A cement kiln thermal diagnosis analysis method and system based on heat balance calculation

Abstract

The application discloses a cement kiln thermal diagnosis analysis method and system based on heat balance calculation, and relates to the field of cement kiln thermal diagnosis.The analysis method comprises the following steps: obtaining the corresponding heat balance state by acquiring the thermal parameters and material composition data of the cement kiln, and analyzing the material balance relationship between the input and output of the cement kiln; taking the ratio between the clinker formation heat of the cement kiln and the total input heat as a thermal efficiency index, evaluating the energy utilization efficiency of the firing system of the cement kiln, combining the heat balance state and the material balance relationship to generate a diagnosis index; comparing the diagnosis index with the historical optimal value to determine the deviation, evaluating the thermal loss anomaly and the running state of the cement kiln according to the deviation, constructing a cause-effect map to predict the development trend of the cement kiln, and obtaining the combustion control result of the cement kiln.The application can help understand how various factors affect the overall performance of the thermal system, and facilitate the optimization and adjustment of different parameters according to the cause-effect relationship in the later stage.

Description

TECHNICAL FIELD

[0001] The present application relates to the field of cement kiln thermal diagnosis, in particular to a cement kiln thermal diagnosis analysis method and system based on heat balance calculation. BACKGROUND

[0002] Cement kiln thermal engineering refers to the application of thermal energy utilization, heat exchange and heat loss and other thermal engineering technologies in the cement production process. As one of the core equipment of cement production, the operation state of the thermal system of the cement kiln directly affects energy consumption, production efficiency, environmental protection emission and product quality. The thermal system of the cement kiln mainly includes a combustion system, a heat recovery system, a kiln body and a cooling system.

[0003] The traditional cement kiln thermal calibration work is mainly completed manually, and the whole process takes as long as 72 hours. Frequent manual data collection is required, and data summarization and analysis rely on manual operation. Not only is the timeliness poor, but errors are also prone to occur. At the same time, this calculation method based on historical data cannot real-time feedback the real state of the system, and it is difficult to meet the demand for efficient and accurate control of modern cement production.

[0004] However, in the prior art, the thermal diagnosis result cannot be used to predict the thermal development trend of the cement kiln in a certain future time, so that the state of the cement kiln cannot be adjusted before the problem actually occurs, and the production efficiency, product quality and environmental protection level of the cement kiln are reduced.

[0005] At present, no effective solution has been proposed for the problems in the related art. SUMMARY

[0006] In view of the problems in the related art, the present application proposes a cement kiln thermal diagnosis analysis method and system based on heat balance calculation to overcome the above technical problems existing in the prior art.

[0007] To this end, the specific technical solutions adopted by the present application are as follows:

[0008] In a first aspect, the present application proposes a cement kiln thermal diagnosis analysis method based on heat balance calculation, which comprises:

[0009] Obtaining the heat balance state corresponding to the cement kiln thermal parameter and material composition data, and analyzing the material balance relationship between the material input and the material output of the cement kiln according to the law of conservation of mass;

[0010] Taking the ratio between the clinker formation heat of the cement kiln and the total input heat as a thermal efficiency index, evaluating the energy utilization efficiency of the burning system of the cement kiln, and generating a diagnosis index in combination with the heat balance state and the material balance relationship;

[0011] The deviation condition is determined by comparing the diagnostic index with the historical optimal value, the thermal loss abnormality and the running state of the cement kiln thermal worker are evaluated according to the deviation condition, and a running cause-effect map of the cement kiln thermal worker is constructed;

[0012] The development trend of the cement kiln thermal worker is predicted based on the running cause-effect map and the target constraint condition, and a dynamic optimization model is constructed by combining the model predictive control technology to obtain the adaptive combustion control result of the cement kiln.

