Research On Radiation Noise Analysis And Structural Optimization Design Method For Reducer Of ZZL1150 Mine Hoist

As one of the important components of the mining hoist,the reducer has a significant effect on the overall performance of the hoisting machine,which requires the hoist with high carrying capacity,low vibration,low noise,and high reliability.However,the current problems exist on the domestic mining hoist reducer are large radiation noise and low power density.Therefore,it is great theoretical and practical value for the problem of lightweight and noise reduction of the hoist reducer.The two-stage planetary reducer of ZZL1150 mine hoist is used as the research object.In order to obtain the radiation noise of the reducer,the analysis model of the acoustics of the reducer is established.The radial basis function neural network is used to construct the surrogate model of the noise power and the housing quality of the reducer,and the housing structure and surrogate model are modified by the global sensitivity analysis.Finally,the multi-objective optimization model of lightweight and noise reduction of the reducer is established,and using the improved multi-objective particle swarm optimization algorithm to solve the model.The main contents of this paper are as follows:1)The vibration velocity of the outer surface of the mine hoist reducer is used as the boundary condition to establish the acoustics analysis model of the reducer housing.The direct boundary element method is used to obtain the surface acoustic pressure and the external sound field radiated noise.2)The main characteristic structure thickness of the reducer housing is designed as the design parameter,and the sound power of the external sound field and the mass of reducer housing are used as the target response.The finite difference method is used to select the parameters which are sensitive to the target response,and the uniform experimental design scheme is established.The surrogate model of the noise and mass of the radiated noise of the reducer is trained by the radial basis function neural network,and the excellent prediction ability of the surrogate model is verified by comparison.Based on the Sobol’ index method and the Latin hypercube sampling technique,the global sensitivity of the design parameters of the reducer housing for the target response is analyzed,and the surrogate model is modified by the analysis.The corresponding structural improvement strategy is proposed for the insensitive design parameters.The simulation verifies the validity of the modification strategy.3)Based on the distance coefficient and elite strategy,this paper proposes a dynamic selection strategy and population diversity strategy to improve the search performance of particle swarm optimization algorithm.Combining niche technology and roulette method,this paper proposes an improved multi-objective particle swarm optimization algorithm,and the validity of the improved algorithm is verified by the test function.4)With the purpose of reducing the weight and noise power of reducer housing,a multi-objective optimization model is developed by employing the design variable which are chosen based on the global sensitivity of the gearbox housing,and the constraint condition with the stress and mass of the housing.The improved multi-objective particle swarm optimization algorithm is used to solve the optimization model,and the effectiveness of the improved strategy is verified by simulation.