Bidirectional Vibration Based On Blade Tip Timing Research On Monitoring System

Rotating blades are widely used in aviation,shipbuilding,electric power and chemical industries.Its stable operation is the key to ensure the safe and reliable transportation of large rotating machinery and to achieve high efficiency production.It is also the core element to achieve energy conversion,and its secure has a direct impact on the efficacy and sustainability of the whole system functions.Therefore,it is necessary to monitor the blade state.However,before the overall monitoring,it is also necessary to test the designed monitor system to assure the consistency and reliability of the whole detecting systems,so as to make timely adjustments.In this paper,a software testing scheme based on blade end timing is proposed.Combining the inherent properties of the measured system and blade rotation characteristics,the accuracy of the measured blade end timing signals of axial and circumferential vibration is verified by software algorithm.On the basis of the determination of the precision of the computation time,an appropriate method is devised to resolve the question of under-sampling of sampling signals caused by blade end timing measurement.A bidirectional vibration monitoring system software based on blade end timing is designed and implemented to complete the monitoring and testing function of blade end timing signal.The main work of this paper is as follows:(1)Aiming at the problem that traditional blade end vibration measurement technology cannot meet the requirements of measurement accuracy,a software vibration measurement method is proposed.Combined with the principle of blade end timing,the relevant measurement parameters of the unit and the measurement vector factors affecting vibration,the overall analysis is made.The mathematical model based on blade end timing is established,and the circumferential and axial arrival time signals of blade end vibration pulses are calculated through the setting of relevant parameters and the calculation method of blade end arrival time.The time domain diagram of vibration equation was obtained based on the analysis of related factors such as sensor characteristics and vibration Angle,and the accuracy of the calculated arrival time signal was further determined by comparing with the equipment in our laboratory.(2)Aiming at the problem of the under-sampling problem of the measured signals based on the blade end timing method,the under-sampled signals are reconstructed based on the sparse reconstructed by algorithms.Due to the high calculation of complexity of general sparse reconstructed algorithms and the potential of vibration phenomenon in the reconstruction process with the increase of the number of iterations,the measurement matrix was analyzed and designed based on the traditional orthogonal matching tracking(OMP)algorithm for further improving the measuring precision.Analysis based on particle swarm optimization(PSO),an improved sparse reconstruction method of PSO-OMP sampled signals is proposed.The standard PSO algorithm updates the locations with random numbers sampled in the mean and Gaussian distribution to improve the search efficiency and accuracy.In order to solve the problem that the improved PSOOMP reconstruction method may cause oscillations in the later period and affect the accuracy of subsequent reconstruction,the Metropolis criteria in the simulated annealing(SA)algorithm is introduced to gradually converge with a specified probability before the particle is updated and to determine if a random perturbation needs to be added to determine whether the particle needs to be updated subsequently,so as to determine whether the particle needs to be updated in the future.Finally,the improved SA-PSO-OMP algorithm was used to reconstruct the signal.(3)The whole system emulation and monitored by the software design has been accomplished,including the emulation calculations of the systems blade tip timing and the design of the monitoring of the system and verification software.The software is designed and developed on QT using C++ language,and the reliability of the proposed algorithm is tested successfully by establishing a perfect experimental platform.