Coal consumption for power supply is an important indicator to measure the economics of thermal power units. On the basis of parallel random forest algorithm

this paper proposes a calculation model of coal consumption for thermal power units. By using a large amount of DCS data of a 600 MW unit

on the platform of Spark big data

the data of stable working condition are determined by threshold

the local outliers are detected and processed by local anomaly factor algorithm

and the different working conditions are determined by K-means clustering algorithm. Finally

38 thermodynamic parameters affecting the coal consumption of power supply and 211 615 sets of data under 10 working conditions are screened out. 4/5 of the data are randomly selected to train and model the coal consumption calculation model based on the parallel random forest algorithm

and the other 1/5 data is tested. The results show that

the calculated value of the model is in good agreement with the actual value. The average absolute error is 1.79 g/(kW·h)

and the relative error is within –3%~3%

indicating this coal consumption calculation model based on parallel stochastic forest algorithm has high accuracy and strong generalization ability

which is suitable for power supply coal consumption calculation.