CN113761794B - Boiler soot blowing optimization method for supplementing pollution factors based on time sequence prediction algorithm - Google Patents
Boiler soot blowing optimization method for supplementing pollution factors based on time sequence prediction algorithm Download PDFInfo
- Publication number
- CN113761794B CN113761794B CN202110942851.8A CN202110942851A CN113761794B CN 113761794 B CN113761794 B CN 113761794B CN 202110942851 A CN202110942851 A CN 202110942851A CN 113761794 B CN113761794 B CN 113761794B
- Authority
- CN
- China
- Prior art keywords
- pollution
- soot blowing
- time period
- time
- heating surface
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
- 239000004071 soot Substances 0.000 title claims abstract description 117
- 238000007664 blowing Methods 0.000 title claims abstract description 102
- 238000004422 calculation algorithm Methods 0.000 title claims abstract description 43
- 238000000034 method Methods 0.000 title claims abstract description 37
- 238000005457 optimization Methods 0.000 title claims abstract description 20
- 230000001502 supplementing effect Effects 0.000 title claims abstract description 11
- 238000010438 heat treatment Methods 0.000 claims abstract description 70
- 230000008859 change Effects 0.000 claims abstract description 15
- 238000004364 calculation method Methods 0.000 claims abstract description 14
- 230000002159 abnormal effect Effects 0.000 claims abstract description 12
- UGFAIRIUMAVXCW-UHFFFAOYSA-N Carbon monoxide Chemical compound [O+]#[C-] UGFAIRIUMAVXCW-UHFFFAOYSA-N 0.000 claims description 21
- 239000003546 flue gas Substances 0.000 claims description 21
- 238000012546 transfer Methods 0.000 claims description 19
- 230000006641 stabilisation Effects 0.000 claims description 11
- 238000011105 stabilization Methods 0.000 claims description 11
- 238000004140 cleaning Methods 0.000 claims description 10
- 238000012216 screening Methods 0.000 claims description 9
- 230000000087 stabilizing effect Effects 0.000 claims description 9
- 238000007477 logistic regression Methods 0.000 claims description 8
- 230000007704 transition Effects 0.000 claims description 6
- 230000000737 periodic effect Effects 0.000 claims description 5
- 230000009471 action Effects 0.000 claims description 4
- 238000012935 Averaging Methods 0.000 claims description 3
- 230000000295 complement effect Effects 0.000 claims description 3
- 238000009499 grossing Methods 0.000 claims description 3
- OIGNJSKKLXVSLS-VWUMJDOOSA-N prednisolone Chemical compound O=C1C=C[C@]2(C)[C@H]3[C@@H](O)C[C@](C)([C@@](CC4)(O)C(=O)CO)[C@@H]4[C@@H]3CCC2=C1 OIGNJSKKLXVSLS-VWUMJDOOSA-N 0.000 claims description 3
- 230000009467 reduction Effects 0.000 claims description 3
- 238000007619 statistical method Methods 0.000 claims description 3
- XLYOFNOQVPJJNP-UHFFFAOYSA-N water Substances O XLYOFNOQVPJJNP-UHFFFAOYSA-N 0.000 claims description 2
- 238000009825 accumulation Methods 0.000 abstract description 6
- 230000009286 beneficial effect Effects 0.000 abstract description 2
- 230000000694 effects Effects 0.000 abstract description 2
- 230000008569 process Effects 0.000 description 7
- 239000002893 slag Substances 0.000 description 4
- 239000000779 smoke Substances 0.000 description 4
- 238000012549 training Methods 0.000 description 4
- 238000011109 contamination Methods 0.000 description 3
- 238000010586 diagram Methods 0.000 description 3
- 239000007789 gas Substances 0.000 description 2
- 239000002184 metal Substances 0.000 description 2
- 238000012986 modification Methods 0.000 description 2
- 230000004048 modification Effects 0.000 description 2
- 238000012544 monitoring process Methods 0.000 description 2
- 230000001932 seasonal effect Effects 0.000 description 2
- 230000003442 weekly effect Effects 0.000 description 2
- 241001123248 Arma Species 0.000 description 1
- QIVBCDIJIAJPQS-VIFPVBQESA-N L-tryptophane Chemical compound C1=CC=C2C(C[C@H](N)C(O)=O)=CNC2=C1 QIVBCDIJIAJPQS-VIFPVBQESA-N 0.000 description 1
- 238000005299 abrasion Methods 0.000 description 1
- 230000006978 adaptation Effects 0.000 description 1
- 238000004458 analytical method Methods 0.000 description 1
- 238000013528 artificial neural network Methods 0.000 description 1
- 239000003245 coal Substances 0.000 description 1
- 238000002485 combustion reaction Methods 0.000 description 1
- 238000005260 corrosion Methods 0.000 description 1
- 230000007797 corrosion Effects 0.000 description 1
- 230000007547 defect Effects 0.000 description 1
- 230000008021 deposition Effects 0.000 description 1
- 230000003203 everyday effect Effects 0.000 description 1
- 229910052500 inorganic mineral Inorganic materials 0.000 description 1
- 238000005259 measurement Methods 0.000 description 1
- 239000011707 mineral Substances 0.000 description 1
- 239000000843 powder Substances 0.000 description 1
- 230000005855 radiation Effects 0.000 description 1
- 230000002441 reversible effect Effects 0.000 description 1
- 239000002699 waste material Substances 0.000 description 1
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F30/00—Computer-aided design [CAD]
- G06F30/20—Design optimisation, verification or simulation
- G06F30/27—Design optimisation, verification or simulation using machine learning, e.g. artificial intelligence, neural networks, support vector machines [SVM] or training a model
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F2119/00—Details relating to the type or aim of the analysis or the optimisation
- G06F2119/08—Thermal analysis or thermal optimisation
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F2119/00—Details relating to the type or aim of the analysis or the optimisation
- G06F2119/12—Timing analysis or timing optimisation
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- Evolutionary Computation (AREA)
- Theoretical Computer Science (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Medical Informatics (AREA)
- Software Systems (AREA)
- Artificial Intelligence (AREA)
- Computer Hardware Design (AREA)
- Geometry (AREA)
- General Engineering & Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Incineration Of Waste (AREA)
Abstract
The invention relates to a boiler soot blowing optimization method for supplementing pollution factors based on a time sequence prediction algorithm, which comprises the following steps: calculating a heating surface pollution factor; removing abnormal data; grabbing in a steady-state working condition time period; and complementing the pollution factor in the unsteady state time period by using a time sequence prediction algorithm. The beneficial effects of the invention are as follows: according to the invention, the pollution factor in the steady-state working condition time period is calculated by grabbing the steady-state working condition, and the pollution factor in the unsteady-state working condition is complemented according to the time sequence prediction algorithm, so that soot blowing is guided. The pollution factor obtained directly according to heat balance calculation is difficult to accurately reflect the actual pollution condition of the heating surface in real time and is easy to fluctuate greatly due to the change of the heat accumulation quantity of the heating surface under the unsteady working condition, so that a great number of false soot blowing operations are avoided. The method is simple and reasonable, has good effect in practical application, does not need to add extra equipment and measuring points, and has remarkable technical and economic benefits.