[0013] Preferably, the deviation condition is determined by comparing the diagnostic index with the historical optimal value, the thermal loss abnormality and the running state of the cement kiln thermal worker are evaluated according to the deviation condition, and a running cause-effect map of the cement kiln thermal worker is constructed, including:

[0014] The historical optimal value and the optimal running state are determined based on the historical running data, the deviation value between the diagnostic index and the historical optimal value is calculated, the diagnostic index with a deviation value greater than a deviation threshold is selected, and a candidate set of abnormal evaluation variables is obtained;

[0015] The abnormal evaluation variable candidate set is detected by using correlation analysis and support vector description technology, and the thermal loss abnormality of the cement kiln thermal worker is evaluated according to the detection result and the enhanced local linear embedding technology;

[0016] The reference sample is selected from the abnormal evaluation variable candidate set to establish an evaluation original matrix, and the distance between the evaluation original matrix and the optimal running state is judged, and the running state of the cement kiln thermal worker is evaluated based on the distance result;

[0017] The key variables affecting the thermal worker state are determined based on the abnormal evaluation variable candidate set, the thermal loss abnormality and the running state of the cement kiln thermal worker, and the running cause-effect map of the cement kiln thermal worker is constructed according to the causal relationship between the key variables.

[0018] Preferably, the development trend of the cement kiln thermal worker is predicted based on the running cause-effect map and the target constraint condition, and a dynamic optimization model is constructed by combining the model predictive control technology to obtain the adaptive combustion control result of the cement kiln, including:

[0019] The target constraint condition is set according to the objective function and the constraint condition of the cement kiln thermal worker control, and the trend elements affecting the development of the cement kiln thermal worker are extracted from the running cause-effect map to construct a development feature vector;

[0020] The development feature vector is analyzed by using semantic decision tree for correlation clustering, a trend target is generated, the motion trend of the trend target is predicted as the development trend of the cement kiln thermal worker by combining the target constraint condition;

[0021] A mathematical model for adaptive combustion control of the cement kiln is established, and feedback compensation is performed on the cement kiln combustion control in combination with the optimal running state to output the optimal combustion control result of the cement kiln.

[0022] In a second aspect, the present application further provides a cement kiln thermal diagnosis analysis system based on heat balance calculation, which comprises:

[0023] a balance relationship acquisition module, configured to acquire a heat balance state corresponding to the cement kiln thermal parameters and material composition data calculation, and analyze a material balance relationship between a material input and a material output of the cement kiln according to the law of conservation of mass;

[0024] a diagnosis index generation module, configured to take a ratio between a clinker formation heat of the cement kiln and a total input heat as a thermal efficiency index, evaluate an energy utilization efficiency of a burning system of the cement kiln, and generate a diagnosis index in combination with the heat balance state and the material balance relationship;

[0025] a causal graph construction module, configured to determine a deviation condition by comparing the diagnosis index with a historical optimal value, evaluate a thermal loss anomaly and a running state of the cement kiln thermal system according to the deviation condition, and construct a running causal graph of the cement kiln thermal system;

[0026] a dynamic optimization control module, configured to predict a development trend of the cement kiln thermal system based on the running causal graph and a target constraint condition, construct a dynamic optimization model in combination with a model prediction control technology, and acquire a self-adaptive combustion control result of the cement kiln.

[0027] The present application has the following beneficial effects:

[0028] 1. The present application realizes self-adaptive optimization of the cement kiln thermal system by heat balance calculation and material balance analysis in combination with a model prediction control technology, improves overall energy efficiency, production stability and environmental protection performance, generates a series of diagnosis indexes (such as thermal loss rate, combustion efficiency, material balance condition, etc.) by processing real-time running data, compares these indexes with historical optimal values, analyzes deviation conditions, clearly presents causal relationships between different operation parameters by constructing a thermal running causal graph, and helps to understand how each factor affects the overall performance of the thermal system, which is convenient for optimizing and adjusting different parameters according to the causal relationships to realize adjustment of the best running state.

[0029] 2. The present application can predict the thermal development trend of the cement kiln in a certain time in the future by the target constraint condition (such as emission standard, production cost, equipment load, etc.) and the running causal graph, can adjust before problems really occur, avoids large-scale energy waste or production stagnation caused by uncontrollable factors, and finally realizes real-time adjustment of combustion parameters, air-coal ratio, kiln body temperature, etc. by model prediction control technology in combination with the thermal state and the target constraint condition, improves production efficiency, product quality and environmental protection level of the cement kiln. BRIEF DESCRIPTION OF DRAWINGS

[0030] In order to more clearly illustrate the technical solutions in the embodiments of the present application or the prior art, the accompanying drawings needed in the embodiments will be briefly introduced as follows. Obviously, the drawings in the following description only constitute some embodiments of the present application, and for those skilled in the art, other drawings can also be obtained based on these drawings without any creative effort.

