| Metal cutting is one of the most important methods of workpiece machining in mechanical machining,and holds the balance of mechanical industry.How to improve machining efficiency and lower the cost of machining is one of the importment task explored and studyed for a long time by machinists.Building cutting tools' database and providing optimized data is one of effective methods to improve machining efficiency and lower the cutting cost.Heavy-duty metal cutting has bigger working area and takes longer time,so that cuting tool and working parameter affects machining efficiency and cutting cost.It's important to build a Heavy-duty metal cutting expert system in order to choose better cutter tool and working parameter.In this paper,traditional expert system's and artificial neural net(ANN)'s basis characteristics were analysed,on which artificial neural net's knowledge acquisition was introduce to expert system ,and a expert system basis on ANN was realized.This system has the ability of obtaining study samples through experimental data,and the ability of self-study,which fetchs up traditional expert system's disadvantages. ANN changes traditional expert system's calculation from symbols to numerical values, and predigests program codes,and improves system's running speed.ANN's BP arithmetic and its betterment arithmetic were metioned in this paper.This paper alse explains this system's network structure.A series of methodes were taken in programming to improve placidity and to quicken convergence rate.A new method to solve local minimization problem was described too.This method can lower system overhead.System knowledge was classified in this paper,The knowledge-base is made up of four parts:database of cutting datas,database of resources,database of knowledg,database of users.Each part was explain through examples.Verified through practice,this system can save much time in tool choice,and cut down on labour,at the same time lower the processing cost.It not only overcomes anthropic factor's effect during tool choice,but also controlls manufacturing quality efficaciously. |