From the substantial evidence , four variable factor of the debt bond market risk have been concluded by use of spssl0 . 0 software . ( 2 ) the basic principle of modern risk management is applied . to aim at the present shortcomings of chinese debt bond market risk that the government debt bond price to deviate value , under the interest rate term , the developed vasicek mode is used by use of eviews software ( 2 )应用现代风险管理的基本原理，以防范化解利率期限下的我国国债市场风险，针对我国传统国债定价弊端，引入扩展型多因子瓦西塞克模型，利用eviews软件nls估计，推导出了在利率期限下，中国国债市场的扩展型四因子瓦西塞克国债定价模型。
Following , making development study from the three directions : the first one is how to reduce calculation when to use markowitz model . this text has improved the efficient frontier of markowitz model utilizing free risk assets , and reduced calculation about revenue rates " co - variance matrix utilizing single or multiple factors , and so on . the second one is to add thinking factors about , such as transaction fee , fund limitation , lowest transaction unit ' s limitation , risk measures and exchange rate risk of international portfolio securities , so as to make markowitz model closer to our country ' s practice 接着，分三今方向对markowitz模型进行了拓展研究：第一个方向是运用markowitz模型时如何减少计算量，本文利用无风险资产来改进markowitz模型的有效边界，利用单因子或多因子模型来减少收益率协方差的计算量等等；第二个方向是增加考虑因素，诸如交易费用、资金限制、最小交易单位限制，风险测度和国际组合证券的汇率风险，使markowitz模型更贴近我国的实际；第三个方向是对markowitz模型进行动态拓展研究，提出了将证券收益率看成是随机序列时的投资决策模型，深入研究了m ? v有效边界随资产品种数增加而发生的漂移，并用解析方法和几何图形描述了漂移的轨迹和方向。
At first , this paper analyzes the factors of water - sand influencing water level of yellow river and the feasibility just using the factors of water - sand to study water level , and collects the corresponding data ; secondly , because there are strong nonlinear relation in the corresponding data , by meticulous theory analysis , this paper integrates basic nonlinear analysis method , theory of random analysis , method of least squares and so on . it puts forward a method which can get the high accuracy simulation of the data , perfects the multi - factor analysis of variable ( over three factors ) of the statistic ; thirdly , it applies the method to the approximation of corresponding water level process which belong to the capacity of sand of middle - high and middle - low , and get the high - accuracy simulation about the typical nonlinear relation ; at last , this paper definitudes the main influence mode that the capacity of sand . it mainly unite with other factors to work on the water level in the yellow river lower reaches ; mor eover , this paper analyzes the difficult point and the direction of improvement to realize the accuracy forecasting of the flood level of erodible - bed channel 首先，系统分析了影响黄河水位的水沙因素，及仅用水沙因素有效研究水位的可行性，并按变量对应思想采集它们的相应数据；其次，由于相应水位过程数据中含极强的非线性关系，本论文经细致的理论分析，将基本的非线性分析方法、统计建模方法、随机分析理论、最小均方误差原则等等数学理论及方法有机揉合，提出了能有效实现这类数据高精度拟合的分层筛选法，并改进了统计学中多因子(三个以上)方差分析法；再次，将这一方法用于黄河中高及中低含沙类洪水相应水位过程的拟合，实现了这一典型非线性关系的高精度拟合，各年汛期上下游相应洪水位过程的拟合误差都较小；最后，明确黄河下游含沙量对水位的主要影响方式，即含沙量主要是与其它因素联合对水位作用；另外分析了要实现变动河床洪水位过程准确预报的困难所在及改进方向。
