The emphasis of the research in thesis lies in identifying the categories of the structures for manufacture and correcting dimensions . and the method adopted is called " character recognition " . using this method , the computer first recognizes the structures in the shaft - like part drawings according to their characteristics , such as slots , through holes and so on , and then corrects the drawings based on the recognition 本文的研究重点为轴类零件图中各种工艺结构的识别和尺寸标注的批改，采用的方法为“特征识别法” ，该方法针对轴类零件图中各种工艺结构的固有特征，从一幅零件图中识别出零件的各种工艺结构，包括槽、通孔等，然后在此基础上进行批改。
Finally , a detailed and intelligent solution is proposed for diagnosing the health condition of the structure , which based on bp neural net - the method identified step by step , combine with a numerical example . the frame structure is a two bays and three stories . through establishing ann model , simulating verification has been processed , which is to predict the damage location and severity . it has been shown that a proper ann model has a comparatively strong ability on damage identification . however , numerical results also show that further research should be done to improve the precision of the damage identification 最后，提出了具体的智能化算法，基于bp网络的结构健康状态的诊断方法? ?分层逐步识别法。结合数值算例，对一个两跨三层框架结构，建立bp网络模型，对结构的损伤位置和损伤程度进行了模拟试验及识别。模拟试验表明恰当的神经网络模型具有较好损伤识别能力，然而对于损伤程度识别的精度问题，还需要进一步深入的研究。
The three kinds of simulated point targets are designed and the rangeprofiles at aspect angle are computed . the radar target recognition method based on the optimal cluster centers is simulated and studied . it is discovered that the algorithm is effective when there are lots of training data , but noneffective when there are only a few training data 2 、对最优聚类中心目标识别法进行仿真实验并研究其识别性能,实验结果表明该算法在大样本训练数据时能得到较高识别率,是一种有效、可行的识别算法,但在少样本训练数据时,所得识别率急剧下降。
According to the data of drilling and other geological investigation on the lpg storage of jinzhou , and taking into account main factors that affecting stability of the storage and other classification standard , this paper selects five evaluating factors . using pattern recognition method and muti - step fuzzy classification clustering method to classify 根据锦州lpg洞库的地质勘察资料，考虑到影响洞库围岩稳定性的主要因素，结合已有的分类标准，本文选取了5个评价因子，用模式识别法与多级模糊聚类法来进行场地岩体的稳定性分类。
This recognition method integrates traits of nearest neighbor method and k - neighbor method . its basic principle is to improve recognition reliability with tow additive thresholds . in addition , ternary threshold neighbor recognition method ’ s fast arithmetic based on c - average clustering and huffman tree is put forward 2 .本文在经典统计模式识别方法的基础上,结合了最近邻识别法和k近邻识别法的特点,给出一种三阈值近邻识别方法,其基本原则是在寻找未知样本的近邻时,通过附加阈值的限制,进一步提高识别的可靠性。
In the second , in allusion to non - stationary the characteristic of the signal , the author introduces to the method that using empirical mode decomposition to analysis the vibration signal so that the signal are made up of some intrinsic mode function , after this process , we can use stochastic subspace identification to identification the mode parameter of the structure and find the same work frequency 其次,针对气阀振动信号的非平稳特点,本文采用了经验模式分解法( empiricalmodedecomposition )对振动信号进行分析处理,使之成为若干个基本模式函数imf ( intrinsicmodefunction )和一个残余量的线性组合。接着采用随机子空间参数识别法对各个基本模式函数其进行结构参数识别,同时找出各种状态的共同工作频带。
