At the same time , this paper discussed the inference machine of failure diagnosis expert system for pressure vessels . the expert system was made with forward reasoning , breadth first search and reliability theory 本文同时讨论了压力容器失效分析专家系统推理机制，选择正向推理和广度优先搜索的推理控制策略，运用可信度理论的不精确推理方法建立了压力容器失效分析专家系统。
After studying basal function of the medical computer - aided diagnoses system , with design methodology of expert system , the system ' s knowledge expressing form has been presented , together with the inference engine , which is relevant to the form 通过研究医疗辅助诊断系统的基本功能，借鉴专家系统设计的技术，建立了知识库和推理控制策略，并讨论了该系统中的知识表示形式及与之相配合的推理方法。
Abstract : a continuous type inferential control algorithm and the proof of the stability and the convergency of the inferential control system are given in this paper . the system has the performance of convergency and robustness . the algorithm which is simple and has only one adjustable parameter q and has been successfully used for the servo control of the tension of the film in the metallized film capacitor automatic winding machine and we get a satisfactory result 文摘:给出了一种适用于工程常见的二阶对象的连续型推理控制算法，以及推理控制系统的稳定性和收敛性证明.该系统无差、稳定，且具有较好的鲁棒性.该控制算法简单，仅有一可调参数q ，并成功地应用于薄膜电容器全自动卷绕机中作为快速张力伺服控制，取得满意效果
This can keep control safe and efficient , to attain the best work appearance . that method , ac cording to field knowledge , adopts the intelligence logic theories and the control ' s way to reason logically . the end target is that egcs , in accord with the actual movement transportation of elevator , the distribute condition of floors waiting for elevators , the direction passenger fly to , can mimicry the intelligence principle in the human brain , proceed the ego study , and will store that study the elevator ' s method and rule inside the knowledge library 动态最佳方法基于领域知识，采用了智能逻辑理论和模糊推理控制的方法，根据电梯运行交通流量、候梯分布状况、乘客流向等信息，模仿人类头脑中的‘智能”原则，进行自我学习，将学习到的电梯调度方法与规则储存到知识库中；在发生厅内召唤时，根据知识库中的知识，通过预测将厅内召唤分配给最适当的电梯，最大限度地提高梯群的整体运行效率。
Developed by oopl , it encapsulates several components of expert system , such as knowledge base and inference engine . it provides multiple knowledge representations including production rule and object oriented , and knowledge acquisition . multiple inference methods and inference strategies and explanation mechanism are properly organized under the model 该模型采用面向对象程序设计语言开发，内部封装了知识库、推理机等专家系统模块，可以提供产生式，面向对象等知识表示方法，以及交互式获得的知识获取方法，能综合运用的基于知识表示方式的推理、不精确推理等推理控制策略以及解释机制。
Second , discussing the detecting and eliminating of inference channel , different channel detecting methord are provided according to static , dynamic and compositional channel , and a new algorithm of eliminating the inference channel by promoting security was presented . third , discussing many problems encountered by transaction , which is from single - level to multilevel , and a transaction schedule algorithm is given based on data replicated architecture . according to the theory mentioned above , a mls / dbms is developed , and the experiment result has showed the correctness of the theory 本文围绕安全数据模型,多级安全环境下的事务处理,推理控制展开了探讨,首先针对空值的二义性和数据穿越问题,提出一种改进的多实例安全模型,并对外键引起的存储隐通道进行了控制,提高了查询的效率;其次探讨推理通道的检测和排除问题,针对静态、动态推理通道和复合通道分别给出了不同的通道检测方法,并提出一种通过提高输入安全密级来消除推理通道的算法;最后讨论了事务从单级拓展到多级所面临的各种难题,提出了一种基于数据复制结构的事务调度算法。
Abstract : a design method of a multi - variable inferential controller based on anew performance index is proposed . the inferential controller uses a v type control structure to avoid inverse transformation of rectangle arriving . this multi - variable inferential control system has the characteristic of simple calculation and only an adjustable parameter q array . this control system output gets the performance of dynamic decoupling of the mutivarable system 文摘:在一种新的性能指标下，提出了一种多变量推理控制器的设计方法，该控制器采用的v规范型控制器结构避免了传递函数矩阵求逆问题.该类多变量推理控制系统的特点是算法简单、仅具有一个可调参数矩阵q ，该系统实现了多变量系统的动态接耦
It is impossible to simulate the imaginary instinct deduction control completely with computer . " increment - decrement and half computing " intelligence control method is pointed out for the single input and single output control system , which is representative of the imaginary instinct deduction control well 若以现有的计算机技术完全模拟人的形象直觉推理控制是不可能的，针对单输入单输出这一类简单的控制系统，提出了“增减对分”智能控制法，较好地模拟了人的形象直觉推理控制。
By the browser / web server / database structure and the technology of asp . net and xml , we achieve the expert system running on the network , extend the application and ensure its backward compatibility ; we have the deep research to the representation of knowledge and apply it to the system , extend the traditional method of rule based knowledge representation ; introduce the fuzzy sets , analyze the generalized possibility interval sets , and combine the linguistic variables and fuzzy sets ; the system adopts the forward and backward chaining , researches the reasoning under uncertainty and inexact reasoning ; and we depart knowledge base and inference engine , make the upgrade simple and maintenance convenient . in the last , we construct the potato produce manage system according to the phases of potato ' s growth to give a detailed instruct to its produc e Net和xml技术，实现了网络化的专家系统，同时扩大了专家系统的应用范围，提高了系统的后向兼容性；对专家系统中的知识表示技术作了深入探讨，在传统的基于规则的知识表示方法的基础上，提出了基于规则扩展的知识表示方法，并较好地将其应用到本系统中；引入了模糊集合，分析并推导了广义可能区间集，将其应用于模糊知识的表示，将语言变量和模糊集合结合起来；提出对系统采用交互式知识获取途径、双向推理控制策略，对不精确推理和不确定性推理的传播算法作了一定的研究，成功地将其应用到马铃薯专家系统的构建之中；采用知识库和推理机相分离的方法，使得系统升级简单易行。
