Coal milling process power consumption diagnostic analysis method and device based on decision tree model

Abstract

The invention relates to the technical field of coal milling process power consumption diagnosis, in particular to a coal milling process power consumption diagnosis analysis method and device based on a decision tree model. The method comprises the following steps: acquiring historical coal milling process data, and acquiring a plurality of effective historical coal milling process data sets from the historical coal milling process data; calculating historical coal milling hour working procedure power consumption of each effective historical coal milling working procedure data set based on historical coal milling hour power consumption and historical hour raw coal yield in each effective historical coal milling working procedure data set; calculating to obtain a Pearson's correlation coefficient corresponding to the parameters based on the parameters in the effective historical coal milling process data set and historical coal milling hour process power consumption, and determining key parameters of the decision tree model based on the Pearson's correlation coefficient. According to the mode, whether the coal milling process power consumption is abnormal or not can be automatically judged in time by obtaining the coal milling working condition score of the current coal milling process data, and when the coal milling process power consumption is abnormal, the coal milling process power consumption abnormity reason can be automatically determined, and an optimization strategy is automatically given.

Description

Technical Field

[0001] This invention relates to the field of coal milling process power consumption diagnosis technology, specifically to a method and apparatus for coal milling process power consumption diagnosis and analysis based on a decision tree model. Background Technology

[0002] The coal mill system is one of the core components of cement clinker production. Its main function is to dry and grind raw coal into pulverized coal (with properties such as fineness, moisture content, and calorific value) that meet the requirements of rotary kiln calcination, providing fuel for the rotary kiln. The coal mill is an important part of clinker calcination. Statistics show that the coal mill system accounts for approximately 8%-12% of electricity consumption, making it a key target for energy conservation and consumption reduction in cement enterprises. When energy consumption in the mill process is abnormal, it is necessary to analyze and diagnose it promptly to achieve the goal of energy conservation and consumption reduction. Currently, the diagnosis of abnormal electricity consumption in the coal mill process at various cement plants mainly relies on manual analysis. However, manual analysis is limited by the technical level of process personnel and the professional limitations of process and quality control, making it difficult to pinpoint the cause and optimize operating parameters in a timely manner when electricity consumption in the mill process is abnormal. Summary of the Invention

[0003] The purpose of this invention is to provide a method and apparatus for diagnosing and analyzing the power consumption of coal milling processes based on a decision tree model. When the power consumption of the coal milling process is abnormal, it can automatically and promptly determine the cause of the abnormal power consumption and automatically provide optimization strategies.

[0004] In a first aspect of the present invention, a method for diagnosing and analyzing power consumption in a coal milling process based on a decision tree model is provided, comprising:

[0005] Historical coal milling process data is obtained, and several effective historical coal milling process data sets are obtained from the historical coal milling process data sets. Based on the historical coal milling hourly power consumption and historical hourly raw coal output in each effective historical coal milling process data set, the historical coal milling hourly process power consumption of each effective historical coal milling process data set is calculated. Based on the parameters in the effective historical coal milling process data sets and the historical coal milling hourly process power consumption, the Pearson correlation coefficient corresponding to the parameters is calculated, and the key parameters of the decision tree model are determined based on the Pearson correlation coefficient.

[0006] Obtain the current coal mill process data set, determine whether the current coal mill process data set is a valid data set. If the current coal mill process data set is a valid data set, calculate the current coal mill hourly process power consumption based on the current coal mill hourly power consumption and the current hourly raw coal output in the current coal mill process data set, and obtain the coal mill operating condition score based on the current coal mill hourly process power consumption and the current coal mill process data set.

[0007] Determine whether the coal mill operating condition score is less than the preset score. When the coal mill operating condition score is less than the preset score, standardize the data values ​​corresponding to the key parameters in the current coal mill process data group and input them into the decision tree model so that the decision tree model outputs a conditional path. Based on the conditional path, determine the cause of abnormal power consumption in the coal mill process, and determine the optimization strategy based on the cause of abnormal power consumption in the coal mill process.

[0008] As a preferred embodiment of the present invention, obtaining several sets of valid historical coal mill process data from historical coal mill process data includes:

[0009] Data sets with mill current greater than 25A, hourly running time equal to 60 minutes, and coal mill feed rate greater than 40t / h were obtained from historical coal mill process data and were used as valid historical coal mill process data sets.

[0010] As a preferred embodiment of the present invention, the key parameters of the decision tree model include: the current of the small high-temperature fan, the feed rate of the coal mill, the grinding pressure of the mill, the current of the main exhaust fan, the frequency of the classifier, the temperature difference between the inlet and outlet of the coal mill, the mill current, the mill differential pressure, the current of the classifier, the fineness of the coal powder, and the moisture content of the raw coal.

[0011] As a preferred embodiment of the present invention, obtaining the coal mill operating condition score based on the current hourly process power consumption of the coal mill and the current coal mill process data set includes:

[0012] The power consumption score is obtained based on the current hourly power consumption of the coal mill process. The model score is obtained based on the current coal mill process data set. The coal mill operating condition score is calculated based on the model score, model score weight, power consumption score, and power consumption score weight.