Description
Technical Field
The invention belongs to the technical field of coal-fired power plants, and particularly relates to a soot blowing optimization method for a coal-fired boiler, which is based on a time sequence prediction algorithm and is used for supplementing pollution factors.
Background
In the coal-fired power station boiler, mineral components such as nonflammable ash and slag can be deposited on a heating surface in the coal-powder combustion process, and the coal-fired power station boiler mainly comprises slag bonding in a high-temperature area such as a water-cooled wall and ash deposition on the heating surface of a tail flue. The ash and dirt thermal resistance deposited on the heating surface is greatly higher than that of the metal pipe wall, and if the metal pipe wall is not cleaned in time, a series of problems such as reduced heat transfer performance of the heating surface, increased smoke discharge temperature, increased flue resistance, high-temperature corrosion, increased abrasion of the heating surface and the like can be caused. The existing large-scale coal-fired power plant boiler avoids serious ash accumulation and slag bonding of a heating surface by a method of installing soot blowers, and the number of the soot blowers of the coal-fired power plant boiler with the level of more than 300MW is generally between 60 and 180, and the soot blowers are arranged at the upper part of a hearth, a horizontal flue, a vertical flue and the like.
At present, most coal-fired power plants adopt a quantitative soot blowing scheme on time, namely, soot blowing operation is carried out for the same time period according to fixed time every day to form a fixed soot blowing period, but the establishment of the fixed soot blowing period only considers the scheduling of operators and lacks data support, the condition of the heated area of the soot cannot be intuitively judged, the heated surface with serious soot accumulation can not be purged in time, the heat transfer performance of the heated surface is reduced, and the main re-steam temperature and the exhaust gas temperature are higher; frequent blowing of the heated surface with slight soot accumulation is also possible, which causes waste of soot blowing steam and shortens the service life of the pipe wall.
Through searching the prior art, part of researchers calculate the pollution factors of each heating surface through heat balance calculation or a neural network, a two-dimensional optimal searching algorithm, so as to judge the soot blowing interval and the frequency. A Soot blowing optimization system or platform such as a Optimax system boiler cleaning module developed by a Switzerland ABB company is a Soot-blower Advisor expert system developed by a foreign New York state electric gas company and a general physical company, and a Smart ProcessTM system or platform of a West house company calculates the actual heat transfer quantity of each Soot blower area according to data such as the inlet and outlet smoke temperature and the working medium temperature of part of heating surfaces, so as to calculate pollution factors, and guide Soot blowing. The China patent of application number 2016107782348 discloses an intelligent soot blowing method for a convection heating surface of a coal-fired unit based on two-dimensional optimization, pollution factors are determined by calculating ideal heat exchange coefficients and actual heat exchange coefficients of the convection heating surface, and cleaning conditions of the convection heating surface are monitored on line. The China patent of application number 2009100332364 discloses a boiler ash and dirt monitoring and soot blowing optimizing method based on coal quality on-line measurement, and the reverse smoke flow carries out on-line monitoring, analysis and calculation on the accumulated ash and slag of each main convection heating surface and the smoke temperature of the outlet of a hearth.
The calculation process of the soot blowing optimization system calculates the actual heat transfer coefficient according to the parameters such as the temperature and flow of the actual flue gas side and the working medium side of each heating surface, and simultaneously can calculate the pollution factor of the heating surface according to the ideal heat transfer coefficient, thereby judging whether the heating surface needs soot blowing or not. However, the above method has the disadvantages that: under the unsteady working condition, for example, the boiler load greatly fluctuates or the heating surface just after the soot blowing is finished in a short time, the pollution factor is difficult to accurately reflect the actual pollution condition of the heating surface in real time due to the change of the heat accumulation amount of the heating surface, so the pollution factor greatly fluctuates under the unsteady working condition, and if the unsteady working condition pollution factor obtained through actual calculation is directly used for judging whether the soot blowing is needed, a large amount of soot blowing errors can be caused.
