阅读说明：本技术 一种电站煤粉锅炉烟气流量软测量方法 (Soft measurement method for flue gas flow of pulverized coal boiler of power station ) 是由 谭鹏 饶德备 李一波 李壮扬 吴小庚 李�权 仲柳 方庆艳 张�成 陈刚 于 2019-09-20 设计创作，主要内容包括：本发明提出一种电站煤粉锅炉烟气流量软测量方法,利用收集煤质数据(包括工业分析和元素分析),使用元素分析计算煤的理论空气量和理论烟气量,通过机器学习算法分别建立煤的工业分析和理论空气量及工业分析和理论烟气量的关系模型,模型可以使用工业分析直接估算煤种的理论烟气量和理论空气量,通过模型得出煤种的理论烟气量和理论空气量,锅炉磨煤机的给煤量、锅炉过量空气系数和烟气中CO占比,即可计算锅炉烟气量。(The invention provides a method for soft measurement of flue gas flow of a pulverized coal boiler in a power station, which comprises the steps of collecting coal quality data (comprising industrial analysis and elemental analysis), calculating theoretical air quantity and theoretical flue gas quantity of coal by using elemental analysis, respectively establishing a relation model of the industrial analysis and the theoretical air quantity of the coal and the industrial analysis and the theoretical flue gas quantity by using a machine learning algorithm, directly estimating the theoretical flue gas quantity and the theoretical air quantity of the coal by using the industrial analysis through the model, obtaining the theoretical flue gas quantity and the theoretical air quantity of the coal, the coal feeding quantity of a coal mill of the boiler, an excess air coefficient of the boiler and a CO ratio in the flue gas through the model, and calculating the flue gas quantity of the boiler.)

1. A soft measurement method for the flue gas flow of a pulverized coal boiler of a power station is characterized by comprising the following steps:

(1) collecting coal quality data:

collecting received base coal quality data comprising industrial analysis and elemental analysis;

(2) calculating theoretical air volume V by elemental analysis using coal quality data⁰And theoretical amount of flue gas

(3) Establishing a machine learning model of industrial analysis and theoretical air quantity:

using the theoretical air quantity calculated in the step (2) and taking industrial analysis data as input, wherein the industrial analysis data comprises moisture, ash, volatile matters, sulfur and low-order calorific value data, the theoretical air quantity is used as output to establish a new data set, carrying out normalization processing, and randomly dividing the data set into a training set and a testing set according to a preset proportion; establishing a machine learning model by using the training set, wherein the obtained model is H₁Verifying the established model by using a test set; the theoretical air volume estimated by the industrial analysis was obtained as:

wherein M is_ar、A_ar、V_ar、Q_net、S_arRespectively representing the receiving of base moisture, the receiving of base ash, the receiving of base volatile components, the receiving of low-level heating value and the receiving of base sulfur;

(4) establishing a machine learning model of industrial analysis and theoretical smoke volume:

using the theoretical flue gas volume calculated in the step (2) and taking industrial analysis data as input, wherein the industrial analysis data comprises moisture, ash, volatile matters, sulfur and low-order calorific value data, the theoretical flue gas volume is used as output to establish a new data set, carrying out normalization processing, and randomly dividing the data set into a training set and a testing set according to a preset proportion; establishing a machine learning model by using the training set, wherein the obtained model is H₂Verifying the established model by using a test set; the theoretical flue gas volume estimated by industrial analysis is as follows:

(5) calculating the actual flue gas amount of each coal type:

the actual smoke gas amount is calculated by the formulaWherein V⁰Andusing an Industrial analytical model H₁And H₂Estimated ofAndreplacing; coefficient of excess airO₂The percentage of oxygen in the flue gas in the dry flue gas is mentioned; discharge V of CO_CO＝CO(ppm)×V_yCO (ppm) is the CO concentration measured at the power station in ppm; the estimate of the actual flue gas quantity is therefore:

(6) calculating the total smoke gas amount of the boiler:

setting the coal feeding quantity of the coal mill X as m (X), and according to the coal quality information of the coal type of the coal mill X, estimating the actual smoke quantity by the step (5) asThe coal feeding amount and the actual flue gas amount of each coal mill are obtained according to the coal feeding amount and the actual flue gas amount; the total smoke amount q of the boiler_sThe sum of the smoke gas amount of each coal mill is as follows:

2. the method for soft measuring the flue gas flow of the pulverized coal fired power plant boiler as claimed in claim 1, characterized in that:

in step (3), modeling is performed using a machine learning algorithm using the theoretical air amount calculated in step (2), with the industrial analysis data as input and the theoretical air amount as output.

