Application of D-S evidence theory in flow regime identification of two-phase horizontal pipe flow

Identification of two-phase flow regimes is deadly required in many industrial processes and is difficult to perform objectively. Multi-sensor data fusion has been applied to the recognition of flow regimes of air/water flow in horizontal pipe. A specific fusion process architecture based on Dasarathy fusion model is introduced, and D-S evidence theory is adopted on the decision fusion level. The fusion data is provided by dual-plane electrical resistance tomography system in which each plane works as a single sensor. Cross-correlation technique is used to match the data from the two electrode planes. It is shown that this flow regime recognition method in which multi-sensor fusion is adopted is an approval strategy.