Mechanism And Data-driven Surface Quality Prediction And Control For Cylinder Head Milling

As the core component of the diesel engine,the cylinder head is made of vermicular graphite cast iron Ru T400 material,which is difficult to process and requires high precision.The cylinder head is a porous,thin-walled box type component,and its processing quality determines the sealing performance and service life of the engine.Milling is the main processing method for cylinder head production,and the surface milling process accounts for more than60% of the man-hours.Therefore,exploring the prediction and control method of milling surface quality of cylinder head has important theoretical and engineering significance to improve the assembly accuracy and sealing performance of cylinder head and ensure the long-term stable operation of engine.Based on this,under the support of the special project of high-end CNC machine tools and basic manufacturing equipment,this project takes the milling process of the bottom surface of the cylinder head of a commercial engine under the CHINA VI standard as the research object.This project deeply explores the milling mechanism,systematic analysis mechanism and datadriven milling surface quality prediction and control method.The main research contents are as follows:(1)In order to solve the problems of difficult acquisition of milling force and milling heat data and high cost in the actual milling process of cylinder head,a model of engine cylinder head milling mechanism is established.Combined with Advant Edge FEM simulation analysis and orthogonal experiments,the influence of process parameters on surface roughness is deeply explored,and the mechanism of milling force and milling heat generation is revealed,which lays a foundation for the prediction and control of subsequent cylinder head milling surface roughness.(2)To solve the low accuracy and long time-consuming problem of the surface roughness prediction model of cylinder head milling,a mechanism and data-driven surface roughness prediction method of cylinder head milling is proposed.The milling speed,the feed per tooth,the amount of back cutting in the process parameters,and the milling force and milling heat data in the output state variables of the milling mechanism model are used as the input of the prediction model.And the improved differential evolution algorithm is used to optimize the internal parameters of the SVR to solve the problem of poor model state representation ability and slow convergence speed.In this way,a fast and precise prediction of the surface roughness of cylinder head milling is achieved.(3)Aiming at the difficulty of surface quality control caused by the time-varying and highly nonlinear of the cylinder head milling process,a surface roughness control model based on a hybrid learning algorithm to optimize the Adaptive Neuro-Fuzzy Inference System is constructed.The hybrid learning algorithm combined with BP algorithm and LSM is used to optimize and adjust the parameters of the front and rear parts of the control model,so as to solve the problem that the fuzzy rules of the control model are set unreasonably and the parameter optimization is easy to fall into the local optimum.In this way,real-time quantitative control of the surface quality of cylinder head milling is realized.