[0031] Figure 1 is a flow chart of a cement kiln thermal diagnosis analysis method based on heat balance calculation according to an embodiment of the present application;

[0032] Figure 2 is a principle block diagram of a cement kiln thermal diagnosis analysis system based on heat balance calculation according to an embodiment of the present application;

[0033] Figure 3 is an online thermal diagnosis schematic diagram of a burning system in a cement kiln thermal diagnosis analysis method based on heat balance calculation according to an embodiment of the present application;

[0034] Figure 4 is a software network architecture diagram in a cement kiln thermal diagnosis analysis method based on heat balance calculation according to an embodiment of the present application;

[0035] Figure 5 is an online thermal diagnosis general diagram in a cement kiln thermal diagnosis analysis method based on heat balance calculation according to an embodiment of the present application;

[0036] Figure 6 is a clinker yield trend analysis diagram in a cement kiln thermal diagnosis analysis method based on heat balance calculation according to an embodiment of the present application;

[0037] Figure 7 is a rotary kiln system design index comparison analysis diagram in a cement kiln thermal diagnosis analysis method based on heat balance calculation according to an embodiment of the present application;

[0038] Figure 8 is a rotary kiln system day-on-day analysis diagram in a cement kiln thermal diagnosis analysis method based on heat balance calculation according to an embodiment of the present application;

[0039] Figure 9 is one of rotary kiln system air core index analysis diagrams in a cement kiln thermal diagnosis analysis method based on heat balance calculation according to an embodiment of the present application;

[0040] Figure 10 is the other of rotary kiln system air core index analysis diagrams in a cement kiln thermal diagnosis analysis method based on heat balance calculation according to an embodiment of the present application;

[0041] Figure 11is one of rotary kiln system heat balance core index analysis charts in a cement kiln thermal diagnosis analysis method based on heat balance calculation according to an embodiment of the present application;

[0042] Figure 12 is the second one of rotary kiln system heat balance core index analysis charts in a cement kiln thermal diagnosis analysis method based on heat balance calculation according to an embodiment of the present application;

[0043] Figure 13 is one of rotary kiln system material balance analysis charts in a cement kiln thermal diagnosis analysis method based on heat balance calculation according to an embodiment of the present application;

[0044] Figure 14 is the second one of rotary kiln system material balance analysis charts in a cement kiln thermal diagnosis analysis method based on heat balance calculation according to an embodiment of the present application;

[0045] Figure 15 is an online thermal diagnosis report chart in a cement kiln thermal diagnosis analysis method based on heat balance calculation according to an embodiment of the present application.

[0046] In the figure:

[0047] 1, balance relationship acquisition module; 2, diagnosis index generation module; 3, causal graph construction module; 4, dynamic optimization control module. DETAILED DESCRIPTION

[0048] To further illustrate the embodiments, the present application provides drawings which are part of the disclosure of the present application, mainly used to illustrate the embodiments, and can explain the operating principle of the embodiments in conjunction with the related description of the specification. With reference to these contents, those skilled in the art should understand other possible implementations and advantages of the present application.

[0049] According to an embodiment of the present application, a cement kiln thermal diagnosis analysis method and system based on heat balance calculation are provided.

[0050] The present application will be further described in conjunction with the drawings and specific embodiments, as shown in the drawings, the cement kiln thermal diagnosis analysis method based on heat balance calculation according to an embodiment of the present application, the analysis method includes: Figure 1

[0051] Step S1, obtain the corresponding heat balance state of the cement kiln thermal parameter and material composition data calculation, and analyze the material balance relationship between the input and output of the cement kiln material according to the law of conservation of mass.

[0052] ​It needs to be explained that in the process of achieving the relationship between thermal equilibrium state and material balance, based on the OPC protocol, the communication connection is established between the data acquisition device and the DCS system of the cement kiln burning system, and 39 kinds of thermal parameters and material composition data are obtained in real time, including temperature parameters (preheater outlet gas temperature of each stage, decomposition furnace temperature, different positions in the kiln, cooling machine gas temperature of each chamber, clinker outlet temperature); pressure parameters (preheater inlet and outlet pressure of each stage, decomposition furnace pressure, related to the combustion and airflow condition in the decomposition furnace; kiln head cover pressure, kiln tail pressure, cooling machine inlet and outlet pressure); air volume parameters (primary air volume, secondary air volume, tertiary air volume, cooling machine chamber air volume, system total exhaust air volume); material composition data (chemical composition of raw material entering the kiln, including calcium oxide, silicon dioxide, aluminum oxide, iron oxide content) and the like, and the raw data collected is preprocessed, and the data after preprocessing is subjected to thermal balance calculation. The thermal balance calculation based on the thermal balance, material balance and thermal efficiency calculation model of GB / T26281-2010 standard, the data after preprocessing is calculated, and each thermal index is obtained. Specifically, the online thermal calibration measurement points of the kiln system are shown in Table 1:

[0053] Table 1: Online thermal calibration measurement points of the kiln system

[0054]

[0055]

[0056]

[0057] The specific process of thermal balance management is as follows:

[0058] (1) Coal combustion heat: the calculation formula is Qrr=mr×Qnet,ar=unit coal combustion heat * received based heat, and the data required to be obtained includes preheater C1 outlet total exhaust gas volume, C1 outlet dust concentration;

[0059] (2) Coal sensible heat: the calculation formula is Q r =(Myr×cyr×ty r +Mfr×cfr×tfr) / Msh=(kiln head coal feeding amount * kiln coal specific heat * kiln coal temperature + decomposition furnace coal feeding amount * decomposition furnace coal specific heat * decomposition furnace coal temperature) / clinker yield; the data required to be obtained includes preheater C1 outlet total exhaust gas volume, C1 outlet dust concentration (measured data);

[0060] (3) Raw material sensible heat entering the preheater: the calculation formula is Qs=ms×cs×ts=unit raw material sensible heat entering the preheater=raw material consumption * raw material specific heat entering the preheater * raw material temperature entering the preheater;

[0061] (4) Primary air sensible heat:

[0062] The formula is Qlk = (Vylk x cylk x tylk + Vysm x cysm x tysm + Vfsm x cfsm x tfsm) / Msh = (primary air volume at kiln head * primary air specific heat at kiln head * primary air temperature at kiln head + coal conveying air volume at kiln head * coal conveying air specific heat at kiln head * coal conveying air temperature at kiln head + coal conveying air volume at decomposing furnace * coal conveying air specific heat at decomposing furnace * coal conveying air temperature at decomposing furnace) / clinker output;

[0063] (5) Cooling drum air sensible heat: the formula is QLk = VLk x cLk x tLk / Msh = cooling drum air volume * average specific heat of cooling drum air * average temperature of cooling drum air / clinker output; data required: total exhaust gas volume at C1 outlet, dust concentration at C1 outlet;

[0064] (6) System air leakage sensible heat: the formula is Qlok = Vlok x ck x tk / Msh = (system air leakage volume * specific heat of ambient air * ambient temperature) / clinker output; data required: total exhaust gas volume at C1 outlet, dust concentration at C1 outlet;

[0065] (7) Total heat income of firing system:

[0066] The formula is Qzr = Qrr + Qr + Qs + Q1k + QLk + Qlok = coal combustion heat + coal sensible heat + raw material sensible heat + primary air sensible heat + cooling drum air sensible heat + system air leakage sensible heat;

[0067] (8) Heat expenditure of firing system:

[0068] The formula is Qsh = 17.19 Al2O3sh + 27.10 MgOsh + 32.01 CaOsh - 2.47 Fe2O3sh - 21.40 SiO2sh = 17.19 * Al2O3 content + 27.10 * MgO content + 32.01 * CaO content - 2.47 * Fe2O3 content - 21.40 * SiO2 content; data required: Al2O3 content, MgO content, CaO content, Fe2O3 content, SiO2 content;

[0069] (9) Water consumption heat of raw material in steam generation: the formula is Qss = ms x Ws / 100 x qqh = actual raw material consumption coefficient of raw material entering kiln * water content of raw material entering kiln / 100 * water vaporization heat; data required: total exhaust gas volume at C1 outlet, dust concentration at C1 outlet, temperature of raw material entering preheater C1;

[0070] (10) Cooling machine clinker sensible heat: the calculation formula is QLsh = mLsh x cLsh x tLsh = cooling machine clinker quantity * cooling machine clinker specific heat * cooling machine clinker temperature; the data required to be obtained are cooling machine to settling chamber exhaust volume (square tube) static pressure, cooling machine exhaust static pressure, cooling machine to settling chamber exhaust (circular tube) static pressure;