Disclosure of Invention
The invention aims to overcome the defects in the prior art and provides a soot blowing optimization method for a coal-fired boiler, which is based on a time sequence prediction algorithm and complements pollution factors.
The soot blowing optimization method for the coal-fired boiler for supplementing pollution factors based on a time sequence prediction algorithm comprises the following steps of:
step1, calculating a heating surface pollution factor;
Step 2, cleaning data according to the upper limit and the lower limit of pollution factors, boiler load, working medium measuring point data quality and other factors, and removing abnormal data;
Step 3, grabbing in a steady-state working condition time period;
step 3.1, grabbing a period of time which is not affected by soot blowing: judging a soot blowing time period by using a soot blower action current measuring point, a soot blower starting instruction measuring point and a soot blower exiting instruction measuring point, and grabbing the soot blowing time period; introducing a stable time t after soot blowing, and determining a soot blowing time period and an unsteady state time period influenced by soot blowing, wherein the remaining time period is a time period not influenced by soot blowing;
step 3.2, grabbing in a boiler load stabilization time period;
Step 3.3, grabbing in a steady-state working condition time period: taking the intersection part of the time period which is not affected by soot blowing and is obtained in the step 3.1 and the boiler load stabilizing time period obtained in the step 3.2 as a steady-state working condition time period, wherein the non-intersection part is a non-steady-state working condition time period;
step 4, complementing the pollution factor in the unsteady state time period by using a time sequence prediction algorithm;
and step 5, judging whether to perform soot blowing or not according to the change trend of the pollution factors.
Preferably, the step 1 specifically includes the following steps:
step 1.1, according to the basic principle of heat balance of the whole and each local heating surface of the boiler, calculating the inlet flue gas temperature of each heating surface section by utilizing the existing online working medium side parameters and flue gas side parameters according to the opposite direction of flue gas flow;
Step 1.2, calculating the actual heat transfer coefficient and the ideal heat transfer coefficient of each heating surface according to the parameters such as the inlet flue gas temperature, the outlet flue gas temperature, the inlet and outlet temperatures of the working medium, the flue gas flow, the working medium flow, the structural size of the heating surface and the like;
step 1.3, calculating the pollution factor of the heating surface according to the formula (1):
In the above formula, K sj is the actual heat transfer coefficient of the heating surface, the unit is W/(m 2*K);Klx is the ideal heat transfer coefficient of the heating surface, the unit is W/(m 2 ×k), and F is the pollution factor.
Preferably, the abnormal data removed in the step 2 includes abnormal data such as shutdown abnormal data and damage to the measuring points.
Preferably, the steady-state working condition in the step 3 is that the heating surface is blowing or blowing is just finished, and the boiler load is stable.
Preferably, step 3.2 comprises the steps of:
Step 3.2.1, screening all boiler load data in a certain time point and a later tau time period to obtain a highest load value D max and a lowest load value D min in the time period tau, and calculating load fluctuation delta D in the time period tau:
ΔD=Dmax-Dmin (2)
In the above formula, the units of D max、Dmin and DeltaD are t/h or MW;
when the load fluctuation delta D in the time period tau meets the following formula (3), the time period tau is a working condition of stable boiler load;
ΔD≤x*Ded (3)
in the above formula, x is the ratio of the preset normal fluctuation range of the load to the rated load; d ed is the rated load of the boiler, and the unit is t/h or MW;
and 3.2.2, screening the boiler load data in all time points and the time period tau after the time points by a moving time window method according to the step 3.2.1, and obtaining the boiler load stabilizing time period.
Preferably, in step 3.2.1, the boiler is in a load stabilizing condition within the fluctuation ratio range.
Preferably, the step 4 specifically includes the following steps:
Step 4.1, calculating pollution factors in a steady-state working condition time period by adopting the calculation method of the heating surface pollution factors in the step 1;
step 4.2, complementing and calculating pollution factors in the unsteady state working condition time period by a time prediction algorithm; the time series prediction algorithm adopts a logistic regression model of prophet algorithm:
F(t)=g(t)+s(t)+h(t) (4)
In the above formula, t is time; f (t) is a function of the pollution factor over time; g (t) represents a growth function for fitting the aperiodic variations; s (t) is used to represent periodic variations, such as weekly, yearly, seasonal, etc.; h (t) represents the change caused by special reasons such as holidays, holidays and the like; i.e. a superposition of time series divided into 3 parts (soot blowing time period, non-steady state time period affected by soot blowing, time period unaffected by soot blowing); for the time series prediction of pollution factors, there is no obvious periodic variation and no obvious holiday variation, so the influence of s (t) and h (t) can be ignored;
the growth of g (t) adopts a logistic regression function model, as shown in the formulas (5) and (6):
In the formulas (5) and (6), t is time; c (t) is a saturation value (load carrying capacity); k is the rate of rise; m is a bias parameter; the function a (t) represents an indication function of the time node as a function of the growth rate k; delta represents the amount of change in the rate of increase at time stamp s j; s j is the moment of the j-th transition point at which the rate of increase changes; gamma is the smoothing offset at s j; a j (t) is an indication function of a time node of the j-th transition point changing with the growth rate k at the moment t;
and 4.3, presetting upper and lower limits of pollution factors, a time interval of data, a prediction period and time sequence data with a time stamp according to prophet algorithm models, and compensating the pollution factors in a non-steady-state working condition time period.
