CN-122020402-A - Decision tree model-based coal grinding process electricity consumption diagnosis and analysis method and device

CN122020402ACN 122020402 ACN122020402 ACN 122020402ACN-122020402-A

Abstract

The invention relates to the technical field of coal grinding process electricity consumption diagnosis, in particular to a decision tree model-based coal grinding process electricity consumption diagnosis analysis method and device. The method comprises the steps of taking historical coal grinding process data, obtaining a plurality of effective historical coal grinding process data sets from the historical coal grinding process data sets, calculating historical coal grinding hour process power consumption of each effective historical coal grinding process data set based on historical coal grinding hour power consumption and historical raw coal yield in each effective historical coal grinding process data set, calculating and obtaining pearson correlation coefficients corresponding to the parameters based on parameters in the effective historical coal grinding process data sets and the historical coal grinding hour process power consumption, and determining key parameters of a decision tree model based on the pearson correlation coefficients. According to the method, whether the electricity consumption of the coal grinding process is abnormal or not can be automatically and timely judged by acquiring the coal grinding working condition score of the current coal grinding process data, and when the electricity consumption of the coal grinding process is abnormal, the reason of the abnormal electricity consumption of the coal grinding process can be automatically determined and an optimization strategy can be automatically given.

Inventors

YUAN YIBIN
WEI CAN
ZHU YONGZHI
ZHU ZHIWEN
Xiao Shouyun
Lai Defa

Assignees

邦业(杭州)智能技术有限公司

Dates

Publication Date: 20260512
Application Date: 20260113

Claims (10)

1. The utility model provides a coal mill process electricity consumption diagnosis analysis method based on decision tree model, which is characterized by comprising the following steps: Acquiring historical coal grinding process data, acquiring a plurality of effective historical coal grinding process data sets from the historical coal grinding process data, calculating the historical coal grinding hour process power consumption of each effective historical coal grinding process data set based on the historical coal grinding hour power consumption and the historical raw coal yield in each effective historical coal grinding process data set, calculating the pearson correlation coefficient corresponding to the parameter based on the parameter in the effective historical coal grinding process data set and the historical coal grinding hour process power consumption, and determining the key parameter of the decision tree model based on the pearson correlation coefficient; Acquiring a current coal grinding process data set, judging whether the current coal grinding process data set is a valid data set, if the current coal grinding process data set is the valid data set, calculating to obtain current coal grinding hour process power consumption based on current coal grinding hour power consumption and current hour raw coal yield in the current coal grinding process data set, and acquiring a coal grinding working condition score based on the current coal grinding hour process power consumption and the current coal grinding process data set; Judging whether the coal grinding working condition score is smaller than a preset score, when the coal grinding working condition score is smaller than the preset score, carrying out standardization processing on data values corresponding to key parameters in a current coal grinding working procedure data set, inputting the data values into a decision tree model to enable the decision tree model to output a condition path, determining the abnormal cause of the electricity consumption of the coal grinding working procedure based on the condition path, and determining an optimization strategy based on the abnormal cause of the electricity consumption of the coal grinding working procedure.
2. The method of claim 1, wherein obtaining a number of valid historical coal grinding process data sets from the historical coal grinding process data comprises: and acquiring a data set with the mill current being more than 25A, the running time being equal to 60 minutes within the hour and the coal mill feeding amount being more than 40t/h from the historical coal mill process data, and taking the data set as an effective historical coal mill process data set.
3. The method of claim 1, wherein the key parameters of the decision tree model include low high temperature fan current, coal mill feed, mill grinding pressure, main exhaust fan current, classifier frequency, coal mill inlet/outlet temperature difference, mill current, mill differential pressure, classifier current, coal fines fineness and raw coal moisture.
4. The method of claim 1, wherein obtaining a coal mill operating condition score based on the current coal mill hour process electricity consumption and the current coal mill process data set comprises: and obtaining an electricity consumption value based on the current coal mill hour process electricity consumption, obtaining a model value based on the current coal mill process data set, and obtaining a coal mill working condition value based on the model value, the model value weight, the electricity consumption value and the electricity consumption value weight.
5. The method of claim 4, wherein obtaining an electricity consumption value based on the current coal mill hour process electricity consumption comprises: Determining current raw coal moisture, current raw coal ash and current raw coal volatile based on the current coal grinding process data set; n effective historical coal grinding process data sets with similar working conditions are found from the effective historical coal grinding process data sets based on the current raw coal moisture, the current raw coal ash and the current raw coal volatile matters; Acquiring neighborhood average power consumption and neighborhood standard deviation power consumption based on N effective historical coal grinding procedure data sets; and obtaining an electricity consumption standardized value based on the current coal mill hour working procedure electricity consumption, the neighborhood average electricity consumption and the neighborhood standard deviation electricity consumption, calculating a standard normal cumulative distribution function value corresponding to the electricity consumption standardized value, and obtaining an electricity consumption value based on the standard normal cumulative distribution function value.
6. The method of claim 4, wherein obtaining a model score based on the current coal grinding process data set comprises: The data values corresponding to the key parameters in the current coal grinding process data set are subjected to standardized processing and then input into a decision tree model, so that the decision tree model outputs the probability of excellent electricity consumption level, the probability of good electricity consumption level and the probability of poor electricity consumption level; And determining an initial model score based on the probability that the electricity consumption level is excellent, the probability that the electricity consumption level is good and the probability that the electricity consumption level is poor, and calculating based on the initial model score to obtain a final model score.
7. The method according to claim 1, further comprising training a decision tree model based on key parameters, comprising in particular: obtaining a model training sample based on a plurality of effective historical coal grinding process data sets and key parameters; And training based on the model training sample and a decision tree algorithm to obtain a decision tree model.
8. A coal grinding process electricity consumption diagnosis and analysis device based on a decision tree model is characterized by comprising: The key parameter determining module is configured to acquire historical coal grinding process data, acquire a plurality of effective historical coal grinding process data sets from the historical coal grinding process data, calculate the historical coal grinding hour process power consumption of each effective historical coal grinding process data set based on the historical coal grinding hour power consumption and the historical raw coal yield in each effective historical coal grinding process data set, calculate the pearson correlation coefficient corresponding to the parameter based on the parameter in the effective historical coal grinding process data set and the historical coal grinding hour process power consumption, and determine the key parameter of the decision tree model based on the pearson correlation coefficient; The coal grinding working condition score obtaining module is configured to obtain a current coal grinding working procedure data set, judge whether the current coal grinding working procedure data set is a valid data set, calculate and obtain current coal grinding working condition score based on current coal grinding hour power consumption and current hour raw coal yield in the current coal grinding working procedure data set if the current coal grinding working procedure data set is the valid data set, and obtain the coal grinding working condition score based on the current coal grinding hour working procedure power consumption and the current coal grinding working procedure data set; the power consumption abnormality cause determining module is configured to determine whether the coal grinding working condition score is smaller than a preset score, when the coal grinding working condition score is smaller than the preset score, input a data value corresponding to a key parameter in a current coal grinding working procedure data set into a decision tree model after standardized processing so that the decision tree model outputs a condition path, determine the power consumption abnormality cause of the coal grinding working procedure based on the condition path, and determine an optimization strategy based on the power consumption abnormality cause of the coal grinding working procedure.
9. A computer program product comprising a computer program which, when executed by a processor, implements the steps of the method of any of claims 1-7.
10. An electronic device comprising one or more processors and memory associated with the one or more processors, the memory for storing program instructions that, when read and executed by the one or more processors, perform the steps of the method of any of claims 1-7.

