Study On Operation Mode Optimization Of Coal Power Generating Units Based On Knowledge Of Power Production Process

With the standardized and intelligent development of the power industry in China,the level of informatization of power production is getting higher and higher.The mass data generated in the process of power production has become one of the important means to improve the coal utilization efficiency and reduce energy consumption and environmental pollution as the data support for the research on the operation mode of coal-fired power units.The data generated in the process of power production has the characteristics of large volume and complex data structure,so it is feasible to study the operation mode of coal power generating units based on the knowledge of power production process.Therefore,this article build knowledge and power production process in the electric power production process knowledge mining method can promote the efficiency of the coal unit operation mode,the use of coal unit produce the large amount of historical data in the operation of the operation parameters of correlation analysis,through clustering decision rules mining to evaluate coal unit operation mode,and put forward feasible coal unit operation method,improve the economic benefit of coal unit operation,reduce energy consumption and environmental pollution.Electric power production in this thesis,based on a power plant distributed control system DCS and SIS database power monitoring information system,draw on a 1000 MW power station a unit operation data in 2018,first of all,based on the time series of data cleaning,building coal unit operation characteristics of the data set,data clustering makes data meet the health sample data of big data analysis;Secondly,the gray relational degree analysis method is used to select the characteristic indexes that have a major impact on the operationmode of coal-fired power units.K-means algorithm is used to select variables with different characteristics and determine the reference value.Combined with GRNN neural network,the target value of the operating mode of coal-fired power units is predicted.Thirdly,boundary conditions of coal power unit load and temperature were used to classify the stable operating conditions,and kmeans clustering algorithm was used to cluster the operating conditions of the divided units,and the correlation degree between the data was calculated to obtain the data combination under excellent sample data.Finally,the method designed in this thesis,the original design method and the actual method are compared to prove the superiority of the method proposed in this thesis to study the operation mode of big data coal-fired power generating units in power production process.The average coal consumption of the proposed method is 305.52g/(k W·h)through experiments,the average coal consumption of the designed method is314.73g/(k W·h),and the average coal consumption of the current actual method is 315.27g/((k W·h).Compared in this thesis,the research methods in the original design method and actual method under the condition of guarantee the same power output lower coal consumption,compared with the current actual average coal consumption reduced 9.75 g/(k W·h),if the generator set to run 8000 hours,the cumulative output 6.26 x 109 k W·h,using the method can save about 26600 tons of standard coal,a year,not only greatly reduces the coal consumption,and reduced emissions of pollutants increased coal unit’s ability to energy saving and emission reduction.