The Design Of The Passenger Train Axle Temperature Alarm System Based On QNX

In China, the Axle Temperature Alarming Device,(ATAD)as an important vehicle monitoring instruments, request each railway passenger vehicles must be installed, and ensure all boot and available. Because of the temperature alarm algorithm is simple, so for a long time, much false-negatives brought to work. Since sixth times railroad large-scale acceleration, the express train began to install the Train Coach Running Diagnosis System (TCDS). Although TCDS enhances ATAD and ground center link, but hotbox misstatement still happen for production.The main contents of the thesis includes as following:Research TCDS axle temperature monitoring subsystem of system structure and technical features, and analyzes TCDS axle temperature monitoring principle. To understand TCDS responsible store-and-forward ATAD's data, the core is still ATAD. Summing up and analyzing all sorts of reasons false-negatives, conclusion is hotbox discriminant model is too simple.The rule of hotbox recognition and understanding can promote hotbox discriminant more accurate. This thesis summarizes the research methods of hotbox rule, analyzes the shortcomings of these methods, and puts forward a new view using grey forecasting theory predict hotbox. In order to improve the prediction precision of grey model, this thesis using a recursive median value filter and the metabolism of GM model. Based on the YW25T675498 Z11 of 2008 with historical data simulation experiment, and verifies the grey model GM (1,1) the applicability and effectiveness of hotbox.Through the temperature of the theoretical analysis and experiments, the results show that:Using the GM (1,1) model can predict the temperature, improve the real-time, to avoid the accident. Using a recursive median value of filtering method can effectively filter, improve clutter hotbox forecast accuracy. For the average temperature environment, reduce the false-negatives winter is very effective.