Data Mining Technologies Of ECC System And Its Application In Pumping Hydraulic System Health Status Evaluation

Along with the formation of the global market and the global environmental consciousness enhancement in recent years, in china, engineering machinery manufacturing industry is facing pressure from both domestic and foreign resources saving and environment friendly requirements, so it’s urgent to develop remanufacturing technology. This paper takes the concrete pump truck ECC system as the research object, recordeding the amount of information of operation and maintenance about20G/day, which obtained from monitoring is massive data in days and months multiplying. The massive data recorded by concrete pump truck ECC system is the basis of product design for remanufacturing, remanufactured parts life prediction and remanufactured product quality assurance. This paper through data mining technology to in-depth analysis these massive data that in order to dig out the potential useful information for concrete pump truck remanufacturing.Firstly, this paper introduces the concrete pump truck ECC system and the related technology of data mining, then identifies the target data mining and data mining methods and processes. Duo to the massive data, which recorded by concrete pump truck ECC system, is comparison messy, according to concept hierarchy of data classification techniques in data mining presented the concept hierarchy tree model of concrete pump truck ECC system, then through C++program to implement management and query of data classification in Microsoft Visual Studio2010, in preparation for the development of information management software system of the various parts of the concrete pump truck.Secondly, in order to determine the history common condition of pump truck boom for the data mining target, using the results of data classification to choose the data set of history condition pump truck boom in boom system. Since the data, which recorded by concrete pump truck ECC system, often exist many errors, such as deficiency of attribute values, missing data and abnormal data, etc; To avoid lead to the failure of the data mining tasks, which need to data preprocessing for missing data set, outliers and extreme value before the data mining tasks. The high quality data set, which after the data preprocessing, by using K-Means clustering analysis to obtain the history common condition of pump truck boom and proportion coefficient of each history common condition accounting for historical conditions, which provides the basis for key parts of pump truck (such as boom and turntable, etc.) residual fatigue life prediction.Finally, this paper introduces the working principle of concrete pump truck pumping hydraulic system, and analysis the failure mechanism of the main cylinder and the main oil pump. Considering the failure forms of the pumping hydraulic system’s key components that recorded in the ECC system and pressure, flow, temperature of pumping hydraulic system that characterization its health status, which taked as indicators for evaluating the health status of corresponding components, then pumping hydraulic system fault hierarchy model is put forward. Using the established pumping hydraulic system fault hierarchical model and expert opinion to build judgment matrix, and to check the consistency of the judgment matrix, and then calculated the weights of evaluation indexes relative to the upper level index; Finally, through fuzzy comprehensive evaluation technology to apply fuzzy evaluation for all levels indicators, then take the assess results as the corresponding key components of pumping hydraulic systems on the basis of remanufacturing, which guide the key parts of concrete pump truck pump hydraulic to remanufacture.