Study And Application On The Retrieval Model Of Monitoring Data In Complicated Engineering System

There are many kinds of supervising system coming from different manufacturers in complex engineering system. These supervising systems are accessible to devices. Each of the device includes high frequency gather of many parameters to feed back timely the health in complex engineering system. This data is structured one. In the process of observation the whole complex engineering system, the supervising system involved in the procedure needs integrating to make sharing the resources and controlling the whole come true, leading to the large amounts of figures, high latitude and enormous categories of data, during which, the first is the most noticeable problem. Because not only does it include the sum, but also it retrieves quantitive data demanded within the justifiable time to analyze the health in the procedure of using.What’s more, considerable figures appear in the form of Chinese document by the accumulation of experience in the complex engineering system. This unstructed data is used for explaining the system characteristics, problems from the figure’s devices feature as well as the solutions.In the process of using the monitoring data in complex engineering system, firstly, a great deal of target data should be retrieved accurately and quickly from the data base or data document to satisfy the the dealing and analysis of data about upper application. As a result, the policy makes can have a whole control of the engineering procedure. It’s greatly practical to complete the retrieval of data real-timely and correctly.It’s found in the research on the retrieval efficiency of the structured data that the traditional data base cannot satisfy the requirement for timeliness of dealing with data because of the time limitation of the disk automatically revolving to fix position and I/O. Aiming at the organization and management of structural data, and considering the current situation where the capacity of the main memory is increasing constantly and the price is decreasing continuously, this thesis will study the essential technology such as the storage structure of the data in main memory database and retrieval technique. In addition, it proposes an optimized T-tree indexing algorithm alignment from the perspective of align the cache line and range queries. A memory copy of data established on the base of that use the high efficiency of access to the memory to improve the efficiency of data retrieval.For the Chinese document retrieval of full-text, the main reasons of affecting the quality of the retrieval are the accuracy and efficiency of segmentation.In order to fulfill the utility requirement,this paper presents a segmentation algorithm based on Chinese dictionary and statistics.The algorithm process consists of pretreatment processes, preliminary segmentation, calculation analysis and amendments of rule. In the module of preliminary segmentation, Using double-words hash dictionary to enhance the efficiency of the preliminary segmentation according to the situation of two words or more than that. Meanwhile, as for the detection of ambiguity field, we propose an optimized maximum two-direction matching detection algorithm. Then using a way based on statistics to eliminate ambiguity.Finally, the segmentation results can be further amended throw artificial regulation base.At last, arranging the retrieval module which is based on the requirement for the retrieval model of monitoring data in a space launch center supervising system to further validate the usefulness of the retrieval model.