Abstract: In the search of parallel hybrid hydraulic excavator power matching, a power prediction model is proposed based on Least Square Support Vector Machine (LS-SVM) algorithm for constant power variable duplex pump. Based on the analysis of the principle of variable capacity pump, the load spectra obtained from actual excavator operations are used as training samples, the pump outlet pressure, the pilot secondary pressure and the variable displacement pressure are taken as the input, and the pump displacement is taken as the output to build the model. Such the model has better adaptability to the working environment. The LS-SVM parameters are optimized with grid-search and cross-validation methods. Experiment results show that the proposed model has good prediction accuracy and generalization capacity.

Key words: power machinery and engineering, hybrid excavator, power matching, pump power prediction, least square support vector machine(LS-SVM)