[0071] (11) Preheater outlet waste gas sensible heat: the formula calculation is Qf = Vf x cf x tf / Msh = preheater C1 outlet total pipe waste gas volume * preheater outlet waste gas specific heat * preheater outlet waste gas temperature / clinker yield;

[0072] (12) Preheater outlet fly ash sensible heat: the calculation formula is Qfh = mfh x cfh x tfh = preheater outlet fly ash quantity * preheater outlet waste gas fly ash specific heat * preheater outlet waste gas fly ash temperature;

[0073] (13) Fly ash dehydration and carbonate decomposition heat consumption: the calculation formula is H2OS = Al2O3s x 36 / 102; CO2S = (CaOS / 100 x 44 / 56 + MgOS / 100 x 44 / 40.3) * 100;

[0074] Qtf = mfh x (100 - Lfh) / (100 - Ls) x H2Os / 100 x 6690 + 〔mfh x (100 - Lfh) / (100 - Ls) x CO2S / 100 - mfh x Lfh / 100〕x 100 / 44 x 1660, that is, fly ash dehydration and carbonate decomposition heat consumption = preheater outlet fly ash quantity * (100 - fly ash loss on ignition) / (100 - raw material loss on ignition) * combined water content in raw material / 100 * 6690 + (preheater outlet fly ash quantity * (100 - fly ash loss on ignition) / (100 - raw material loss on ignition) * CO2 content in raw material / 100) * 100 / 44 * 1660;

[0075] (14) Cooling machine to settling chamber exhaust sensible heat: the calculation formula is Qyp = Vyp x cyp x typ / Msh = cooling machine to settling chamber exhaust volume (square tube) * cooling machine to settling chamber exhaust specific heat * cooling machine to settling chamber exhaust temperature / clinker yield;

[0076] (15) Cooling machine to settling chamber exhaust fly ash sensible heat (square tube): the calculation formula is Qyph = myph x cyph x typ = cooling machine to settling chamber exhaust fly ash quantity (square tube) * cooling machine to settling chamber exhaust fly ash specific heat * cooling machine to settling chamber exhaust fly ash temperature;

[0077] (16) Cooling machine de-sedimentation chamber exhaust sensible heat (circular pipe): the calculation formula is Qmp=Vmp*cmp*tmp / Msh=cooling machine de-sedimentation chamber exhaust volume (circular pipe)*cooling machine de-sedimentation chamber exhaust specific heat*cooling machine de-sedimentation chamber exhaust temperature / clinker output;

[0078] (17) Cooling machine de-sedimentation chamber exhaust fly ash sensible heat (circular pipe): the calculation formula is Qmph=mmph*cmph*tmph=cooling machine de-sedimentation chamber exhaust fly ash amount (circular pipe)*cooling machine de-sedimentation chamber exhaust fly ash specific heat*cooling machine de-sedimentation chamber exhaust fly ash temperature;

[0079] (18) Cooling machine exhaust sensible heat: the calculation formula is Qfp=Vfp*cfp*tfp / Msh=cooling machine exhaust volume*cooling machine specific heat*cooling machine temperature / clinker output;

[0080] (19) Cooling machine exhaust fly ash sensible heat: the calculation formula is Qfph=mfph*cfph*tfph=cooling machine exhaust amount*cooling machine exhaust fly ash specific heat*cooling machine exhaust fly ash temperature;

[0081] (20) Chemical incomplete combustion heat:

[0082] The calculation formula is Qhb=Vf*(1-H2Of / 100)*COf / 1000000*12630 / Msh=preheater C1 outlet total pipe exhaust volume*(1-preheater outlet gas water vapor volume percentage content / 100)*preheater outlet gas CO dry gas volume content / 1000000*12630 / clinker output;

[0083] (21) Mechanical incomplete combustion heat: the calculation formula is Qjb=Lsh / 100*33874=clinker loss amount / 100*33847;

[0084] (22) Sintering system surface heat dissipation amount: the calculation formula is Qbs=Qbsh / Msh=sintering system per hour surface heat dissipation amount / clinker output;

[0085] (23) System other heat expenditure:

[0086] The calculation formula is:

[0087] Qqt=Qzr-(Qsh+Qss+QLsh+Qf+Qfh+Qtf+Qyp+Qyph+Qfp+Qfph+Qhb+Qjb+Qbs)=system heat total income-(all loss heat);

[0088] (24) Rotary kiln thermal efficiency=clinker formation heat / (pulverized coal combustion heat+combustion heat of combustible substances in raw materials)*100;

[0089] (25) System surface heat dissipation test results: preheater total = C1+C2+C3+C4+C5+C6 (heat dissipation per hour); sintering system surface heat dissipation = grate cooler + kiln hood + flue + tertiary air pipe + kiln body surface + decomposition furnace.