[0013] As a preferred embodiment of the present invention, obtaining the power consumption score based on the current hourly power consumption of the coal mill process includes:

[0014] Determine the current raw coal moisture content, ash content, and volatile matter content based on the current coal milling process data set;

[0015] Based on the current raw coal moisture content, current raw coal ash content, and current raw coal volatile matter, find N valid historical coal mill process data sets with similar operating conditions from several valid historical coal mill process data sets;

[0016] The average power consumption and standard deviation power consumption of the neighborhood are obtained based on N valid historical coal mill process data sets.

[0017] Based on the current hourly process power consumption of coal mill, the average power consumption of the neighborhood, and the standard deviation of the power consumption of the neighborhood, the standardized value of power consumption is obtained. The standard normal cumulative distribution function value corresponding to the standardized value of power consumption is calculated, and the power consumption score is obtained based on the standard normal cumulative distribution function value.

[0018] As a preferred embodiment of the present invention, obtaining the model score based on the current coal milling process data set includes:

[0019] After standardizing the data values ​​corresponding to the key parameters in the current coal mill process data set, the data is input into the decision tree model so that the decision tree model outputs the probability of an excellent power consumption level, the probability of a good power consumption level, and the probability of a poor power consumption level.

[0020] The initial model score is determined based on the probability of having an excellent power consumption level, a good power consumption level, and a poor power consumption level. The final model score is then calculated based on the initial model score.

[0021] As a preferred embodiment of the present invention, it further includes training a decision tree model based on key parameters, which specifically includes:

[0022] Model training samples were obtained based on several valid historical coal mill process data sets and key parameters.

[0023] The decision tree model is obtained by training the model training samples and the decision tree algorithm.

[0024] In a second aspect of the present invention, a coal milling process power consumption diagnostic analysis device based on a decision tree model is provided, comprising:

[0025] The key parameter determination module is configured to acquire historical coal mill process data, obtain several valid historical coal mill process data sets from the historical coal mill process data, calculate the historical coal mill hourly process power consumption of each valid historical coal mill process data set based on the historical coal mill hourly power consumption and historical hourly raw coal output in each valid historical coal mill process data set, calculate the Pearson correlation coefficient corresponding to the parameters based on the parameters in the valid historical coal mill process data set and the historical coal mill hourly process power consumption, and determine the key parameters of the decision tree model based on the Pearson correlation coefficient.

[0026] The coal mill operating condition score acquisition module is configured to acquire the current coal mill process data group, determine whether the current coal mill process data group is a valid data group, and if the current coal mill process data group is a valid data group, calculate the current coal mill hourly process power consumption based on the current coal mill hourly power consumption and the current hourly raw coal output in the current coal mill process data group, and acquire the coal mill operating condition score based on the current coal mill hourly process power consumption and the current coal mill process data group.

[0027] The module for determining the cause of abnormal power consumption is configured to determine whether the coal mill operating condition score is less than a preset score. When the coal mill operating condition score is less than the preset score, the data values ​​corresponding to the key parameters in the current coal mill process data group are standardized and then input into the decision tree model, so that the decision tree model outputs a conditional path. Based on the conditional path, the module determines the cause of abnormal power consumption in the coal mill process and determines the optimization strategy based on the cause of abnormal power consumption in the coal mill process.

[0028] In a third aspect of the present invention, a computer program product is provided, including a computer program that, when executed by a processor, implements the method provided according to the first aspect.

[0029] In a fourth aspect of the present invention, an electronic device is provided, including one or more processors and a memory associated with the one or more processors, the memory being used to store program instructions that, when read and executed by the one or more processors, perform the method provided in the first aspect.

[0030] In summary, the present invention has the following beneficial effects:

[0031] The method of this invention first determines the key parameters of the decision tree model based on the Pearson correlation coefficient, then obtains the coal mill operating condition score based on the current hourly coal mill process power consumption and the current coal mill process data set. Next, when the coal mill operating condition score is less than a preset score, the data values ​​corresponding to the key parameters in the current coal mill process data set are standardized and then input into the decision tree model so that the decision tree model can automatically output the conditional path. Finally, the method automatically determines the cause of abnormal coal mill process power consumption and optimization strategy based on the conditional path. This method can automatically and timely determine whether the coal mill process power consumption is abnormal by obtaining the coal mill operating condition score of the current coal mill process data. When the coal mill process power consumption is abnormal, it can automatically determine the cause of the abnormal coal mill process power consumption and automatically give optimization strategies.

[0032] Further or more detailed beneficial effects will be described in conjunction with specific embodiments in the detailed implementation. Attached Figure Description

[0033] Figure 1 A flowchart of the coal milling process power consumption diagnosis and analysis method based on a decision tree model according to an embodiment of the present invention is shown;

[0034] Figure 2 A block diagram of a coal milling process power consumption diagnostic analysis device based on a decision tree model according to an embodiment of the present invention is shown.

[0035] Figure 3 A block diagram of an electronic device according to an embodiment of the present invention is shown. Detailed Implementation

[0036] Embodiments of the present invention will now be described in more detail with reference to the accompanying drawings. While some embodiments of the invention are shown in the drawings, it should be understood that the invention can be implemented in various forms and should not be construed as limited to the embodiments set forth herein. Rather, these embodiments are provided to provide a more thorough and complete understanding of the invention. It should be understood that the accompanying drawings and embodiments are for illustrative purposes only and are not intended to limit the scope of protection of the invention.

