基于多参数的电站风机监测技术的试验研究

采用试验研究的方法，对电站通风机的性能、非稳态流动和机械振动的多参数监测技术进行了研究。基于无因次性能曲线反映出的性能参数间稳定、良好的映射规律，采用具备优良逼近特性RBF网络逼近风机无因次性能曲线，推导出基于参数映射的流量监测模型，实现了风机性能的在线监测。通过对4-73风机的吸力侧、压力侧旋转失速、进口涡流的频率特性研究，分析了三种非稳态流动的特点，给出了准确描述三种非稳态流动的联合特征参数。将通风机机械振动特征分为谐波特征、能量特征和奇异性特征，采用分频段技术、二进小波变换方法导出了谐波监测指标、能量指标和奇异性指标。

Experimental Investigation of Multiple Parameter-based Monitoring Technology for a Power Plant Air Blower

By using a experimental study method an investigation was conducted of a multiple parameter-based monitoring technology involving the performance, non-steady state flow and mechanical vibrations of a power station air blower. On the basis of the stable and good mapping mechanism existing among the performance parameters as reflected by non-dimensional performance curves, a RBF (Radial Basis Function) network featuring excellent approximation characteristics was employed to approximate the non-dimensional performance curves of the air blower. As a result, a parameter mapping-based flow-monitoring model was derived, thereby realizing the on-line monitoring of the air blower performance. Through a study of the rotating stall at the 4-73 air blower suction and pressure side and the frequency characteristics of inlet vortex flow and an analysis of three kinds of non-steady flow specific features given are combined eigen parameters capable of accurately describing three kinds of non-steady state flows. Mechanical vibration characteristics of the air blower are divided into harmonic, energy and singularity characteristics. By using frequency-division section technology and a binary small-wave transformation method derived are harmonic monitoring indexes, energy and singularity indexes.