Based on the grey - forecasting model with unfixed parameter , the primary model was ameliorated through optimized analysis on the non - equidistant sequence to improve its fining and forecasting accuracy especially when there are only few experimental data 摘要针对变参数灰色模型在实验数据少时误差较大的问题，提出了对非等间距序列优化的方法，使原来模型得到改进，从而提高了模型的拟合和预测精度，拓宽了应用范围。
And then , the paper analyses the reason engaged in the directional property of underwater acoustic transducer and array , discusses the physical parameter of the directional property . to point acoustic source equal spacing linear array , the dissertation makes experiment analysis and simulation 分析了水声换能器及基阵指向特性形成的原因、表征指向特性的物理参数，并针对点声源等间距线列阵的指向性进行了数学、实验分析与方针。
Based on original data of physical investigation for junior conducted in the cities of china and using the non - equidistance gm ( 1 , 1 ) model , the expect model for junior growth in the cities is established , the growth trend of junior in the future of 10 years is predicated 摘要依据国家先后4次测试公布的全国城市汉族儿童青少年体质调研的原始数据，运用灰色非等间距gm ( 1 , 1 )模型，建立城市儿童青少年生长发育预测值模型，预测未来10年城市儿童青少年的生长发育趋势。
( 4 ) through level division of the predict result which favorable for mineralize district in the studied areas . we put up a graduate method which is used to divide graduations of statistical result . this kind of dividing method is different from traditional interval or non - interval graduate method , it can avoid a subjectivity for level division and is favorable to divide the high value of the result . ( 5 ) in the light of rank points and weight of the predicting factors , minerogenetic prediction map had been made . on the basis of the map , six favorable districts to mineralize had been determined 这种划分方法不同于传统的等间距、不等间距划分方法，避免了划分的主观性，更有利于对预测结果高值区域的划分； ( 5 )根据预测因子的级别分数和权重，通过gis叠加分析操作，生成一系列的成矿预测系列图件，根据预测图件，在研究区圈定6个成矿有利区。
The dataset collected , called vch - m1 ( virtual chinese human male 1 ) , were characterized in the slice interval of the same 0 . 2mm , including 9232 anatomical serially sectioned slices ( up to 161 . 56 gb ) , and radiological data composed of 1720 ct images with scan thickness of 1mm and 780 mri images with that of 2mm . we succeeded in eliminating the deformation of the chest and the abdomin during embedding , using a technique of fixation under salty environment with low temperature , and also a technique of sequentially embedding with enough ice pieces 在对中国虚拟人女1号( vch - f1 ) 0 . 2mm等间距断面切削建模的基础上，我们对数字建模工艺的一些步骤进行了改进，获取了vch - m1的0 . 2mm等间距断面数据集，包括冰冻切片图像9232张，数据量为161 . 56gb ；影像数据包括全身断层ct1720张，层厚1mm ； mri780张，扫描间距2mm 。
The research achievements are as followed : studying the shortages and the improved methods of gm ( 1 , 1 ) grey prediction model , considering the characteristic of the transformer chromatographic data , bring forward the method for converting a series of data which are sampled in different interval into a series of data in the same interval . the weakening operator is applied to reconstruct the transformer chromatographic data for attenuating or eliminating the influence of randomicity . the improved prediction model for power transformer interior fault is constructeded 主要取得了以下研究成果：通过对gm ( 1 , 1 )灰色预测模型的缺陷及其改进方法的深入研究，针对变压器色谱数据序列的特有规律，提出了原始非等间距色谱数据序列的等间距处理方法，运用弱化算子改造原始序列，淡化或消除原始色谱数据序列由于受各种随机因素影响所具有的随机性，给出了适用范围更广的变压器内部故障改进灰色预测模型的建模方法。
A normalized mapping rule of raw grey series is introduced by analysis of exponential function characteristics of the whitening differential equation , which makes the non - equigap grey prediction model ngm ( 1 , 1 ) fit for universal raw grey series , improves prediction precision greatly , meanwhile makes innovation to series with negative values and enlarges grey prediction theory 通过对白化微分方程解的指数函数特性的分析，研究了原始灰序列的归一化映射规则，使非等间距灰预测模型ngm ( 1 ， 1 )适应一般灰序列，灰预测精度也大大提高。这亦解决了含负值灰序列预测的理论问题，拓宽了灰色预测理论。
To a degree , the influence of artificial factor and non - objectivity lies in the design process of membership . every fuzzy set of membership has several variables so the optimal question becomes very complex , especially , more the number of variables is , more obvious the question is 在常规隶属函数的设计过程中，通常采用对称相等、均匀分布(等间距)的函数曲线，这样人为因素的影响较大并缺乏客观性；另外在每个模糊子集的隶属函数中均有几个参数，因此优化的问题相当复杂，特别是当参数过多时，此问题由为突出。
At first , buck - boost circuit is detailed analyzed and derived strictly in the theory , and on the base , the current of the inductance and the output voltage are analyzed elaborately . when the input voltage is the ac voltage with the nomal frequency , the two variable quantity are analyzed and some conclusions are concluded by means of equal length separation methods and introducing function methods and derivative - equation methods . and this is the basis of the next proceeding 首先，对buck - boost电路进行了严谨地理论推导，先对buck - boost电路在输入为直流电压时的工作情况进行了分析；在此基础上，对输入为工频交流电压时的电感电流、输出电压作了详细分析，利用等间距离散化方法、引入函数法、微分方程法分别对变量进行了推导和总结，为下一步工作奠定了理论基础。
To diminish the error in the non - equidistant grey model for forecasting , considering the characteristic of the sequence , this model combined neural network and three models of grey theory with energy consumption data , and proposed the combination model of energy consumption 摘要针对非等间距灰色系统预测中存在误差较大的问题，结合序列本身的特点，利用世界能源消费的历史数据，采用3种灰色预测模型与神经网络进行组合优化，建立了灰色神经网络的能源消费量组合预测模型。
