An effective selective assembly model for spinning shells based on the improved genetic simulated annealing algorithm (IGSAA)

For cylinder shell parts produced in batches, computer-aided selective assembly can not only obtain higher product matching accuracy, but also reduce the remaining number of parts, ensuring the welding assembly quality and improving the production efficiency. Aiming at the selective assembly problem for spinning shells with electron beam welding, a selective assembly model based on an improved genetic simulated annealing algorithm was proposed. By analyzing the assembly process characteristics of spinning shells, mapping association matrix of assembly constraints was built to describe the assembly relationship between the different cylinder of spinning shells. Considering the multi-assembly quality loss function using SNR and assembly yield, a multi-objective comprehensive optimization model was established. Based on the measured internal diameter of the parts, a specific coding method and the adaptive cross mutation operator based on the sigmoid curve is introduced to apply an improved genetic simulated annealing algorithm (IGSAA), solving the assembly selection problem of 5 shell parts case. The results show that the model established has a good applicability to the spinning shell parts matching problem, which can effectively improve the success rate of parts matching and assembly accuracy, and meet the production needs of enterprises. Moreover, the produced assembly difference through improved genetic simulated annealing algorithm (IGSAA) is even better than manual selection in matching accuracy and efficiency.