3. The method for soft measuring the flue gas flow of the pulverized coal fired power plant boiler as claimed in claim 1, characterized in that:

in the step (4), modeling is carried out by using a machine learning algorithm by using the theoretical flue gas volume calculated in the step (2), industrial analysis data is used as input, and the theoretical flue gas volume is used as output.

4. The method for soft measuring the flue gas flow of the pulverized coal fired power plant boiler as claimed in claim 2 or 3, characterized in that:

the modeling by using a machine learning algorithm is specifically modeling by using a neural network algorithm.

5. The method for soft measuring the flue gas flow of the pulverized coal fired power plant boiler as claimed in claim 2 or 3, characterized in that:

the modeling by using a machine learning algorithm is specifically modeling by using a support vector machine.

6. The method for soft measuring the flue gas flow of the pulverized coal fired power plant boiler as claimed in claim 1, characterized in that:

the steps (1) to (4) can be carried out off-line, and the steps (5) and (6) must be carried out on-line, wherein the excess air coefficient alpha and the CO concentration in the step (5), the coal quality and the coal feeding quantity of each coal mill in the step (6) are real-time data of the boiler SIS/DCS system and the coal blending and burning system.

7. The method for soft measuring the flue gas flow of the pulverized coal fired power plant boiler as claimed in claim 1, characterized in that:

in step (3) and step (4), the ratio of the total weight of the mixture is 7: a scale of 3 randomly divides the data set into a training set and a test set.

Technical Field

The invention belongs to the technical field of coal combustion, and particularly relates to a soft measurement method for the smoke flow of a pulverized coal boiler of a power station.

Background

The boiler flue gas volume plays an important role in environmental protection index calculation and boiler operation control, but because the situations of coal mixing, use of non-designed coal types and coal type non-element analysis frequently occur in the actual operation process of a power plant, the calculation of the flue gas volume is very difficult, and a power station often directly uses the theoretical flue gas volume and the theoretical air volume of the designed coal types for substitution, so that when the coal types used by the power station deviate from the designed coal types, the calculation of the flue gas volume has large errors. The total amount of emissions in the partial operation control including the SCR ammonia injection control and the calculation of the environmental protection index is adversely affected.

Boiler flue gas amount is needed in the SCR ammonia spraying process, but the control of the ammonia spraying amount is greatly influenced due to the fact that a large deviation exists in the calculation method of the boiler flue gas amount at the present stage. In addition, the boiler flue gas amount is needed when the emission total amount of each pollutant is calculated, and large calculation deviation of the emission total amount can be caused due to inaccuracy of the boiler flue gas amount. Therefore, the research and development of a more accurate boiler flue gas amount estimation method are of great significance.

Disclosure of Invention

Aiming at least one of the defects or the improvement requirements in the prior art, in particular to overcome the problem of flue gas volume calculation at the present stage, the invention provides a method for soft measurement of the flue gas volume of the pulverized coal boiler of the power station.

The method comprises the following specific steps:

(1) collecting coal quality data

The received base coal quality data including industrial and elemental analyses is collected.

(2) Calculating theoretical air volume V by elemental analysis using coal quality data⁰And theoretical amount of flue gas

1kg of theoretical air volume V required for complete combustion of the base fuel⁰Comprises the following steps:

theoretical amount of flue gas generated by complete combustion of 1kg received base fuelComprises the following steps:

(3) machine learning model for establishing industrial analysis and theoretical air quantity

Establishing a new data set by using the theoretical air amount calculated in the step (2) and using industrial analysis (including moisture, ash, volatile matter, sulfur and low calorific value) data as input and the theoretical air amount as outputAnd carrying out normalization processing, and randomly dividing the data set into a training set and a test set according to the ratio of 7: 3. Establishing a machine learning model by using the training set, wherein the obtained model is H₁And verifying the established model by using the test set. The theoretical air volume estimated by the industrial analysis was obtained as:

(4) machine learning model for establishing industrial analysis and theoretical smoke volume

