Research On Price Forecast And Purchase Models Of Nickel And Their Application

Research on raw material price forecast and purchase model has important academic value and practical value,which has a widely practical background in fields such as oil price forecast,iron ore purchase,and so on.Nickel is not only an important raw material in the manufacture of stainless steel,but also an indicator of the stainless steel raw material market and product market.Moreover,it is a raw material with commodity and financial attributes.Influenced by the global macro economic environment,global supply and demand of nickel,production cost of nickel,US dollar index,nickel substitutes,emergencies and even black swan events,as well as many other factors,nickel price fluctuates frequently and violently.A large number of facts or experimental results show that scientific forecast of the nickel price trend helps stainless steel production companies to make the most reasonable purchase at the right time and the right price.Therefore,the price forecast and purchase models of nickel are an important research topic of concern.Currently,studies on the price forecast and purchase models of raw materials have drawn the attention of the academia and the business community,and achieved certain results.However,such studies seldom combined the raw material price forecast model with the purchase model effectively,and have not formed a complete research system.Therefore,it is of great theoretical and practical significance to systematically study the price forecast and purchase models of nickel and build a targeted price forecast model and purchase models of nickel.In this paper,the price forecast and purchase models of nickel were deeply discussed and studied.The research was finished as follows.First,the problems of nickel price forecast were described,on this basis,and a nickel price forecast model based on the improved support vector machine(SVM)was proposed.The specific research work included the following steps:1)analyzing the basic idea,main characteristics,algorithm and kernel function of SVM,and improving the existing traditional SVM methods mainly through optimization of the number of training samples;2)analyzing and describing the problems in nickel price forecast;3)on the basis of nickel price analysis,discussing the feasibility of price forecast by using the SVM principle;4)building a nickel price forecast model based on the improved SVM through fuzzy granulation of historical nickel price data and generation of SVM training samples.Second,a nickel purchase model was proposed.Based on the characteristics of nickel price fluctuation,and with reference to related studies on purchase models,this paper constructed a nickel purchase model.The specific research work included the following steps:1)analyzing the purchase theory,basic purchase model and case-based reasoning technologies,which laid a theoretical foundation for further research of constructing a nickel purchase model;2)proposing methods of determining short-term and long-term nickel purchase ratio based on similar case analysis;3)building a nickel rolling purchase model under price uncertainty.Third,practical application research of the nickel price forecast model support system at company B was put forward.The specific research work included the following steps:1)analyzing the background of company B which produced stainless steel,including its basic facts,the importance of nickel price forecast to the stable operation of the company,and the necessity of building a nickel price forecast model support system at the company;2)designing and developing a nickel price forecast model support system;3)using the developed support system for nickel price forecast.Fourth,practical application research of the nickel purchase model at company B was put forward.The specific research work included the following steps:1)analyzing the background of company B which produced stainless steel,including the existing problems of company B in nickel purchase,and the necessity of using a nickel purchase model to develop a suitable nickel purchase strategy;2)collecting the historical case data and extracting similar cases,and then determining short-term and long-term nickel purchase ratio;3)developing a nickel rolling purchase strategy for company B based on optimization model.In summary,this paper proposed an S VM-based nickel price forecast method through the optimization of the number of training samples.This method can effectively improve the accuracy of nickel price forecast and provide a scientific and reliable nickel price basis for buyers.Moreover,this paper also provided a method of determining short-term and long-term nickel purchase ratio based on similar case analysis,and a nickel rolling purchase strategy under price uncertainty,thus further improving and developing the theories and methods of raw material purchase.Therefore,such in-depth study on raw material price forecast and purchase models lays a solid foundation for further research and application,and can be used as a reference and guidance at both the theoretical knowledge and technical approach levels.