Constrained multiobjective optimization design for ordinary shovel attachment of hydraulic excavator based on evolutionary algorithm

Abstract

Shovel attachment is one of the most commonly used working attachments in hydraulic excavators. Its mechanical design remains a challenging optimization problem. To tackle this issue, the optimizing model of ordinary shovel attachment is firstly established which contains twenty-three design variables, six objective functions, and sixty-three constraints. Then, an improved decomposition-based constrained many-objective evolution algorithm is proposed, which integrates advanced push & pull search constraint handling technique and differential evolution operator to deal with the optimization problems of shovel attachment. Finally, the proposed algorithm has demonstrated its superior optimization ability for designing shovel attachments through the case studies of a 70-ton face-shovel hydraulic excavator, outperforming seventeen state-of-the-art constrained multiobjective evolutionary algorithms. Furthermore, the first quantitative comparison between ordinary shovel attachment and TriPower shovel attachment under the same main machine is presented by the optimized results. The result demonstrates the complexity of the optimal design for shovel attachment and the effectiveness of multiobjective evolutionary algorithm.