Establishing a machine learning model of the industrial analysis and the theoretical smoke volume calculated in the step (2) similarly to the step (3), wherein the obtained model is H₂The theoretical air volume estimated by the industrial analysis is:

(5) calculating the actual smoke gas amount of single coal

The actual smoke gas amount is calculated by the formulaWherein V⁰Andusing Industrial analysis and model H₁And H₂Estimated ofAndinstead. The excess air coefficient alpha is calculated by the oxygen amount at the inlet of the boiler air preheater (or the oxygen amount at other representative measuring points)The discharge amount of CO is calculated by the measured concentration of CO in the power station_CO＝CO(ppm)×V_y. Estimation of the actual flue gas quantityThe calculation value is:

(6) calculating the total smoke gas quantity of the boiler

Setting the coal feeding quantity of the coal mill A as m (A), and according to the coal quality information of the coal type of the coal mill A, estimating the actual smoke quantity by the step (5) asBy analogy, the coal feeding amount and the actual flue gas amount of each coal mill of the BCDEF are obtained according to the format. The total smoke amount q of the boiler_sThe sum of the smoke gas amount of each coal mill is as follows:

the above-described preferred features may be combined with each other as long as they do not conflict with each other.

Generally, compared with the prior art, the above technical solution conceived by the present invention has the following beneficial effects:

1. compared with the prior art, the soft measurement method for the flue gas flow of the pulverized coal boiler of the power station can calculate the flue gas flow of the boiler by using the theoretical flue gas flow and the air quantity estimated by industrial analysis and combining with the real-time data of the SIS/DCS system of the power station, and has important significance for the operation of the boiler and the calculation of environmental protection indexes.

2. The method for soft measurement of the flue gas flow of the pulverized coal boiler of the power station can obtain the flue gas flow with higher precision without element analysis, and is convenient for practical use in the operation of the power station.

Drawings

FIG. 1 is a schematic flow chart of a method for soft measurement of flue gas flow of a pulverized coal boiler of a power station in an embodiment of the invention.

Detailed Description

In order to make the objects, technical solutions and advantages of the present invention more apparent, the present invention is described in further detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative of the invention and are not intended to limit the invention. In addition, the technical features involved in the embodiments of the present invention described below may be combined with each other as long as they do not conflict with each other. The present invention will be described in further detail with reference to specific embodiments.

As a preferred embodiment of the invention, as shown in FIG. 1, the invention provides a method for soft measurement of flue gas flow of a pulverized coal fired boiler of a power station. The method comprises the steps of collecting coal quality data (comprising industrial analysis and elemental analysis), calculating theoretical air quantity and theoretical flue gas quantity of coal by using elemental analysis, respectively establishing a relation model of the industrial analysis and the theoretical air quantity of the coal and the theoretical flue gas quantity by using a machine learning algorithm, directly estimating the theoretical flue gas quantity and the theoretical air quantity of coal by using the industrial analysis through the model, obtaining the theoretical flue gas quantity and the theoretical air quantity of the coal through the model, and calculating the boiler flue gas quantity by using the coal feeding quantity of a boiler coal mill, the excess air coefficient of the boiler and the CO ratio in the flue gas. The method comprises the following specific steps:

(1) collecting coal quality data:

the received base coal quality data including industrial and elemental analyses is collected.

(2) Calculating theoretical air volume V by elemental analysis using coal quality data⁰And theoretical amount of flue gas

1kg of theoretical air volume V required for complete combustion of the base fuel⁰Comprises the following steps:

wherein C is_ar、H_ar、S_ar、O_arRespectively, the content of the received radical C, H, S, O element;

theoretical amount of flue gas generated by complete combustion of 1kg received base fuelComprises the following steps:

wherein N is_ar、M_arThe content of the received basic N element and the received basic moisture are respectively, and the rest are the same as before.

(3) Establishing a machine learning model of industrial analysis and theoretical air quantity:

and (3) establishing a new data set by using the theoretical air quantity and the theoretical flue gas quantity calculated in the step (2) and taking industrial analysis (including moisture, ash, volatile matters, sulfur and low calorific value) data as input, taking the theoretical air quantity and the theoretical flue gas quantity as output, carrying out normalization processing, and randomly dividing the data set into a training set and a testing set according to the ratio of 7: 3. Establishing a machine learning model by using the training set, wherein the obtained model is H₁And verifying the established model by using the test set. The theoretical air volume estimated by the industrial analysis was obtained as:

wherein M is_ar、A_ar、V_ar、Q_net、S_arRespectively represents the receipt of base moisture, the receipt of base ash, the receipt of base volatile matter, the receipt of low-grade calorific value and the receipt of base sulfur.

