with a multi-sensor system applied to an unknown curved metal surface cutting robot system. The measurements were performed by sensors set on an array of the tip of a five axis manipulator. The sensor array is carried to the target surface by moving the manipulator. The manipulator approaches the surface by using sensor outputs. To approach the work fast, the system should use long measurement range sensors. For precise cutting and a fast approach, the system should use both high accuracy sensors and long measurement range sensors. To use these sensors effectively, the multi-sensor integration system was based on neural network and fuzzy inference techniques. As a result, the system can consider the angle between the sensors and the object. The proposed system was shown to be effective through extensive experiments.<>