[0037] In the description of embodiments of the present invention, the term "comprising" and similar terms should be understood as open-ended inclusion, i.e., "including but not limited to". The term "based on" should be understood as "at least partially based on". The term "one embodiment" or "the embodiment" should be understood as "at least one embodiment". The terms "first", "second", etc., may refer to different or the same objects. Other explicit and implicit definitions may also be included below.

[0038] Figure 1 A flowchart of a coal milling process power consumption diagnosis and analysis method 100 based on a decision tree model, according to an embodiment of the present invention, is shown. The method 100 includes:

[0039] Step 102. Obtain historical coal mill process data. Obtain several valid historical coal mill process data sets from the historical coal mill process data. Calculate the historical coal mill hourly process power consumption and historical hourly raw coal output of each valid historical coal mill process data set. Calculate the Pearson correlation coefficient corresponding to the parameters based on the parameters in the valid historical coal mill process data sets and the historical coal mill hourly process power consumption. Determine the key parameters of the decision tree model based on the Pearson correlation coefficient.

[0040] This embodiment assumes today is November 8th, XX year. Therefore, the historical coal mill process data can be data from January to October of XX year. This step can group the historical coal mill process data by hour, with each hour's data forming one group of historical coal mill process data. For example, the data from 6:00 AM to 7:00 AM on January 8th, XX year is one group of historical coal mill process data, and the data from 7:00 AM to 8:00 AM on January 8th, XX year is another group of historical coal mill process data. Each group of historical coal mill process data includes parameters such as the current of the small high-temperature fan, the coal mill feed rate, the mill grinding pressure, the current of the main exhaust fan, the frequency of the classifier, the temperature difference between the coal mill inlet and outlet, the mill current, the mill differential pressure, the classifier current, the fineness of the coal powder, the moisture content of the raw coal, the ash content of the raw coal, the volatile matter of the raw coal, the hourly raw coal output, and the hourly power consumption of the coal mill.

[0041] This step also determines the validity of each set of historical coal mill process data. If the mill current in a set of historical coal mill process data is greater than 25A, the hourly operating time is equal to 60 minutes, and the coal mill feed rate is greater than 40 t / h, then that set of historical coal mill process data is considered a valid set of historical coal mill process data. This embodiment assumes that 1000 sets of valid historical coal mill process data were ultimately obtained.

[0042] Each set of valid historical coal mill process data contains one historical hourly raw coal output and one historical hourly coal mill power consumption. Dividing the historical hourly coal mill power consumption by the historical hourly raw coal output yields the historical hourly coal mill process power consumption. Since there are 1000 sets of valid historical coal mill process data, 1000 historical hourly coal mill process power consumptions can be obtained.

[0043] Regarding the parameter of "small high-temperature fan current," each valid historical coal mill process data set contains one historical small high-temperature fan current. Since there are 1000 valid historical coal mill process data sets, 1000 historical small high-temperature fan currents can be obtained. Based on these 1000 historical small high-temperature fan currents and 1000 historical coal mill hourly process power consumption data, a Pearson correlation coefficient corresponding to the "small high-temperature fan current" can be calculated (the specific calculation method for the Pearson correlation coefficient is existing technology). Finally, the magnitude of this Pearson correlation coefficient can be used to determine whether the "small high-temperature fan current" is a key parameter of the decision tree model. In this embodiment, the "small high-temperature fan current" is a key parameter of the decision tree model.

[0044] The same procedure applies to other parameters in the valid historical coal mill process data set. In this embodiment, the key parameters of the final decision tree model include the current of the small high-temperature fan, the coal mill feed rate, the mill grinding pressure, the current of the main exhaust fan, the frequency of the classifier, the temperature difference between the coal mill inlet and outlet, the mill current, the mill differential pressure, the classifier current, the fineness of the coal powder, and the moisture content of the raw coal.

[0045] Step 104. Obtain the current coal mill process data set and determine whether the current coal mill process data set is a valid data set. If the current coal mill process data set is a valid data set, calculate the current coal mill hourly process power consumption based on the current coal mill hourly power consumption and the current hourly raw coal output in the current coal mill process data set, and obtain the coal mill operating condition score based on the current coal mill hourly process power consumption and the current coal mill process data set.

[0046] This embodiment assumes that one set of coal mill process data from 8:00 AM to 9:00 AM on November 8th, XX year has been obtained. This set of data is the current coal mill process data set. Then, it is necessary to determine whether the current coal mill process data set is a valid data set. If the mill current in the current coal mill process data set is greater than 25A, the hourly running time is equal to 60 minutes, and the coal mill feed rate is greater than 40 t / h, then the current coal mill process data set is a valid data set; otherwise, the current coal mill process data set is an invalid data set. If the current coal mill process data set is invalid, then the next set of coal mill process data (one set of coal mill process data from 9:00 AM to 10:00 AM on November 8th, XX year) is obtained, and this next set of coal mill process data is used as the current coal mill process data set, and then the determination of whether the current coal mill process data set is a valid data set is repeated.

[0047] If the current coal mill process data group is a valid data group, then obtain the current hourly raw coal output and the current hourly coal mill power consumption in the current coal mill process data group, and divide the current hourly coal mill power consumption by the current hourly raw coal output to obtain the current hourly coal mill process power consumption.