Preferably, the step 5 specifically includes the following steps:
Step 5.1, carrying out statistical analysis on historical pollution factors of a heating surface, setting the condition that the pollution factors are smaller than or equal to fixed values as a cleaning working condition of the heating surface, and defining the fixed values of the pollution factors as minimum pollution factors F min;
step 5.2, counting the change amplitude of the pollution factor of the heating surface before and after the multi-time soot blowing, and averaging to obtain the decrease amplitude delta F of the pollution factor of the heating surface after the soot blowing:
ΔFi=Fb,i-Fa,i (7)
In the formulas (7) to (8), deltaF i is the variation amplitude of the heating surface pollution factors before and after the ith soot blowing; f b,i is a pollution factor of a heating surface before the ith soot blowing; f a,i is a pollution factor of a heating surface after the ith soot blowing; n is the number of statistics, preferably at least 1 year of data;
Step 5.3, calculating a critical pollution factor F lj and generating a critical pollution factor table;
Flj=Fmin+ΔF (9)
in the above formula, F min is the minimum pollution factor F min, and DeltaF is the reduction amplitude of the pollution factor of the heated surface after soot blowing;
Step 5.4, judging whether the actual pollution factor is larger than the critical pollution factor, if the actual pollution factor is larger than the critical pollution factor, further judging whether parameters such as boiler load, main re-steam temperature and the like meet the necessary condition of soot blowing, and performing soot blowing when the necessary condition of soot blowing is met, otherwise, not performing soot blowing; if the actual pollution factor is smaller than or equal to the critical pollution factor, the critical pollution factor is continuously calculated, and a critical pollution factor table is generated.
Preferably, the minimum contamination factor F min in step 5.1 is the average value of all the historical contamination factors minus 2 times the standard deviation of the historical contamination factors:
Fmin=μ-2σ (10)
In the above formula, μ is the average value of the historical pollution factors, and σ is the standard deviation of the historical pollution factors.
Preferably, in step 5.4, working conditions are subdivided according to parameters such as unit load, water supply flow and the like, and a critical pollution factor table under each subdivision working condition is generated.
Preferably, the critical pollution factor can also be determined by judging the energy balance of soot blowing gain and soot blowing steam consumption according to the completed historical pollution factor data obtained in the steps 1 to 4 in the steps 5.1 to 5.3.
The beneficial effects of the invention are as follows: according to the invention, the pollution factor in the steady-state working condition time period is calculated by grabbing the steady-state working condition, and the pollution factor in the unsteady-state working condition is complemented according to the time sequence prediction algorithm, so that soot blowing is guided. The pollution factor obtained directly according to heat balance calculation is difficult to accurately reflect the actual pollution condition of the heating surface in real time and is easy to fluctuate greatly due to the change of the heat accumulation quantity of the heating surface under the unsteady working condition, so that a great number of false soot blowing operations are avoided. The method is simple and reasonable, has good effect in practical application, does not need to add extra equipment and measuring points, and has remarkable technical and economic benefits.
Drawings
FIG. 1 is a flow chart of model training in a second embodiment of the invention;
Fig. 2 is a flowchart of the model operation in the second embodiment of the present invention.
FIG. 3 is a schematic diagram of a boiler load stabilization period in a second embodiment of the present invention.
Detailed Description
The invention is further described below with reference to examples. The following examples are presented only to aid in the understanding of the invention. It should be noted that it will be apparent to those skilled in the art that modifications can be made to the present invention without departing from the principles of the invention, and such modifications and adaptations are intended to be within the scope of the invention as defined in the following claims.
Example 1
The embodiment of the application provides a soot blowing optimization method for a coal-fired boiler, which is based on a time sequence prediction algorithm to complement pollution factors, and comprises the following steps:
step1, calculating a heating surface pollution factor;
Step 2, cleaning data according to the upper limit and the lower limit of pollution factors, boiler load, working medium measuring point data quality and other factors, and removing abnormal data;
Step 3, grabbing in a steady-state working condition time period;
step 3.1, grabbing a period of time which is not affected by soot blowing: judging a soot blowing time period by using a soot blower action current measuring point, a soot blower starting instruction measuring point and a soot blower exiting instruction measuring point, and grabbing the soot blowing time period; introducing a stable time t after soot blowing, and determining a soot blowing time period and an unsteady state time period influenced by soot blowing, wherein the remaining time period is a time period not influenced by soot blowing;
step 3.2, grabbing in a boiler load stabilization time period;
Step 3.2.1, screening all boiler load data in a certain time point and a later tau time period to obtain a highest load value D max and a lowest load value D min in the time period tau, and calculating load fluctuation delta D in the time period tau:
ΔD=Dmax-Dmin (2)
In the above formula, the units of D max、Dmin and DeltaD are t/h or MW;
when the load fluctuation delta D in the time period tau meets the following formula (3), the time period tau is a working condition of stable boiler load;
ΔD≤x*Ded (3)
in the above formula, x is the ratio of the preset normal fluctuation range of the load to the rated load; d ed is the rated load of the boiler, and the unit is t/h or MW;
step 3.2.2, screening the boiler load data in all time points and the time period tau after the time points by a moving time window method according to the step 3.2.1 to obtain a boiler load stabilizing time period;
Step 3.3, grabbing in a steady-state working condition time period: taking the intersection part of the time period which is not affected by soot blowing and is obtained in the step 3.1 and the boiler load stabilizing time period obtained in the step 3.2 as a steady-state working condition time period, wherein the non-intersection part is a non-steady-state working condition time period;
step 4, complementing the pollution factor in the unsteady state time period by using a time sequence prediction algorithm;
Step 4.1, calculating pollution factors in a steady-state working condition time period by adopting the calculation method of the heating surface pollution factors in the step 1;
step 4.2, complementing and calculating pollution factors in the unsteady state working condition time period by a time prediction algorithm; the time series prediction algorithm adopts a logistic regression model of prophet algorithm:
F(t)=g(t)+s(t)+h(t) (4)
In the above formula, t is time; f (t) is a function of the pollution factor over time; g (t) represents a growth function for fitting the aperiodic variations; s (t) is used to represent periodic variations, such as weekly, yearly, seasonal, etc.; h (t) represents the change caused by special reasons such as holidays, holidays and the like; i.e. a superposition of time series divided into 3 parts (soot blowing time period, non-steady state time period affected by soot blowing, time period unaffected by soot blowing); for the time series prediction of pollution factors, there is no obvious periodic variation and no obvious holiday variation, so the influence of s (t) and h (t) can be ignored;
the growth of g (t) adopts a logistic regression function model, as shown in the formulas (5) and (6):
In the formulas (5) and (6), t is time; c (t) is a saturation value (load carrying capacity); k is the rate of rise; m is a bias parameter; the function a (t) represents an indication function of the time node as a function of the growth rate k; delta represents the amount of change in the rate of increase at time stamp s j; s j is the moment of the j-th transition point at which the rate of increase changes; gamma is the smoothing offset at s j; a j (t) is an indication function of a time node of the j-th transition point changing with the growth rate k at the moment t;
Step 4.3, presetting upper limit and lower limit of pollution factors, time interval of data, prediction period and time sequence data with time stamp according to prophet algorithm model, and compensating pollution factors in unsteady state working condition time period;
and step 5, judging whether to perform soot blowing or not according to the change trend of the pollution factors.
