Research On Abnormal Energy Efficiency State Pattern Recognition And On-line Performance Test Of Coal-fired Units

In this paper,the abnormal pattern recognition of energy efficiency status and on-line performance test of coal-fired generating units are studied by means of scientific theory and method,which provides the basis for operation optimization and maintenance decision-making of units.Based on the analysis of thermal system and equipment,the energy efficiency characteristic parameter system is established.The cluster method is adopted to select the decision samples of energy efficiency benchmark state,the kernel density estimation model is employed to estimate the reference value of characteristic parameters and the interval of deterioration threshold under the operating condition of the system by using the stochastic forest method.Through the analysis of the professional mechanism of the system,the potential failure modes which affect the abnormal energy efficiency state of the unit are analyzed by FMEA,and the abnormal energy efficiency pattern tree of the unit is established by using the FTA method.The real-time energy efficiency monitoring system detects that the energy efficiency characteristic parameters exceed the threshold value,which will result in abnormal energy efficiency events and transmit them to the pattern recognition system.The pattern recognition system adopts the method based on FTA to realize the recognition of the abnormal energy efficiency state pattern,and gives the corresponding treatment suggestions according to the FMEA analysis.Based on the overhaul period of the unit,according to the principle and calculation method of thermal test,the overall energy efficiency of the unit,system and equipment is analyzed and evaluated by using the thermodynamic parameters measured online by the unit.Compare the changes of the overall energy efficiency state of the unit before and after overhaul and renovation under different adjustment modes.Establish a comprehensive evaluation model of equipment performance,track the deterioration trend of equipment performance,guide the maintenance time and maintenance project decision.