[0048] Next, it is necessary to obtain the coal mill operating condition score. In this step, obtaining the coal mill operating condition score based on the current hourly process power consumption of the coal mill and the current coal mill process data set specifically includes: obtaining the power consumption score based on the current hourly process power consumption of the coal mill, obtaining the model score based on the current coal mill process data set, and calculating the coal mill operating condition score based on the model score, model score weight, power consumption score, and power consumption score weight.

[0049] In this step, the electricity consumption score is obtained based on the current hourly process electricity consumption of the coal mill, including:

[0050] Step 31. Determine the current raw coal moisture content, ash content, and volatile matter content based on the current coal milling process data set. This embodiment assumes that the current raw coal moisture content is 8.9%, the current raw coal ash content is 22.1%, and the current raw coal volatile matter content is 28.3%.

[0051] Step 32. Based on the current raw coal moisture content, current raw coal ash content, and current raw coal volatile matter, find N valid historical coal milling process data sets with similar operating conditions from several valid historical coal milling process data sets. For one valid historical coal milling process data set, historical raw coal moisture content, historical raw coal ash content, and historical raw coal volatile matter can be obtained from this set. One coal quality distance can be calculated using these historical raw coal moisture content, historical raw coal ash content, historical raw coal volatile matter, current raw coal moisture content, current raw coal ash content, and current raw coal volatile matter (specifically, first calculate the variance of historical raw coal moisture content and current raw coal moisture content, simultaneously calculate the variance of historical raw coal ash content and current raw coal ash content, and simultaneously calculate the variance of historical raw coal volatile matter and current raw coal volatile matter, then sum the three variances and take the square root). Because there are 1000 valid historical coal milling process data sets, 1000 coal quality distances can be obtained. Next, the 1000 coal quality distances are sorted from smallest to largest to find the N valid historical coal milling process data sets with the smallest coal quality distance. In this embodiment, N can be 50, so step 32 obtains 50 valid historical coal milling process data sets.

[0052] Step 33. Obtain the neighborhood average power consumption and neighborhood standard deviation power consumption based on N valid historical coal mill process data sets.

[0053] In step 102, the power consumption for each historical coal milling process data set (one hour) has been calculated. This step has 50 valid historical coal milling process data sets, resulting in 50 historical hourly power consumption figures. Averaging these 50 historical hourly power consumption figures yields a neighborhood average power consumption, and calculating the standard deviation of these figures yields a neighborhood standard deviation power consumption (the specific calculation method for the standard deviation is described in existing technology).

[0054] Step 34. Obtain the standardized value of power consumption based on the current hourly process power consumption of coal mill, the average power consumption of the neighborhood, and the standard deviation of the power consumption of the neighborhood. Calculate the standard normal cumulative distribution function value corresponding to the standardized value of power consumption, and obtain the power consumption score based on the standard normal cumulative distribution function value.

[0055] In this step, the standardized power consumption value is obtained by subtracting the neighborhood average power consumption from the current hourly power consumption of the coal mill and then dividing by the neighborhood standard deviation power consumption. Assuming the standardized power consumption value is 0.75, the corresponding standard normal cumulative distribution function value is 0.773, which can be obtained using the formula... The calculation yields the power consumption score based on the standard normal cumulative distribution function value. Specifically, this is done by first subtracting the standard normal cumulative distribution function value from 1, then multiplying by 40, and finally adding 60. This is achieved through the formula... The power consumption score can then be calculated. In this embodiment, the calculated power consumption score is 69.

[0056] In this step, the model score obtained based on the current coal mill process data set includes:

[0057] Step 41. After standardizing the data values ​​corresponding to the key parameters in the current coal mill process data group, input them into the decision tree model so that the decision tree model outputs the probability of excellent power consumption level, good power consumption level, and poor power consumption level.

[0058] In step 102 of this embodiment, 1000 sets of valid historical coal mill process data have been obtained, and the key parameters have been determined as follows: small high-temperature fan current, coal mill feed rate, mill grinding pressure, main exhaust fan current, classifier frequency, coal mill inlet and outlet temperature difference, mill current, mill differential pressure, classifier current, coal powder fineness, and raw coal moisture content.

[0059] Taking the current of a small high-temperature fan as an example, 1000 historical small high-temperature fan currents can be obtained through 1000 sets of valid historical coal mill process data. From these 1000 historical small high-temperature fan currents, the maximum and minimum historical small high-temperature fan currents can be obtained. First, the minimum historical small high-temperature fan current is subtracted from the current small high-temperature fan current in the current coal mill process data set to obtain the first difference. Then, the minimum historical small high-temperature fan current is subtracted from the maximum historical small high-temperature fan current to obtain the second difference. Finally, the standardized value of the current small high-temperature fan current is obtained by dividing the first difference by the second difference.

[0060] Other key parameters are handled similarly. This step ultimately yields the standardized values ​​for the current of the small high-temperature blower, the current feed rate of the coal mill, the current grinding pressure of the mill, the current current of the main exhaust fan, the current frequency of the classifier, the current temperature difference between the inlet and outlet of the coal mill, the current of the mill, the current differential pressure of the mill, the current of the classifier, the fineness of the pulverized coal, and the moisture content of the raw coal.

