摘要
矿产品价格是矿业项目投资经济评价的重要参数。矿产品价格的合理确定是一个十分复杂的问题,也是涉及矿业项目投资经济评价可靠性、可行性的关键。本文首先讨论了矿产品价格的定价原理,指出其具有较强的不确定性和时序性,在此基础上,建立了基于径向基神经网络(RBF)的矿产品价格非线性预测模型,由3层前向神经网络组成,并以高斯函数作为基函数,该模型具有结构自适应、易于收敛和外推能力强等优点。应用建立的预测模型时某金属的中长期价格进行仿真,结果表明具有较好的可靠性和实用性。
The price of mineral products is a important parameter in the economic evaluation of mining project＇ s investment.It is very troublesome to determine adequately the price of mineral products,and it is the kev to evaluate the reliability and feasibility of mining project＇s investment economically.At first,the principle to fix mineral products price is discussed and draw the conclusion that it has obvious uncertainty and time-series.Then,the nonlinear model to predict mineral products price is set up based on RBF neural network,which is composed of 3 forward neural lavers and use Gauss function as its basic function.The model has the merits of self-adaption in structure.ease in convergence and extrapolating.The appling of the model to simulate a metal medium and long term price indicates that it has good reliability and practicability.
出处
《中国矿山工程》
2014年第1期65-67,共3页
China Mine Engineering
关键词
矿产品
价格预测
非线性
神经网络
mineral products
price prediction
nonlinear
neural network