Example two
On the basis of the first embodiment, the second embodiment of the application provides an application of the soot blowing optimization method of the coal-fired boiler based on the time sequence prediction algorithm to complement pollution factors in the soot blowing optimization process of a certain 1000MW coal-fired boiler:
1. Model training
The process of model training is shown in fig. 1.
Step1, calculating a heating surface pollution factor;
The boiler is provided with a low-temperature superheater, a low-temperature reheater inlet section, a flue gas temperature online measuring point of all heating surface inlets and outlets of a downstream flue and a working medium temperature measuring point of all heating surface inlets and outlets, so that the flue gas temperatures of a high-temperature reheater inlet, a high-temperature superheater inlet and a screen superheater inlet are calculated section by section according to the opposite directions of flue gas flow by utilizing the existing online working medium side parameters and the existing flue gas side parameters and considering the influence of radiation heat transfer and convection heat transfer. And calculating an actual heat transfer coefficient and an ideal heat transfer coefficient according to parameters such as the temperature of a flue gas inlet and outlet of each heating surface, the temperature of a working medium inlet and outlet, the flow of flue gas and working medium, the structural size of the heating surface and the like, and calculating a pollution factor of the heating surface according to a formula (1), thereby obtaining a pollution factor calculation value.
In the above formula, K sj is the actual heat transfer coefficient of the heating surface, the unit is W/(m 2*K);Klx is the ideal heat transfer coefficient of the heating surface, the unit is W/(m 2 ×k), and F is the pollution factor.
Step2, data cleaning;
and (3) cleaning data according to factors such as load, upper and lower limits of pollution factors, data quality of relevant flue gas and working medium measuring points, and the like, and eliminating abnormal data such as shutdown, measuring point damage and the like. The following data were mainly washed out:
1) The load is less than 0MW or greater than 1050MW;
2) The pollution factor calculated value is smaller than or equal to 0 or larger than or equal to 1;
3) The flue gas temperature, the working medium temperature and the working medium flow measuring point have no reading or abnormal reading.
Step 3, grabbing in a steady-state working condition time period;
And 3.1, grabbing in a time period which is not affected by soot blowing. And judging a soot blowing time period by using measuring points such as a soot blower action current, a soot blower starting instruction, a soot blower exiting instruction and the like, setting the stabilization time after soot blowing to be 30min, and setting the remaining time period to be a time period which is not influenced by soot blowing.
Step 3.2, grabbing in a boiler load stabilization time period;
1) Screening all boiler load data at a certain time point and within a period tau thereafter to obtain a highest load value D max and a lowest load value D min within the period, so as to calculate load fluctuation delta D within the period, as shown in a formula (2):
ΔD=Dmax-Dmin (2)
In the above formula, the units of D max、Dmin and DeltaD are t/h or MW;
And delta D is less than or equal to x and D ed is taken as a necessary condition for stabilizing the load of the boiler. When the value of tau is smaller and the value of x is larger, the number of segments of the load stabilization time period obtained by corresponding grabbing is larger, but the actual load stability is poorer; on the contrary, when τ is larger and x is smaller, the number of segments is smaller in relation to the load stabilization time period in which the condition is satisfied, but the obtained load stabilization time period is also better in actual load stability. In this embodiment, τ is selected in the range of 10-60 min and x is selected in the range of 3% -10%.
2) And screening the boiler load data in all time points and the post tau time period by using a moving time window method according to the method to obtain the working condition time period with stable boiler load.
The schematic diagram of the boiler load stabilization time period obtained by screening is shown in fig. 3, and dark curve part, standard square frame part and standard oval frame part in the diagram are load stabilization working condition time periods obtained by sequentially and newly increasing different tau and x values (according to the directions of reducing tau and increasing x).
Step 3.3, grabbing in a steady-state working condition time period;
the intersection part of the time period which is not affected by soot blowing and the working condition time period with stable boiler load is the steady-state working condition time period, and the other time periods are unsteady-state working condition time periods.