[0061] Next, the standardized values ​​of the current small high-temperature fan current, the current coal mill feed rate, the current mill grinding pressure, the current main exhaust fan current, the current classifier frequency, the current coal mill inlet and outlet temperature difference, the current mill current, the current mill differential pressure, the current classifier current, the current coal powder fineness, and the current raw coal moisture content are input into the decision tree model. The decision tree model can automatically output the probabilities of an excellent power consumption level, a good power consumption level, and a poor power consumption level. For example, the decision tree model outputs a probability of 0.15 for an excellent power consumption level, 0.3 for a good power consumption level, and 0.55 for a poor power consumption level.

[0062] Step 42. Determine the initial model score based on the probability of an excellent power consumption level, the probability of a good power consumption level, and the probability of a poor power consumption level. Calculate the final model score based on the initial model score.

[0063] In this step, the initial model score equals "the probability of an excellent power consumption level multiplied by 1" plus "the probability of a good power consumption level multiplied by 0" plus "the probability of a poor power consumption level multiplied by -1". For example, when the probability of an excellent power consumption level is 0.15, the probability of a good power consumption level is 0.3, and the probability of a poor power consumption level is 0.55, the initial model score is -0.4. The final model score can be obtained by adding "80" to "the initial model score multiplied by 20". In this embodiment, the calculated model score is 72.

[0064] Finally, this step calculates the coal mill operating condition score using the model score, model score weight, power consumption score, and power consumption score weight. Specifically, the coal mill operating condition score equals "model score multiplied by model score weight" plus "power consumption score multiplied by power consumption score weight". In this embodiment, the model score weight can be 0.8, and the power consumption score weight can be 0.2. When the model score is 72 and the power consumption score is 69, the final calculated coal mill operating condition score is 71.4.

[0065] In addition, before performing step 104, this embodiment also includes training a decision tree model based on key parameters, specifically including:

[0066] Model training samples were obtained based on several valid historical coal mill process data sets and key parameters.

[0067] In step 102 of this embodiment, 1000 sets of valid historical coal mill process data have been obtained, and the key parameters have been determined as follows: small high-temperature fan current, coal mill feed rate, mill grinding pressure, main exhaust fan current, classifier frequency, coal mill inlet and outlet temperature difference, mill current, mill differential pressure, classifier current, coal powder fineness, and raw coal moisture content.

[0068] Taking the current of a small high-temperature fan as an example, 1000 historical small high-temperature fan currents can be obtained through 1000 sets of valid historical coal mill process data. From these 1000 historical small high-temperature fan currents, the maximum and minimum historical small high-temperature fan currents can be obtained. Similarly, through 1000 sets of valid historical coal mill process data, the maximum and minimum historical coal mill feed rates, the maximum and minimum historical mill grinding pressures, and so on, can also be obtained.

[0069] Taking the first set of valid historical coal mill process data as an example, the standardized value of the historical small high-temperature fan current can be obtained through the current of the small high-temperature fan, the maximum value of the historical small high-temperature fan current, and the minimum value of the historical small high-temperature fan current in this set of valid historical coal mill process data. Similarly, the standardized value of the historical coal mill feed rate can be obtained through the historical coal mill feed rate, the maximum value of the historical coal mill feed rate, and the minimum value of the historical coal mill feed rate in this set of valid historical coal mill process data. Finally, the standardized values ​​of the historical small high-temperature fan current, the historical coal mill feed rate, the historical mill grinding pressure, the historical main exhaust fan current, the historical classifier frequency, the historical coal mill inlet and outlet temperature difference, the historical mill current, the historical mill differential pressure, the historical classifier current, the historical coal powder fineness, and the historical raw coal moisture content can be obtained. In addition, the power consumption level can be determined as excellent, good, or poor by using the historical hourly power consumption of the first group of valid historical coal mill process data (for example, when the historical hourly power consumption of the coal mill is less than or equal to 25 kWh / t, the power consumption level is determined as excellent and can be represented by the label "0"; when the historical hourly power consumption of the coal mill is less than 26.5 kWh / t but greater than 25 kWh / t, the power consumption level is determined as good and can be represented by the label "1"; when the historical hourly power consumption of the coal mill is greater than or equal to 26.5 kWh / t, the power consumption level is determined as poor and can be represented by the label "2"). The standardized values ​​of historical small high-temperature fan current, historical coal mill feed rate, historical mill grinding pressure, historical main exhaust fan current, historical classifier frequency, historical coal mill inlet and outlet temperature difference, historical mill current, historical mill differential pressure, historical classifier current, historical coal powder fineness, and historical raw coal moisture content of the first effective historical coal mill process data group are used as the input of the first model training sample, and the power consumption level is used as the output of the first model training sample. That is, the first model training sample has been determined at this time.

[0070] The same process can be applied to other groups of valid historical coal mill process data, ultimately yielding 1000 model training samples. Once the model training samples are ready, a decision tree model can be trained using these samples and the decision tree algorithm. Subsequently, only a set of standardized values ​​corresponding to the key parameters needs to be input into the decision tree model, and the model will automatically output the probabilities of an excellent power consumption level, a good power consumption level, and a poor power consumption level.

