Transform About Capability Testing Standard Of Friction Materials By Computer

In recent years, friction materials product of my country enter into international market more and more rapidly and the quantity increase too. The product of one company should product according to the GB5763—98, GB/T5764—98 of our country,meet the require of domestic marketAt the same time ,they should meet international quality require, mensurate friction capability at the chase machine according to the SAEJ661 and classify them to different rank.If capability data under different testing machine can transform each other by a program,that can bring huge convenience to the company of friction materials for their produce and quality control.My thesis aim at the above problem,choose the discal brake patch used by car which occupy the most market share as the research object.Bring forward three different methods,namely:curve fitting toolbox of matlab,write 'm text' by tool function of matlab to build nerve network model and use nerve network toolbox to build model directly.the three methods all can finish the standard transform task,error are also in the allowable range.But three methods have their own characteristic,suit for different people and occasion. Curve fitting toolbox suits for lower error require and little data quantity.Because this method operate simply,the people who have bad program writing ability and don't know other transform methods can use it.Writting 'm text' suit for higher error require and enough data quantity.compare with the first method,this one is more scientific.If you are familiar with matlab language ,use it.The last method suit for the people who know something about nerve network knowledge,know how to build nerve network model,but not familiar with matlab language.