An improved NBA-based STEP design intention feature recognition

STEP design intent feature recognition is an important way in the research field of STEP semantic exchange in the computer aided design and manufacturing. The paper proposes an improved NBA algorithm to optimize BP neural network algorithm for semantic feature recognition. At first, an improved AAG is put forward. It can extract the compound features by the aggregation of similar features factors which the traditional AAG method cannot do. Then the compound feature coding method is given. The neural network semantic feature recognition based on improved NBA Algorithm is proposed. At last, simulation results show the effectiveness of the proposed algorithm.