[0071] In addition, users only need to train the model using training samples and execute the decision tree algorithm to directly obtain the decision tree model. The specific training principle of the decision tree model is based on existing technology. The conditional path of the decision tree model can be, for example, "small high-temperature fan current ≤ 0.03 → main exhaust fan current ≤ -0.697 → classifier frequency > -0.598 → feed rate > -0.443", or "small high-temperature fan current > 0.03 → main exhaust fan current ≤ -0.697 → classifier frequency > -0.598 → feed rate > -0.443", or "small high-temperature fan current ≤ 0.03 → main exhaust fan current > -0.697 → feed rate > -0.281 → grinding pressure > 0.487".

[0072] Step 106. Determine whether the coal mill operating condition score is less than the preset score. When the coal mill operating condition score is less than the preset score, standardize the data values ​​corresponding to the key parameters in the current coal mill process data group and input them into the decision tree model so that the decision tree model outputs the condition path. Based on the condition path, determine the cause of abnormal power consumption in the coal mill process, and determine the optimization strategy based on the cause of abnormal power consumption in the coal mill process.

[0073] After obtaining a coal mill operating condition score in step 104, the device immediately executes step 106 to determine whether the coal mill operating condition score is less than a preset score. In this embodiment, the preset score can be 75. If the coal mill operating condition score is greater than or equal to 75, the device returns to step 104 to obtain the next set of current coal mill process data. If the coal mill operating condition score is less than 75, the device standardizes the data values ​​corresponding to the key parameters in the current coal mill process data set and inputs them into the decision tree model.

[0074] By using the current current of the small high-temperature fan and the historical maximum and minimum values ​​of the small high-temperature fan current in the current coal mill process data group, the standardized value of the current small high-temperature fan current can be obtained... Finally, the standardized values ​​of the current small high-temperature fan current, current coal mill feed rate, current mill grinding pressure, current main exhaust fan current, current classifier frequency, current coal mill inlet and outlet temperature difference, current mill current, current mill differential pressure, current classifier current, current coal powder fineness, and current raw coal moisture content can be obtained.

[0075] This step only requires inputting the following values ​​into the decision tree model: the current standardized value of the small high-temperature blower current, the current standardized value of the coal mill feed rate, the current standardized value of the mill grinding pressure, the current standardized value of the main exhaust fan current, the current standardized value of the classifier frequency, the current standardized value of the coal mill inlet and outlet temperature difference, the current standardized value of the mill current, the current standardized value of the mill differential pressure, the current standardized value of the classifier current, the current standardized value of the coal powder fineness, and the current standardized value of the raw coal moisture content. The decision tree model will automatically output the conditional path.

[0076] This embodiment assumes that the current output conditional path of the decision tree model is "small high-temperature fan current ≤ 0.03 → main exhaust fan current ≤ -0.697 → feed rate < -0.281 → grinding pressure ≤ 0.487". The device will automatically determine the cause of abnormal power consumption in the coal mill process based on the conditional path and determine the optimization strategy based on the cause of abnormal power consumption in the coal mill process. For example, in this conditional path, the main exhaust fan current ≤ -0.697 and feed rate < -0.281 are the causes of abnormal power consumption in the coal mill process. The optimization strategy is to increase the main exhaust fan current, adjust the system coal-air ratio, and increase the mill feed rate. After obtaining the optimization strategy, the device can send the process equipment, coal mill operating condition score, current hourly process power consumption of the coal mill, cause of abnormal power consumption in the coal mill process, and optimization strategy to relevant personnel via WeChat.

[0077] Figure 2 A block diagram of a coal milling process power consumption diagnostic analysis device 200 based on a decision tree model according to an embodiment of the present invention is shown. The device 200 includes:

[0078] The key parameter determination module 202 is configured to acquire historical coal mill process data, obtain several valid historical coal mill process data sets from the historical coal mill process data, calculate the historical coal mill hourly process power consumption of each valid historical coal mill process data set based on the historical coal mill hourly power consumption and historical hourly raw coal output in each valid historical coal mill process data set, calculate the Pearson correlation coefficient corresponding to the parameters based on the parameters in the valid historical coal mill process data set and the historical coal mill hourly process power consumption, and determine the key parameters of the decision tree model based on the Pearson correlation coefficient.

[0079] The coal mill operating condition score acquisition module 204 is configured to acquire the current coal mill process data group, determine whether the current coal mill process data group is a valid data group, and if the current coal mill process data group is a valid data group, calculate the current coal mill hourly process power consumption based on the current coal mill hourly power consumption and the current hourly raw coal output in the current coal mill process data group, and acquire the coal mill operating condition score based on the current coal mill hourly process power consumption and the current coal mill process data group.

[0080] The module 206 for determining the cause of abnormal power consumption is configured to determine whether the coal mill operating condition score is less than a preset score. When the coal mill operating condition score is less than the preset score, the data values ​​corresponding to the key parameters in the current coal mill process data group are standardized and then input into the decision tree model, so that the decision tree model outputs a conditional path. The cause of abnormal power consumption in the coal mill process is determined based on the conditional path, and an optimization strategy is determined based on the cause of abnormal power consumption in the coal mill process.