[0090] Sintering system material balance calculation (incoming material):

[0091] (1) Pulverized coal consumption: the calculation formula is mr= (Myr+Mfr) / Msh (pulverized coal consumption = (kiln head coal feeding amount + decomposition furnace coal feeding amount) / clinker output);

[0092] (2) Raw material consumption: the calculation formula is ms=Ms / Msh (raw material consumption = raw material feeding amount / clinker output);

[0093] (3) Primary air amount: the calculation formula is mlk= (Vy1k+Vysm+Vfsm) / Msh x pK (primary air amount = (kiln head primary air volume + kiln head coal conveying air volume + decomposition furnace coal conveying air volume) / clinker output * air density);

[0094] (4) Cooling machine air blowing amount: the calculation formula is mLK=VLK / Msh x pK (cooling machine air blowing amount = cooling machine air blowing volume / clinker output * air density);

[0095] (5) Total incoming material of sintering system: the calculation formula is mzr=mr+ms+mlk+mLk+mlok (total material incoming = pulverized coal consumption + raw material consumption + primary air amount + cooling machine air blowing amount + system air leakage amount).

[0096] Sintering system material balance calculation (outgoing material):

[0097] (1) Clinker output from cooling machine: the calculation formula is mLsh=1-(Vyp x dyp+Vfp x dfp+VMp x dMp) / 1000 / Msh (clinker output from cooling machine = 1- (cooling machine sedimentation chamber exhaust air volume (square pipe) * cooling machine sedimentation chamber exhaust air dust concentration (square pipe) + cooling machine exhaust air volume * cooling machine exhaust air dust concentration + cooling machine sedimentation chamber exhaust air volume (circular pipe) * cooling machine sedimentation chamber exhaust air dust concentration (circular pipe)) / 1000 / clinker output);

[0098] (2) Preheater outlet exhaust air amount: the calculation formula is mf=Vf x pF / Msh (preheater outlet exhaust air amount = preheater C1 outlet total pipe exhaust air volume * preheater outlet exhaust air density / clinker output);

[0099] (3) Preheater outlet fly ash amount: the calculation formula is mfh = Vf x δf / 1000 / Msh (preheater outlet fly ash amount = preheater C1 outlet total pipe exhaust gas volume * preheater C1 outlet total pipe exhaust gas dust concentration / 1000 / clinker output);

[0100] (4) Cooling machine de-sedimentation chamber exhaust amount (square tube): the calculation formula is myp = Vyp x pk / Msh (cooling machine de-sedimentation chamber exhaust amount (square tube) = cooling machine de-sedimentation chamber exhaust volume (square tube) * air density / clinker output);

[0101] (5) Cooling machine de-sedimentation chamber exhaust fly ash amount (square tube): the calculation formula is myph = Vyp x δyp / 1000 / Msh (cooling machine de-sedimentation chamber exhaust fly ash amount (square tube) = cooling machine de-sedimentation chamber exhaust volume (square tube) * cooling machine de-sedimentation chamber exhaust dust concentration (square tube) / 1000 / clinker output);

[0102] (6) Cooling machine de-sedimentation chamber exhaust amount (round tube): the calculation formula is mmph = Vmp x δmp / 1000 / Msh (cooling machine de-sedimentation chamber exhaust amount (round tube) = cooling machine de-sedimentation chamber exhaust volume (round tube) * air density / clinker output)

[0103] (7) Cooling machine exhaust amount: the calculation formula is mfp = Vfp x pk / Msh (cooling machine exhaust amount = cooling machine exhaust exhaust volume * air density / clinker output);

[0104] (8) Cooling machine exhaust fly ash amount: the calculation formula is mfp = Vfp x pk / Msh (cooling machine exhaust amount = cooling machine exhaust exhaust volume * air density / clinker output);

[0105] (9) Cooling machine exhaust fly ash amount: the calculation formula is mfph = Vfp x δfp / 1000 / Msh (cooling machine exhaust fly ash amount = cooling machine exhaust exhaust volume * cooling machine exhaust dust concentration / 1000 / clinker output.

