Abstract
A two-stage computational framework is proposed to optimize the radiated noise and weight of a large mining planetary gear reducer under the rated conditions, based on a combination of response surface methodology and multi-objective optimization. The well-established transient dynamic analysis model of a large mining planetary gear reducer, which is used to analyze the mechanical strength and acoustic characteristics of the gear reducer. A unified experimental design is developed to obtain the response surface of the gearbox radiated noise and the mass of the gearbox housing. After obtaining the multi-objective optimization function, the multi-objective optimization problem for a lightweight and low-noise gearbox is performed using non-dominated sorting from the Genetic Algorithm II (NSGA-II). The research results demonstrates the effectiveness of the proposed optimization method in reducing vibrating amplitude and weight of the gearbox. This is crucial for minimizing energy consumption and enhancing the overall performance of the system. Additionally, the optimized gearbox design not only saves energy but also contributes to the reduction of carbon emissions, making it environmentally friendly.