[0081] Figure 3A block diagram of an electronic device 300 according to some embodiments of the present invention is shown. The device 300 includes a processor 301, which performs various appropriate actions and processes based on computer program instructions loaded into random access memory (RAM) 303 according to computer program instructions stored in read-only memory (ROM) 302. Various programs and data required for the operation of the device 300 may also be stored in RAM 303. The processor 301, ROM 302, and RAM 303 are interconnected via a bus 304. An input / output (I / O) interface 305 is also connected to the bus 304.

[0082] The various processes and procedures described above, such as method 100, can be executed by processor 301. For example, in some embodiments, method 100 may be implemented as a software program tangibly contained in a machine-readable medium. In some embodiments, part or all of the software program may be loaded and / or installed on device 300 via ROM 302. When the software program is loaded into RAM 303 and executed by processor 301, one or more actions of method 100 described above may be performed.

[0083] The functions described above in this document can be performed at least in part by one or more hardware logic components. For example, exemplary types of hardware logic components that can be used, without limitation, include: field programmable gate arrays (FPGAs), application-specific integrated circuits (ASICs), application-specific standard products (ASSPs), systems-on-a-chip (SoCs), payload programmable logic devices (CPLDs), and so on.

[0084] The program code used to implement the methods of the present invention can be written in any combination of one or more programming languages. This program code can be provided to a processor or controller of a general-purpose computer, special-purpose computer, or other programmable data processing device, such that when executed by the processor or controller, the program code causes the functions / operations specified in the flowcharts and / or block diagrams to be implemented. The program code can be executed entirely on the machine, partially on the machine, as a standalone software package partially on the machine and partially on a remote machine, or entirely on a remote machine or server.

[0085] This invention can be a method, apparatus, system, and / or program product. The program product may include a machine-readable storage medium on which machine-readable program instructions for performing various aspects of the invention are loaded. The machine-readable program instructions described herein can be downloaded from the machine-readable storage medium to various computing / processing devices, or downloaded via a network, such as the Internet, a local area network, a wide area network, and / or a wireless network, to an external computer or external storage device. The network may include copper transmission cables, fiber optic transmissions, wireless transmissions, routers, firewalls, switches, gateway computers, and / or edge servers. A network adapter card or network interface in each computing / processing device receives the machine-readable program instructions from the network and forwards them for storage in the machine-readable storage medium of the respective computing / processing device.

[0086] Machine program instructions used to perform the operations of this invention may be assembly instructions, instruction set architecture (ISA) instructions, machine instructions, machine-dependent instructions, microcode, firmware instructions, state setting data, or source code or object code written in any combination of one or more programming languages, including object-oriented programming languages ​​such as Smalltalk, C++, etc., and conventional procedural programming languages ​​such as the "C" language or similar programming languages. The machine-readable program instructions may be executed entirely on the user's computer, partially on the user's computer, as a standalone software package, partially on the user's computer and partially on a remote computer, or entirely on a remote computer or server. In cases involving a remote computer, the remote computer may be connected to the user's computer via any type of network—including a local area network (LAN) or a wide area network (WAN)—or may be connected to an external computer (e.g., via the Internet using an Internet service provider). In some embodiments, electronic circuitry, such as programmable logic circuitry, field-programmable gate arrays (FPGAs), or programmable logic arrays (PLAs), is personalized by utilizing state information from the machine-readable program instructions. This electronic circuitry can execute the machine-readable program instructions to implement various aspects of the invention.

[0087] In the context of this invention, a machine-readable medium can be a tangible medium that may contain or store a program for use by or in conjunction with an instruction execution system, apparatus, or device. A machine-readable medium can be a machine-readable signal medium or a machine-readable storage medium. Machine-readable media can include, but are not limited to, electronic, magnetic, optical, electromagnetic, infrared, or semiconductor systems, apparatus, or devices, or any suitable combination of the foregoing. More specific examples of machine-readable storage media include electrical connections based on one or more wires, portable computer disks, hard disks, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM or flash memory), optical fibers, portable compact disk read-only memory (CD-ROM), optical storage devices, magnetic storage devices, or any suitable combination of the foregoing. Furthermore, although the operations are depicted in a specific order, this should be understood as requiring that such operations be performed in the specific order shown or in sequential order, or requiring that all illustrated operations be performed to achieve the desired result. In certain environments, multitasking and parallel processing may be advantageous. Similarly, while several specific implementation details are included in the above discussion, these should not be construed as limiting the scope of the invention. Certain features described in the context of individual embodiments may also be implemented in combination in a single embodiment. Conversely, various features described in the context of a single embodiment may also be implemented individually or in any suitable sub-combination in multiple embodiments.

[0088] Although the subject matter has been described using language specific to structural features and / or methodological logic, it should be understood that the subject matter defined in the appended claims is not necessarily limited to the specific features or actions described above. Rather, the specific features and actions described above are merely illustrative examples of implementing the claims.