Step 4, complementing the pollution factor in the unsteady state time period by using a time sequence prediction algorithm;
and 4.1, calculating a pollution factor in a time period of a steady-state working condition by adopting the method of the step 1.
Step 4.2, the pollution factor in the unsteady state working condition time period is obtained by complementation calculation of a time prediction algorithm; the time series prediction algorithm may employ, but is not limited to, a logistic regression model of prophet algorithm, and other time series algorithms such as a piecewise function model of prophet algorithm, an ARMA algorithm model, etc. may be used for the calculation of this step. The logistic regression model adopting prophet algorithm needs to preset the upper and lower limits of pollution factors, the time interval of data, the prediction period and time sequence data with time stamps, and can calculate the pollution factors in the unsteady working condition time period.
And 4.3, obtaining pollution factors in all time periods (including steady-state and unsteady-state working conditions).
Step 5, calculating critical pollution factors and generating a critical pollution factor table;
And 5.1, carrying out statistical analysis on historical pollution factors of a heating surface, wherein a calculation method for setting the pollution factors as minimum pollution factors F min.Fmin when the pollution factors are smaller than or equal to a certain value are set as cleaning working conditions of the heating surface, and the calculation method adopts the following formula (8), namely the average value of all the historical pollution factors is reduced by 2 times of standard deviation.
And 5.2, calculating the change amplitude of the pollution factor of the heating surface before and after each normal soot blowing in history, and averaging to obtain the pollution factor reduction amplitude delta F of the heating surface after soot blowing.
And 5.3, calculating a critical pollution factor F lj according to the following formula (7), and generating a critical pollution factor table.
ΔFi=Fb,i-Fa,i (7)
2. Model operation
The process of model operation is shown in fig. 2.
The steps 1, 2, 3 and 4 of the model running process are the same as the model training process.
The step 5 of the model operation process is as follows: judging whether the actual pollution factor is larger than the critical pollution factor, if the actual pollution factor is larger than the critical pollution factor, judging that the heating surface has soot blowing necessity, and if not, continuing to calculate.
Claims (8)
1. The boiler soot blowing optimization method for supplementing pollution factors based on the time sequence prediction algorithm is characterized by comprising the following steps of:
step1, calculating a heating surface pollution factor;
Step 2, cleaning data according to the upper limit and the lower limit of the pollution factor, the boiler load and the data quality of the working medium measuring point, and eliminating abnormal data;
Step 3, grabbing in a steady-state working condition time period;
step 3.1, grabbing a period of time which is not affected by soot blowing: judging a soot blowing time period by using a soot blower action current measuring point, a soot blower starting instruction measuring point and a soot blower exiting instruction measuring point, and grabbing the soot blowing time period; introducing a stable time t after soot blowing, and determining a soot blowing time period and an unsteady state time period influenced by soot blowing, wherein the remaining time period is a time period not influenced by soot blowing;
step 3.2, grabbing in a boiler load stabilization time period;
Step 3.3, grabbing in a steady-state working condition time period: taking the intersection part of the time period which is not affected by soot blowing and is obtained in the step 3.1 and the boiler load stabilizing time period obtained in the step 3.2 as a steady-state working condition time period, wherein the non-intersection part is a non-steady-state working condition time period;
step 4, complementing the pollution factor in the unsteady state time period by using a time sequence prediction algorithm;
The step 4 specifically comprises the following steps:
Step 4.1, calculating pollution factors in a steady-state working condition time period by adopting the calculation method of the heating surface pollution factors in the step 1;
step 4.2, complementing and calculating pollution factors in the unsteady state working condition time period by a time prediction algorithm; the time series prediction algorithm adopts a logistic regression model of prophet algorithm:
F(t)=g(t)+s(t)+h(t) (4)
In the above formula, t is time; f (t) is a function of the pollution factor over time; g (t) represents a growth function for fitting the aperiodic variations; s (t) is used to represent a periodic variation; h (t) represents a change caused by a specific cause;
the growth of g (t) adopts a logistic regression function model, as shown in the formulas (5) and (6):
in the formulas (5) and (6), t is time; c (t) is a saturation value; k is the rate of rise; m is a bias parameter; the function a (t) represents an indication function of the time node as a function of the growth rate k; delta represents the amount of change in the rate of increase at time stamp s j; s j is the time of the j-th transition point; gamma is the smoothing offset at s j; a j (t) is an indication function of a time node of the j-th transition point changing with the growth rate k at the moment t;
Step 4.3, presetting upper limit and lower limit of pollution factors, time interval of data, prediction period and time sequence data with time stamp according to prophet algorithm model, and compensating pollution factors in unsteady state working condition time period;
and step 5, judging whether to perform soot blowing or not according to the change trend of the pollution factors.
2. The boiler soot blowing optimization method based on the time sequence prediction algorithm for supplementing pollution factors according to claim 1, wherein the step 1 specifically comprises the following steps:
Step 1.1, calculating the inlet flue gas temperature of each heating surface section by section according to the opposite direction of flue gas flow by utilizing the existing online working medium side parameters and flue gas side parameters;
step 1.2, calculating the actual heat transfer coefficient and the ideal heat transfer coefficient of each heating surface according to the inlet flue gas temperature, the outlet flue gas temperature, the inlet and outlet temperatures of the working medium, the flue gas flow, the working medium flow and the structural size of the heating surface;
Step 1.3, calculating the pollution factor of the heating surface according to the formula (1):
In the above formula, K sj is the actual heat transfer coefficient of the heating surface, the unit is W/(m 2*K);Klx is the ideal heat transfer coefficient of the heating surface, the unit is W/(m 2 ×k), and F is the pollution factor.
