| As key components in aerospace inertial navigation products,the platform-like structural components are ideal for aerospace field research due to several characteristics including compact,complexity,thin wall and low stiffness,which have provided high-precision mounting references to inertial instruments and reference prism components,such as gyroscopes and accelerometers.Compared with other aerospace materials,high-strength aluminum alloy materials(such as 7000 series,etc.)and new aluminum-lithium alloy materials(Such as 2050 and 2060 series,etc.)are the most commonly adopted in manufacturing the typical structural components of aerospace inertial navigation currently.Aluminum alloys have the characteristics of high strength,low density,good molding and processing performance,high cost performance and good machinability,but for the typical structural components of aerospace inertial navigation products,aluminum alloys also have the shortcomings of complex shape structure,large material removal amount and thin wall deformation.And it has high requirements on various aspects of technical indicators such as machining accuracy,quality and processing efficiency.Therefore,how to achieve efficient processing of aluminum alloys in the aerospace field is a goal that many aerospace manufacturers have been researching and pursuing.In this paper,the aluminum alloy in the aerospace field was used as the main processing material,and the aerospace inertial platform structural components were the main research objects.The research focuses on the effective construction of the milling force prediction model of space inertial navigation platform structure through the finite element simulation software Advant Edge and RBF neural network,and the improvement of the milling efficiency by reducing the amount of calculation while satisfying the accuracy requirements.Having taken the aerospace aluminum alloy milling force as the research target,the target size constraint range was set,and the optimized cutting parameters were obtained through the BP neural network,which provided support for the cutting parameters of the enterprise and the operator.The main research contents as follows:1.Carrying on theoretical analysis of typical structural components about aerospace inertial navigation products,including typical structural products,metal cutting,finite element simulation software Third Wave Systems Advant Edge,RBF neural network and BP neural network,by which the theoretical foundation was been provided for the milling simulation and milling force prediction system of aluminum alloy construction.2.Depending on the finite element analysis software Advant Edge,Orthogonal test method was been used in the simulation experiment of aluminum alloy milling,and the simulation data was obtained,which provided data support for the selection of milling parameters and milling force.3.The prediction model of milling force of aluminum alloy had been constructed according to Simulation data and actual experimental data were used as training samples and test samples of RBF neural network respectively,by which those forecast data had been availed.The experimental results had been provided that the establishing predictive model had been used to save resources and Increase efficiency under the conditions of accuracy requirements of actual milling experiments according to compare the forecast data with actual experimental data.4.With the aim of milling force,according to the actual needs of the project group,the constraint range of milling force was set.Based on the BP neural network,the optimal milling parameters satisfying the constraint range were obtained by the error back propagation method,and substituted into the constructed RBF neural network milling force prediction model for verification.The results showed that the optimized milling parameters obtained by the BP neural network meet the milling accuracy requirements. |
PDF Full Text Request