Claims

1. A method for diagnosing and analyzing power consumption in coal milling processes based on a decision tree model, characterized in that, include: Historical coal milling process data is acquired, and several valid historical coal milling process data sets are obtained from the historical coal milling process data. Based on the historical hourly power consumption and historical hourly raw coal output in each valid historical coal milling process data set, the historical hourly process power consumption of each valid historical coal milling process data set is calculated. Based on the parameters in the valid historical coal milling process data set and the historical hourly process power consumption, the Pearson correlation coefficient corresponding to the parameters is calculated, and the key parameters of the decision tree model are determined based on the Pearson correlation coefficient. Obtain the current coal mill process data set, determine whether the current coal mill process data set is a valid data set, and if the current coal mill process data set is a valid data set, calculate the current coal mill hourly process power consumption based on the current coal mill hourly power consumption and the current hourly raw coal output in the current coal mill process data set, and obtain the coal mill operating condition score based on the current coal mill hourly process power consumption and the current coal mill process data set; Determine whether the coal mill operating condition score is less than a preset score. When the coal mill operating condition score is less than the preset score, standardize the data values ​​corresponding to the key parameters in the current coal mill process data group and input them into the decision tree model so that the decision tree model outputs the conditional path. The causes of abnormal power consumption in the coal milling process are determined based on the conditional path, and optimization strategies are determined based on the causes of abnormal power consumption in the coal milling process.

2. The method according to claim 1, characterized in that, Several sets of valid historical coal mill process data were obtained from the historical coal mill process data, including: Data sets with mill current greater than 25A, hourly running time equal to 60 minutes, and coal mill feed rate greater than 40t / h were obtained from historical coal mill process data and were used as valid historical coal mill process data sets.

3. The method according to claim 1, characterized in that, Key parameters of the decision tree model include: current of the small high-temperature fan, coal mill feed rate, mill grinding pressure, current of the main exhaust fan, frequency of the classifier, temperature difference between the inlet and outlet of the coal mill, mill current, mill differential pressure, classifier current, fineness of coal powder, and moisture content of raw coal.

4. The method according to claim 1, characterized in that, The coal mill operating condition score is obtained based on the current hourly process power consumption of the coal mill and the current process data set, including: The power consumption score is obtained based on the current hourly power consumption of the coal mill process, the model score is obtained based on the current coal mill process data group, and the coal mill operating condition score is calculated based on the model score, model score weight, power consumption score, and power consumption score weight.

5. The method according to claim 4, characterized in that, The energy consumption score is obtained based on the current hourly process energy consumption of the coal mill, including: Determine the current raw coal moisture content, ash content, and volatile matter content based on the current coal milling process data set; Based on the current raw coal moisture content, current raw coal ash content, and current raw coal volatile matter, find N valid historical coal mill process data sets with similar operating conditions from several valid historical coal mill process data sets; The average power consumption and standard deviation power consumption of the neighborhood are obtained based on N valid historical coal mill process data sets. Based on the current hourly process power consumption of coal mill, the average power consumption of the neighborhood, and the standard deviation of the power consumption of the neighborhood, the standardized value of power consumption is obtained. The standard normal cumulative distribution function value corresponding to the standardized value of power consumption is calculated, and the power consumption score is obtained based on the standard normal cumulative distribution function value.

6. The method according to claim 4, characterized in that, The model score obtained based on the current coal milling process data set includes: After standardizing the data values ​​corresponding to the key parameters in the current coal mill process data set, the data is input into the decision tree model so that the decision tree model outputs the probability of an excellent power consumption level, the probability of a good power consumption level, and the probability of a poor power consumption level. The initial model score is determined based on the probability of having an excellent power consumption level, a good power consumption level, and a poor power consumption level. The final model score is then calculated based on the initial model score.

7. The method according to claim 1, characterized in that, It also includes training decision tree models based on key parameters, specifically including: Model training samples were obtained based on several valid historical coal mill process data sets and key parameters. The decision tree model is obtained by training the model training samples and the decision tree algorithm.

8. A coal milling process power consumption diagnosis and analysis device based on a decision tree model, characterized in that, include: The key parameter determination module is configured to acquire historical coal mill process data, obtain several valid historical coal mill process data sets from the historical coal mill process data, calculate the historical coal mill hourly process power consumption of each valid historical coal mill process data set based on the historical coal mill hourly power consumption and historical hourly raw coal output in each valid historical coal mill process data set, calculate the Pearson correlation coefficient corresponding to the parameters based on the parameters in the valid historical coal mill process data set and the historical coal mill hourly process power consumption, and determine the key parameters of the decision tree model based on the Pearson correlation coefficient. The coal mill operating condition score acquisition module is configured to acquire the current coal mill process data group, determine whether the current coal mill process data group is a valid data group, and if the current coal mill process data group is a valid data group, calculate the current coal mill hourly process power consumption based on the current coal mill hourly power consumption and the current hourly raw coal output in the current coal mill process data group, and acquire the coal mill operating condition score based on the current coal mill hourly process power consumption and the current coal mill process data group; The module for determining the cause of abnormal power consumption is configured to determine whether the coal mill operating condition score is less than a preset score. When the coal mill operating condition score is less than the preset score, the data values ​​corresponding to the key parameters in the current coal mill process data group are standardized and then input into the decision tree model, so that the decision tree model outputs the conditional path. The causes of abnormal power consumption in the coal milling process are determined based on the conditional path, and optimization strategies are determined based on the causes of abnormal power consumption in the coal milling process.

9. A computer program product, characterized in that, Includes a computer program, which, when executed by a processor, implements the steps of the method according to any one of claims 1-7.

10. An electronic device, characterized in that, It includes one or more processors and memory associated with the one or more processors, the memory being used to store program instructions that, when read and executed by the one or more processors, perform the steps of the method according to any one of claims 1-7.

Images