A data fusion algorithm based on adaptive neural networks-fuzzy inference system(ANFIS) and D-S evidence theory is proposed.The establishment of the basic probability assignment is solved
which is rather difficult in evidence theory.Neural networks are used to extract fuzzy rules and adjust the fuzzy membership function
and then the outputs of ANFIS can be transformed into the basic probability assignment in evidence theory.Fuzzy inference
neural networks and evidence theory are integrated effectively.Evidence theory is provided with the ability of intelligent information processing.The algorithm is implemented for target recognition on IR/MMW Radar compound guidance system.The simulation results show that the degree of confidence is increased and the rate of false alarm is decreased
furthermore the result of the fusion is more reasonable.