In the establishment of knowledge base , in order to find a good way to represent the knowledge which is fit for our system , we introduced a way to represent our knowledge by the rule branch and the rule number , combining the theory of expert system and on the basis of collecting and summarizing , analyzing the deficient elements of the crops . considering the using for farmers , the characteristic of this system and it ' s reliability and accuracy , we adopted the uncertainty reasoning model which included certainty and weight factors . we successfully established a crops nutrients diagnostic expert system under specific conditions 在知识库建立中，为了能找到一套适合于本系统的知识表示模式，在收集并总结和分析了作物缺素症状的基础上，结合专家系统的理论提出了采用规则族与规则号的知识表示方法对该系统的知识进行表示；在推理过程中，为了更好地提高系统的实用性、可靠性和准确性，从农民用户使用角度着想并结合系统本身的特点采用了带有可信度因子加权的不确定性推理方法。
In this thesis we firstly made all kinds of statistics on chinese character stroke information of chinese national standard code for information interchange ( gb2312 - 80 ) , such as the average strokes of each chinese character and each character which uses utility frequency as its weight , the average strokes of each character been added weight or not that can be differed from the other characters , the number of chinese characters which begin with each stroke , the times of each stroke in the chinese character set , the chinese characters which have same strokes in the chinese character set , and the frequency of adjacent stroke , etc . according to the statistic , we have devised a new chinese character keyboard input method named " stroke code " by adopting the stroke sequence of hand - written chinese character as its input codes and designed a kind of key arrangement for the method 论文首先统计了国标二级字库中汉字笔画信息的各种数据，这些数据主要包括：汉字的平均笔画数及按使用频度加权的平均笔画数、能与其它字区分开的汉字前若干笔画的平均数与加权平均数、以各种笔画起笔的汉字数、各种笔画在汉字字库中的出现次数、汉字字库中笔画相同的汉字以及汉字字库中相邻笔画的频度等。根据这些统计数据，我们采用书写汉字时的笔画顺序作为汉字输入码，设计了笔画码汉字输入法和实现该输入方法的键盘。
This method utilizes an on - line algorithm based upon lssvm ( least square support vector machine ) , which can build adaptive models to predict the cod values of unknown water samples quickly and accurately . in the modeling process , every training sample is also assigned a prior weight to take their significance to the final predictive model into account 该方法是一种基于最小二乘支持向量机的在线自适应加权算法，这种算法可以自适应地选取和未知水样最相近的标准样本进行建模，同时在建模中又利用加权的方法分别考虑了各个标准样本重要性。
The innovation of this paper is concentrate the merit of the above two recommendation technology , namely , it is a hybrid method that use the conception of customers ’ preference and weighed clv , and what is joited is to use the conception of threshold , so it can be settled on the joit of the two method and avoid the error of use the two recommendation seprately . in this hybrid method , we can firstly identify the loyalty of researched customer use clv , what is mean is firsly the wrfm technology is used to classify and secondly identify the customer is high loyalty or not , if he belongs to the 本文的创新点正是集中以上两种推荐技术的优点,使用了基于客户偏好的协同过滤和基于加权的客户生命周期值两种理论形成的混合推荐方法。这种方法在两种技术的结合点上采用了“阀值”的概念,这样就很好地解决了两种方法的中间过渡并且有效避免了两种推荐方法单独使用时的偏差。
In chapter 3 , the group decision - making is applied in the safety evaluation , and the credit degree is put forward to considering the experts " appraisal level . the principle of set - valued statistics and the fuzzy gravity center based decision theory are put into a new method , and are applied in the uncertain type of ahp . hence , the weights of evaluation indices that reflect the experts " level are gained . this measure , which reflects the objective of the evaluation process and the subjective of the experts , can concentrate various opinions and lower the deviation of the evaluation process . a new method of calculating weights and variable weight is presented in the safety evaluation , and it is discussed that how the method of variable weight is applied in several evaluation methods 在原有的集值统计原理的基础上，引入对专家评判水平的评判? ?专家的权重，形成加权的集值统计原理，并与重心决策理论相结合，计算出考虑了专家水平差异的桥梁安全性评价指标的权重。这样处理反映了评价的客观性和参评人员的主观性，能很方便地集中多种不同意见，减少了评价过程中的随机误差。在桥梁安全性评价中，为了反映桥梁状态发展的不均衡性，使评价过程更趋于合理，评价结果更可靠，将变权原理应用于桥梁安全性评价，探讨变权原理在多种评价方法中的运用。
A novel coding algorithm was presented for still ima ge using biorthogonal wavelet transform . combined with the visibility preaccentua tion before the quantization ， a new data structure named extended q _ tree was pro posed to organize wavelet transformation coefficients . experimental results show that the new algorithm is very practical and with good performance 研究了一种基于双正交小波变换的静止图象编码算法.结合带视觉加权的标量量化技术，提出了一种新的数据结构? ?扩展四叉树，从而实现了一种性能较好、复杂度较低的静止图象编码方案
In order to allocate different bandwidth to different flows in a network supporting multiple priorities according to a predetermined proportion , wfriday ( weighted friday ) , which is an extension version of friday , is presented in chapter v . to control different bandwidth for different priority , wfriday employs the concept of virtual arriving rate 为了在多忧先级的环境中，根据预定的带宽分配比例，用缓存管理技术为不同优先级的流分配不同的带宽，本文在第五章提出wfriday （加权的friday ）算法。该算法是对friday算法的拓展。为支持多优先级的带宽分配， wfriday引入了虚拟到达速率的概念。
Moreover , this thesis proposes two novel strategies : the weighted fuzzy control strategy for multi objects and the behavior control strategy based on emotion evaluation and associative learning , and these two strategies are applied on an invert - pendulum system and a simulative agent system , respectively 最后，本文分别提出了基于对多性能指标分别加权的专家模糊控制策略，以及基于人工情感评估机制与联想学习的行为控制策略，并分别将两种策略应用于非精确倒立摆系统和仿真agent系统。
In part two of this dissertation , a general interaction programming of interactive system which has m subsystems is introduced according to the productions and resources distribution model ( prdm ) . also , we define three kinds of solutions of it . after transforming the ipp into a single - level and single - object optimum problem , the relation between their solutions > s discussed 第三章针对资源?产品分配模型( prdm )给出了数学模型? ?一个有多个子系统的一般交又规划模型，提出了较为合理的几种解的定义，并利用加权的方法将交叉规划转化为一单层单目标问题，然后讨论了他们的解之间的关系
Based on the framework of particle filter , this paper proposed and implemented different tracking algorithms for single mouse target and multi mice targets separately , which are as follows : we combined the weighted color histogram , which is considered as observation model , into the framework of particle filter and realized the tracking of single mouse target 本文在粒子滤波框架基础上,分别就单个鼠类目标和多个鼠类目标的情况,提出和实现了不同的跟踪算法。具体内容如下:提出将加权的颜色直方图作为观察模型融入粒子滤波器算法框架,实现了对单个鼠类目标的跟踪。
