Research On The Real-time Monitoring And Early Warning Method For The Process Of The Engine Bench Test

The engine in bench test bears great practical alternating mechanical load and thermal load. Its speed will have several mutations in a short time and a lot of test loop in a test process. With the change of speed, the load will change. The engine with design defects or poor performance in the test process will be premature failure. If the failure of the components has not been timely found, it will lead to the failure of other parts. Therefore, it is necessary to study a set of monitoring and warning system, which will monitor the engine and test process. Before a major failure will happen the system will set an early warning and stop the test to protect the testing personnel and test bench. For this purpose, the research and innovation of this paper are as follows:1) Research of monitoring method for engine test process. By analyzing the failure type and cause of engine main kinematic pair and gearbox, the vibration signal is proposed to monitor the engine test process and the torque of engine performance parameters will be an an aid for monitoring parameters. Propose to use the vibration signal characteristic value RMS to realize the engine monitoring during the test time history and the composite power spectrum of corresponding order bands to monitor the pitting wear, tooth fracture and so on of engine and gearbox.2) Research of vibration limits established method using intelligent learning. The paper proposes the engine speed and torque are related with RMS which reflects the energy trend of engine. RMS is related to the torque and square of engine speed. The paper establishes a three-dimensional model which formed with the engine speed, torque and limits of RMS. In the paper, author uses the variable interval variable speed and torque conditions to establish the actual classification model. Use the statistical theory to learn RMS of different engine condition and the method of 3σto determine the upper and lower limits.3) Research of early warning of in the engine test procedure. The author constructs an early warning model based on D-S information fusion in the engine test procedure and uses the relatively badness to measure the state of monitored signals. In the paper, by the example the author verified the early warning model can reduce the uncertainty of the information and improve the classification recognition capacity to engine work state of system. 4) Design of real-time monitoring and early warning system. The chip TMS320C32 to process the digital signal afete AD conversion, and the signal will transmitted to computer through the USB module. The virtual instrument technology is used to design the subroutine and signal analysis module.All the above work are verified by the examples. The monitoring method is reasonable and early warning model is accurate. the system can refer a safe working environment to the engine testing room and avoid major accidents.