3. The boiler soot blowing optimization method based on the time sequence prediction algorithm to complement pollution factors according to claim 1, wherein the method comprises the following steps: and (3) the abnormal data removed in the step (2) comprise shutdown abnormal data and measuring point damage abnormal data.
4. The boiler soot blowing optimization method based on the time series prediction algorithm for supplementing pollution factors according to claim 1, wherein the step 3.2 comprises the following steps:
Step 3.2.1, screening all boiler load data in a certain time point and a later tau time period to obtain a highest load value D max and a lowest load value D min in the time period tau, and calculating load fluctuation delta D in the time period tau:
ΔD=Dmax-Dmin (2)
In the above formula, the units of D max、Dmin and DeltaD are t/h or MW;
when the load fluctuation delta D in the time period tau meets the following formula (3), the time period tau is a working condition of stable boiler load;
ΔD≤x*Ded (3)
in the above formula, x is the ratio of the preset normal fluctuation range of the load to the rated load; d ed is the rated load of the boiler, and the unit is t/h or MW;
and 3.2.2, screening the boiler load data in all time points and the time period tau after the time points by a moving time window method according to the step 3.2.1, and obtaining the boiler load stabilizing time period.
5. The boiler soot blowing optimization method based on the time sequence prediction algorithm for supplementing pollution factors according to claim 4, wherein the method comprises the following steps of: in the step 3.2.1, the boiler is in a load stabilizing working condition in a fluctuation proportion range.
6. The boiler soot blowing optimization method based on the time sequence prediction algorithm for supplementing pollution factors according to claim 1, wherein the step 5 specifically comprises the following steps:
Step 5.1, carrying out statistical analysis on historical pollution factors of a heating surface, setting the condition that the pollution factors are smaller than or equal to fixed values as a cleaning working condition of the heating surface, and defining the fixed values of the pollution factors as minimum pollution factors F min;
step 5.2, counting the change amplitude of the pollution factor of the heating surface before and after the soot blowing for a plurality of times, and averaging to obtain the decrease amplitude delta F of the pollution factor of the heating surface after the soot blowing:
ΔFi=Fb,i-Fa,i (7)
In the formulas (7) to (8), deltaF i is the variation amplitude of the heating surface pollution factors before and after the ith soot blowing; f b,i is a pollution factor of a heating surface before the ith soot blowing; f a,i is a pollution factor of a heating surface after the ith soot blowing; n is the statistics times;
Step 5.3, calculating a critical pollution factor F lj and generating a critical pollution factor table;
Flj=Fmin+ΔF (9)
in the above formula, F min is the minimum pollution factor F min, and DeltaF is the reduction amplitude of the pollution factor of the heated surface after soot blowing;
step 5.4, judging whether the actual pollution factor is larger than the critical pollution factor, if the actual pollution factor is larger than the critical pollution factor, further judging whether the parameters meet the necessary condition of soot blowing, and performing soot blowing when the necessary condition of soot blowing is met, otherwise, not performing soot blowing; if the actual pollution factor is smaller than or equal to the critical pollution factor, the critical pollution factor is continuously calculated, and a critical pollution factor table is generated.
7. The optimization method for soot blowing of a boiler based on the completion of pollution factors by a time series prediction algorithm according to claim 6, wherein the minimum pollution factor F min in step 5.1 is the average value of all the historical pollution factors minus 2 times the standard deviation of the historical pollution factors:
Fmin=μ-2σ (10)
In the above formula, μ is the average value of the historical pollution factors, and σ is the standard deviation of the historical pollution factors.
8. The optimization method for soot blowing of the boiler for supplementing pollution factors based on the time sequence prediction algorithm according to claim 6, wherein in the step 5.4, a critical pollution factor table under each subdivision condition is generated according to unit load and feed water flow subdivision conditions.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202110942851.8A CN113761794B (en) | 2021-08-17 | 2021-08-17 | Boiler soot blowing optimization method for supplementing pollution factors based on time sequence prediction algorithm |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202110942851.8A CN113761794B (en) | 2021-08-17 | 2021-08-17 | Boiler soot blowing optimization method for supplementing pollution factors based on time sequence prediction algorithm |
Publications (2)
Publication Number | Publication Date |
---|---|
CN113761794A CN113761794A (en) | 2021-12-07 |
CN113761794B true CN113761794B (en) | 2024-05-31 |
Family
ID=78790083
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN202110942851.8A Active CN113761794B (en) | 2021-08-17 | 2021-08-17 | Boiler soot blowing optimization method for supplementing pollution factors based on time sequence prediction algorithm |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN113761794B (en) |
Families Citing this family (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN114285032B (en) * | 2021-12-23 | 2024-05-03 | 广东电网有限责任公司 | Power distribution network line load peak value prediction method and device, computer equipment and medium |
CN115510904B (en) * | 2022-09-26 | 2023-08-01 | 天津大学 | Boiler heating area ash monitoring method based on time sequence prediction |
Citations (7)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
