Abstract: The method for compressor efficiency prediction and parameter optimization by establishing the reciprocating air compressor digital twin model has the advantage of flexibility, low cost, and good versatility. However, the traditional twin model based on the BP neural network (BPNN) has lots of shortcomings, such as longer training time to establish a module, easily falling into the local optimal solution, and difficulty in achieving the global optimal solution. To solve these problems, a novel digital twin model based on the CIWOA-BPNN algorithm is put forward to determine the key indexes by the principal component analysis method, in which a CIWOA algorithm is introduced to improve the BPNN’s performance. The results show that the new CIWOA-BPNN twin model effectively avoids falling the local optimal problem. The relative error of CIWOA-BPNN is less than 0.6%, and the coefficient of determination is 0.997 75, which greatly improves the prediction accuracy compared with the traditional model.

Key words: reciprocating air compressor, efficiency, BP neural network, improved whale optimization algorithm