Research On Condition Monitoring And Fault-Tolerant Technology For Multi-sensor Down-hole Tool

Multi-sensor Down-hole Tool (MDT) is the important equipment used in oil extraction, which works underground and provides parameters such as temperature, pressure, discharge flux, vibration and current leakage, according to which the operator could monitor the oil extraction equipment. But the precondition of all above is that the MDT should be very reliable. However, the problem is that the MDT works in very poor condition, some weak parts of the system, especially the sensors etc. are more easily to be broken-down compare with those working in condition with normal temperature and pressure, and it is hard to repair or replace them. So, it's very significant and worthful to make research of the condition monitoring and fault tolerance of the MDT.The MDT software based on Windows system is designed, its function and characteristics are summarized and its program design is recited in detail. In order to analyze the reliability of the whole system and find out the weak parts in theory, the method based on Fault Tree Analysis (FTA) is used and the conception of Triangular Fuzzy Number (TFN) and the median are inducted.To enhance the reliability, for the pressure sensors which are related to the temperature parameter, the data fusion processing technology to calculate the pressure is applied. The experiment validates the ability of the method on enhancing its temperature adaptability and precision. For the temperature sensors, a model is presented, which is based on prepared redundancy of double sensors, combining the time series predicting theory based on the RBFNN. The K-means algorithm is improved, the alternate time is reduced and the calculate speed is enhanced while reserving the inhere precision of the K-means algorithm. It actualizes the condition monitoring and fault tolerance for the MDT. For the fault recognition of the sensors, a new way is presented to identify the fault byanalyzing the change of the sensors' characteristic parameters in the paper.Besides, this paper elucidates some other aspects discovered among the test that will affect the reliability of the MDT, such as fault tolerance of the resistances and capacitance for modulation, wave sieve and fault tolerance of the software etc. and presents the author's consideration and suggestion.