JP2013206363A (en) * | 2012-03-29 | 2013-10-07 | Hitachi Ltd | Plant control device and thermal power generation plant control device |
CN103604132A (en) * | 2013-12-06 | 2014-02-26 | 太原理工大学 | System for online monitoring of gathered dust on convection heating surface of boiler |
CN106322412A (en) * | 2016-08-30 | 2017-01-11 | 上海交通大学 | Coal-fired unit convection heating surface intelligent soot blowing method based on two-dimensional optimization |
CN108278625A (en) * | 2018-01-19 | 2018-07-13 | 华能国际电力股份有限公司上海石洞口第电厂 | A kind of coal-burning boiler high temperature reheater ash-blowing method |
CN112283689A (en) * | 2020-10-29 | 2021-01-29 | 西安工业大学 | On-line monitoring system and detection method for accumulated ash on heating surface of coal-fired power station boiler |
CN112307650A (en) * | 2020-11-27 | 2021-02-02 | 浙江浙能技术研究院有限公司 | Multi-step prediction method for ultra-supercritical boiler heating surface pipe wall overtemperature early warning |
CN112989694A (en) * | 2021-03-08 | 2021-06-18 | 浙江大学 | Segmented monitoring system and method for ash on heating surface |
-
2021
- 2021-08-17 CN CN202110942851.8A patent/CN113761794B/en active Active
Patent Citations (7)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
JP2013206363A (en) * | 2012-03-29 | 2013-10-07 | Hitachi Ltd | Plant control device and thermal power generation plant control device |
CN103604132A (en) * | 2013-12-06 | 2014-02-26 | 太原理工大学 | System for online monitoring of gathered dust on convection heating surface of boiler |
CN106322412A (en) * | 2016-08-30 | 2017-01-11 | 上海交通大学 | Coal-fired unit convection heating surface intelligent soot blowing method based on two-dimensional optimization |
CN108278625A (en) * | 2018-01-19 | 2018-07-13 | 华能国际电力股份有限公司上海石洞口第电厂 | A kind of coal-burning boiler high temperature reheater ash-blowing method |
CN112283689A (en) * | 2020-10-29 | 2021-01-29 | 西安工业大学 | On-line monitoring system and detection method for accumulated ash on heating surface of coal-fired power station boiler |
CN112307650A (en) * | 2020-11-27 | 2021-02-02 | 浙江浙能技术研究院有限公司 | Multi-step prediction method for ultra-supercritical boiler heating surface pipe wall overtemperature early warning |
CN112989694A (en) * | 2021-03-08 | 2021-06-18 | 浙江大学 | Segmented monitoring system and method for ash on heating surface |
Non-Patent Citations (4)
Title |
---|
A Neural-Network-Based Method for Ash Fouling Prediction of Heat Transfer Surface in Coal-Fired Power Plant Boiler;YUANHAO SHI 等;IEEE Access;20210811;第9卷;109584-109604 * |
The algorithm of steam soot blowers operation based on the monitoring of fouling factors of heating surfaces of a coal-fired boiler under operating conditions;Maciej Bujalski 等;E3S Web of Conferences;20191231;第82卷;1-12 * |
基于时序聚类的吹灰预测模型;何源 等;计算机工程;20080531;第34卷(第10期);244-246 * |
姚凯 等.某 480 t/h 锅炉结渣原因分析及处理.电站系统工程.2018,第34卷(第6期),24-26. * |
Also Published As
Publication number | Publication date |
---|---|
CN113761794A (en) | 2021-12-07 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN113761794B (en) | Boiler soot blowing optimization method for supplementing pollution factors based on time sequence prediction algorithm | |
US4466383A (en) | Boiler cleaning optimization with fouling rate identification | |
CN111342459B (en) | Power demand decision analysis system and method | |
CN109654518B (en) | Soot blowing optimization method for heating surface of coal-fired power plant boiler | |
CN117013606B (en) | Intelligent energy storage control system for photovoltaic power generation based on artificial intelligence | |
CN115111594B (en) | Intelligent regulation and control system and method for heat accumulating type thermal oxidation furnace | |
CN112836935A (en) | Wisdom water affairs management platform suitable for power plant | |
CN112381210B (en) | Coal-fired unit water-cooling wall temperature prediction neural network model | |
CN115471014A (en) | Power grid enterprise electricity carbon consumption prediction method based on deep neural network | |
CN110888403A (en) | Intelligent soot blowing closed-loop control system based on minimum loss boiler convection heating surface | |
CN109858125A (en) | A kind of fired power generating unit net coal consumption rate calculation method based on radial base neural net | |
CN112783115A (en) | Online real-time optimization method and device for steam power system | |
CN115198047B (en) | Hot blast stove combustion monitoring system and method based on big data analysis | |
CN112304657A (en) | Method for dividing key easy-scaling groups of air preheater and monitoring heat exchange efficiency state | |
CN116720446A (en) | Method for monitoring thickness of slag layer of water-cooled wall of boiler in real time | |
CN116227322A (en) | Online self-adaption method for digital twin system model of nuclear power plant | |
CN112668751B (en) | Method and device for establishing unit optimization scheduling model | |
CN115183318A (en) | Heat supply adjusting method based on load prediction and heat progress consistency and related equipment | |
CN113946789A (en) | Method for calculating waste incineration amount of four-furnace three-machine waste incineration power plant in real time | |
CN111852596B (en) | Method for predicting operation parameters and relative power generation coal consumption rate of boiler steam turbine generator unit | |
CN114091734A (en) | Optimization method for improving boiler soot blowing large yield based on segmented threshold | |
CN114779722B (en) | Intelligent combustion optimization control system and method for coal-fired power station boiler | |
CN113433826B (en) | Automatic control method for optimization of ash discharge period | |
CN116734307A (en) | Heating system control method based on adaptive model predictive control technology | |
CN115855542A (en) | Energy flow direction and loss data online monitoring system and monitoring method |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
PB01 | Publication | ||
PB01 | Publication | ||
SE01 | Entry into force of request for substantive examination | ||
SE01 | Entry into force of request for substantive examination | ||
GR01 | Patent grant | ||
GR01 | Patent grant |