Abstract: Aiming at the lack of effective means for the assembly quality inspection of the combine harvester and the low accuracy of the existing methods, a method for the assembly quality inspection of the combine harvester based on the improved whale algorithm optimized least square support vector machine (IWOA-LSSVM) was proposed. First of all, in view of the weak search ability and early maturity of the whale optimization algorithm (WOA), the cosine control factor and the sine time-varying adaptive weight were introduced to improve, and the benchmark function was used to verify the general adaptability of the algorithm.Then, the signal was denoised by complementary ensemble empirical mode decomposition (CEEMD), and the information entropy of each component and the time-frequency domain features were extracted to construct a fusion feature set.Finally,an assembly quality detection model of combine harvester based on IWOA-LSSVM-CEEMD was established, and it was tested on Dongfanghong 4LZ-9A2 combine harvester to verify the validity of the detection model; the method was compared with other methods. The research results show that the average prediction accuracy rate of the method proposed is 91.3%, which is an increase of 7% compared with the unimproved method, and the standard deviation of the average prediction accuracy rate is reduced by 0.68%. The results verify the superiority of the proposed method in the application of combine harvester assembly quality inspection.

Key words: agricultural machinery; mechanical assembly quality inspection; improved whale optimization algorithm (IWOA); least square support vector machine (LSSVM); complementary ensemble empirical mode decomposition (CEEMD); signal noise reduction decomposition