Abstract: In response to the issues of reliance on manual expertise,slow straightening speed,and low yield rate of quality products in traditional straightening processes,a straightening intelligent optimization process prediction model based on the Dung Beetle Optimizer (DBO) algorithm optimized Back Propagation (BP) neural network was proposed. Considering the influences of parameters such as plate thickness,elastic modulus,yield strength,and plastic ratio of the plate during the straightening process,as well as the issues of BP neural networks easily falling into local optima and weak generalization ability,the DBO algorithm was introduced.The model was trained using a training set consisting of 1 000 data points. A comparison between the BP neural network prediction model and the particle swarm algorithm optimized BP prediction model was conducted.Results show that the percentage errors of the DBO algorithm optimized BP neural network prediction model for the adjustment amounts of the head and tail rollers are within 0.5 % and 0.6 % respectively,and the total straightening force percentage error is within 0.6 %.The proposed model demonstrates high reference value for accurate prediction of the straightening process.

Key words: straightening process, Dung Beetle Optimizer (DBO) algorithm, Back Propagation (BP) neural network, predic-tion model