Study On The Signal Of Compressor Stall And Online Identification Arithmetic

Through a detailed research on the feature of stall inception in low speed axial compressor with different inlet conditions and on the development feature of three sets of typical high speed axial compressors under different speed,obtained stall inception feature of rotating stall / surge,and on this basis,developing online identification software applied to compressor stall.Analysis of the compressor stall inception feature under the condition of uniform and swirl distortion,results show that:(1)The compressor's stall inception type is modal wave,different inlet conditions do not affect the type of stall inception.(2)Under the effect of swirl distortion,the air flow's inlet angle was increased in the compressor;With the blade angle increased,the airflow angle increased,the more easy for the compressor to develop into stall and the time from the stall inception to full stall was shorter.(3)For low speed axial compressor,the modal wave signal was obvious before stall,and the amplitude increased gradually,so it could be thought the compressor was going to fall into stall when the disturbance exceeded the threshold value that set before,which achieved the purpose of online identification.Studied the rotating stall / surge feature of three kinds typical high speed axial compressor,results showed that:(1)Stall development of the three compressors was different.However,for the on-line identification,the measurement points should be arranged in the downstream where captured the signal relatively early.(2)The signal of the pressure amplitude was similar to modal wave before stall was appeared,and the amplitude increased with the decrease of the flow rate.When the signal suddenly increased,the compressor fell into stall.(3)Near design speed,the compressor inlet appeared stall inception signal before stall,and in some cases the rotor inlet fell into stall firstly,and then the post.(4)When deviated from the design speed and low speed,the inlet flow effect the first rotor stage blades,with the throttle closed,rotor blades were first separated and developed into stall and propagated to the upstream and downstream.Based on the above research,an online identification algorithm for compressor stall was developed.The algorithm was simple and fast,and can capture the stall signal as early as possible.Stall feature of the development as the threshold reference standard,embedded into online identification algorithm,which can be as the standard for the compressor to fall into stall.Making a judgment in the sampling period of the collected data,once it reaches a set threshold value,a warning forecast was sent out.