Turbine Blades Temperature Feature Extraction And Measurement Software Implementation

In recent years,security and energy-saving are two important principles in the reform of electricity power industry. In order to meet such demand, the gas turbine power generation mode using natural gas as energy is gradually replacing the coal combustion and obtaining more and more extensive application. As the transmission part of gas turbine, health condition of the turbine blade has great influences on the equipment safety. Therefore, monitoring the blade health condition during the gas turbine operation process is of great significance to guarantee the smooth operation of equipment.In this paper,the turbine blade temperature monitoring system is designed and realized.The sensors of this system convert radiation energy of the blade to electrical signals by radiation temperature measurement technology. The signal acquisition and analysis part executes digitization sampling, then transfers the results to the industrial computer for completing the calculation, display, analysis and storage using virtual instrument technology.The temperature measurement software developed in the Lab Windows is the core of the monitoring system, as well as the focus of work in this paper.So as to determine the characteristic information of the blade conditions, in this paper,the data analysis method is firstly discussed combined with the blade physical structure and the actual operation records of a certain type of turbine blade. The maximum temperature,minimum temperature and average temperature of each blade are defined as the basic temperature values. The basic temperature center distance is proposed as the turbine blade temperature uniformity index, the basic temperature deviation is proposed to locate the suspected failure blade. Taking blade temperature waveform as the research object, this paper proposes the feature vector as evaluating index of the blade health conditions. Discussions are executed combined with multi clustering methods. Determine the cluster center can be used as standard feature vector in health assessment by the fuzzy clustering method. The K mean clustering algorithm can achieve further classification of the blade. It is proved that the feature vector has better classification effect through applying multiple distance criterions for system clustering.In this paper, after determining the information content that should be paid attention to in calculation process, development of the temperature monitoring system is realized in the Lab Windows. The software which consist signal acquisition module, data processing module,interface interaction module and data management module has ability to control the external equipment collecting voltage signal and temperature of each blade through calculation and analysis. The turbine blade temperature monitoring system can realize online monitoring of the blade temperature, so as to guarantee the smooth operation of the gas turbine and provide actual data for the theoretical research of blade health assessment.