Description

Decision tree model-based coal grinding process electricity consumption diagnosis and analysis method and device Technical Field The invention relates to the technical field of coal grinding process electricity consumption diagnosis, in particular to a decision tree model-based coal grinding process electricity consumption diagnosis analysis method and device. Background The coal grinding system is one of the key links of cement clinker production, and has the main functions of drying raw coal and grinding the raw coal into coal dust (indexes such as fineness, moisture, heat value and the like) meeting the firing requirements of the rotary kiln, so as to provide fuel for the rotary kiln calcination, and the coal grinding is an important component in the clinker firing. According to statistics, the electricity consumption of the coal grinding system accounts for about 8% -12%, is a key target point of energy conservation and consumption reduction of cement enterprises, and needs to be analyzed and diagnosed when the energy consumption of the grinding process is abnormal, so that the purposes of energy conservation and consumption reduction are achieved. At present, the abnormal electricity consumption diagnosis of the coal grinding process in each cement factory mainly depends on manual analysis, the manual analysis is limited by the technical level of process personnel and the professional limit of process and quality, and the reasons are difficult to position in time and optimize the operation parameters when the abnormal electricity consumption of the grinding process is generated. Disclosure of Invention The invention aims to provide a decision tree model-based coal grinding process electricity consumption diagnosis and analysis method and device, which can automatically and timely determine the cause of abnormal coal grinding process electricity consumption and automatically give out an optimization strategy when the coal grinding process electricity consumption is abnormal. In a first aspect of the embodiment of the present invention, a method for diagnosing and analyzing power consumption of a coal grinding process based on a decision tree model is provided, including: Acquiring historical coal grinding process data, acquiring a plurality of effective historical coal grinding process data sets from the historical coal grinding process data, calculating the historical coal grinding hour process power consumption of each effective historical coal grinding process data set based on the historical coal grinding hour power consumption and the historical raw coal yield in each effective historical coal grinding process data set, calculating the pearson correlation coefficient corresponding to the parameters based on the parameters in the effective historical coal grinding process data sets and the historical coal grinding hour process power consumption, and determining key parameters of a decision tree model based on the pearson correlation coefficient; Acquiring a current coal grinding process data set, judging whether the current coal grinding process data set is a valid data set, if the current coal grinding process data set is the valid data set, calculating to obtain current coal grinding hour process electricity consumption based on current coal grinding hour electricity consumption and current hour raw coal yield in the current coal grinding process data set, and acquiring a coal grinding working condition score based on the current coal grinding hour process electricity consumption and the current coal grinding process data set; Judging whether the coal grinding working condition score is smaller than a preset score, when the coal grinding working condition score is smaller than the preset score, carrying out standardization processing on data values corresponding to key parameters in the current coal grinding working procedure data set, inputting the data values into a decision tree model, enabling the decision tree model to output a condition path, determining the abnormal cause of the electric consumption of the coal grinding working procedure based on the condition path, and determining an optimization strategy based on the abnormal cause of the electric consumption of the coal grinding working procedure. Preferably, the obtaining a plurality of valid historical coal grinding process data sets from the historical coal grinding process data includes: and acquiring a data set with the mill current being more than 25A, the running time being equal to 60 minutes within the hour and the coal mill feeding amount being more than 40t/h from the historical coal mill process data, and taking the data set as an effective historical coal mill process data set. As the optimization of the embodiment of the invention, key parameters of the decision tree model comprise small high-temperature fan current, coal mill feeding amount, mill grinding pressure, main exhaust fan current, powder concentrator frequency, c