(4) Establishing a machine learning model of industrial analysis and theoretical smoke volume:

similar to the step (3), the theoretical flue gas volume calculated in the step (2) is utilized, industrial analysis data are used as input, industrial analysis comprises moisture, ash content, volatile matter, sulfur content and low-order calorific value data, the theoretical flue gas volume is used as output to establish a new data set, normalization processing is carried out, and the data set is randomly divided into a training set and a testing set according to a preset proportion; establishing a machine learning model by using the training set, wherein the obtained model is H₂Verifying the established model by using a test set; the theoretical flue gas volume estimated by industrial analysis is as follows:

(5) calculating the actual flue gas amount of each coal type:

the actual smoke gas amount is calculated by the formulaWherein V⁰Andusing Industrial analysis and model H₁And H₂Estimated ofAndinstead. The excess air coefficient alpha is calculated by the oxygen amount at the inlet of the boiler air preheater (or the oxygen amount at other representative measuring points)O₂The percentage of oxygen in the flue gas accounts for the dry flue gas, and the emission of CO is calculated by the measured concentration of CO in the power station_CO＝CO(ppm)×V_y，

CO (ppm) is the CO concentration measured at the plant in ppm. The estimate of the actual flue gas quantity is therefore:

(6) calculating the total smoke gas amount of the boiler:

setting the coal feeding quantity of the coal mill A as m (A), and according to the coal quality information of the coal type of the coal mill A, estimating the actual smoke quantity by the step (5) asBy analogy, the coal feeding amount and the actual smoke amount of each coal mill of the BCDEF are determined according to the coal feeding amount and the actual smoke amountAnd (5) obtaining the format. The total smoke amount q of the boiler_sThe sum of the smoke gas amount of each coal mill is as follows:

the present invention will be described in detail below by taking the 11127 coal quality data collected as an example.

The American coal quality data 7428 group and the Chinese coal quality data 3699 group are collected, 11127 groups of coal quality data are calculated in total for modeling, and the theoretical air quantity and the theoretical flue gas quantity of different coal types are calculated by utilizing the elemental analysis in the coal quality data. Two new data sets are established by taking industrial analysis (including moisture, ash, volatile matters, sulfur and low calorific value) in the 11127 group coal quality data as input and respectively taking the calculated theoretical air quantity and the theoretical flue gas quantity as output. After the data set is normalized, the data are randomly scrambled according to the following steps of 7: the ratio of 3 is assigned to the training set and the test set. Optimizing optimal parameters c and g of a support vector machine by using a genetic algorithm, wherein the optimizing result is shown in table 1, establishing a support vector machine by using parameters obtained by training and optimizing, and testing the model precision by using a test set, wherein the training result and the test result are shown in table 2:

TABLE 1 parameter optimization results

TABLE 2 model Effect

As can be seen from the table 2, the precision of the 2 models is high, and the theoretical air quantity and the theoretical flue gas quantity can be directly estimated by using industrial analysis data through the models without element analysis, so that the actual use of the power station is greatly facilitated.

And (5) obtaining the real-time boiler flue gas amount with higher precision through the steps (5) and (6) by combining the oxygen amount, the CO concentration, the coal quality and the coal supply amount in the SIS system of the power station.

Compared with the prior art, the soft measurement method for the flue gas flow of the pulverized coal boiler of the power station can calculate the flue gas flow of the boiler by using the theoretical flue gas flow and the air quantity estimated by industrial analysis and combining with the real-time data of the SIS/DCS system of the power station, and has important significance for the operation of the boiler and the calculation of environmental protection indexes.

The method for soft measurement of the flue gas flow of the pulverized coal boiler of the power station can obtain the flue gas flow with higher precision without element analysis, and is convenient for practical use in the operation of the power station.

It will be understood by those skilled in the art that the foregoing is only a preferred embodiment of the present invention, and is not intended to limit the invention, and that any modification, equivalent replacement, or improvement made within the spirit and principle of the present invention should be included in the scope of the present invention.