[0106] When performing material balance calculation, according to the law of conservation of mass, the balance relationship between the input amount of materials such as raw meal, coal powder and air into the kiln and the output amount of materials such as clinker, exhaust gas and fly ash out of the kiln is analyzed, the conversion and distribution of materials in the system are determined by accurately calculating the mass and composition change of each material (mainly involving: the relationship between C1 dust content and feeding amount, clinker output = theoretical material consumption / (1- production loss rate), the theoretical material consumption is determined according to the coal ash mixing amount and the raw meal loss on ignition amount, and the production loss mainly refers to the amount of raw meal flying off; CO content in flue gas and CO content in smoke chamber evaluate the combustion condition of in-furnace and in-kiln combustion), which provides basis for subsequent process adjustment.

[0107] Step S2, the ratio between the cement kiln clinker formation heat and the total input heat is taken as the thermal efficiency index to evaluate the energy utilization efficiency of the cement kiln firing system, and the diagnosis index is generated in combination with the heat balance state and the material balance relationship.

[0108] It should be explained that the thermal efficiency is calculated, and the ratio between the clinker formation heat and the total input heat is taken as the thermal efficiency index to evaluate the energy utilization efficiency of the cement kiln firing system. Through the calculation of the thermal efficiency, the energy-saving effect of the system can be directly understood, and key data support is provided for optimizing the production process.

[0109] The calculation results of the heat balance, the material balance and the thermal efficiency are analyzed in depth, the actual values of the indexes are compared with the historical optimal values and the design values, and the deviation of the indexes is determined. By establishing a scientific comparison and analysis model (the comparison model is established by determining the key comparison indexes, such as the thermal efficiency, the heat consumption, the waste gas quantity per unit of clinker, etc.), the problems existing in the system running process and the potential optimization space can be accurately found out. The reasons for the index deviation are quantitatively analyzed, and the influence of various factors on the index deviation is comprehensively considered.

[0110] For example, if the sensible heat of the cooling machine waste gas increases, the data analysis and logical reasoning function of the system (for example: the normal air volume range of the cooling machine and the corresponding clinker temperature range are set; the abnormal conditions of the parameters related thereto are analyzed: the secondary air temperature, the tertiary air temperature, the waste heat extraction power generation air temperature, the coal mill extraction waste gas air temperature, the clinker temperature, the cooling effect of the cooling machine is evaluated; a data-based evaluation model is established) is used to analyze whether the cooling effect of the clinker is poor due to the excessive air volume of the cooling machine or other reasons, and the contribution proportion of each factor to the deviation is determined through accurate calculation. A diagnosis report containing core index comparison, deviation reason quantitative analysis and process adjustment suggestions for the deviation is generated, the core indexes include clinker yield, heat consumption, cooling machine heat recovery efficiency, etc., which are the key parameters for measuring the performance of the cement kiln firing system; the process adjustment suggestions are made according to the specific deviation conditions, for example, if the heat consumption is too high, it is suggested to adjust the air volume, optimize the combustion parameters or check the equipment sealing condition; if the clinker yield is insufficient, the raw material batching can be adjusted, the running speed of the kiln can be increased, etc.

[0111] It should be explained that the firing system thermal diagnosis system is deployed in the production management network of the cement enterprise, penetrates the DCS production control network and the production management office network, connects the DCS production network for field layer data acquisition through the OPC communication protocol, can be self-set for data entry module, realizes continuous online firing system thermal diagnosis through digital technologies such as data acquisition, data storage and data calculation, supports PC and mobile access, meets the network security specifications, can be independently deployed, or can be deployed as a sub-module of the production control digital MES platform, and the software network architecture is as follows Figure 4as shown.

[0112] As Figure 3 shown, the determination of thermal performance is carried out in accordance with GB / T26282-2010 "Cement Rotary Kiln Heat Balance Determination Method", and the processing of determination data is carried out in accordance with GB / T26281-2010 "Cement Rotary Kiln Heat Balance, Thermal Efficiency, Comprehensive Energy Consumption Calculation Method". Through the analysis of the operation condition and data of the determination system, the automatic determination and comprehensive evaluation of the thermal performance of the kiln system are regularly completed. The online thermal diagnosis system overview large screen page of the firing system realizes the daily online evaluation of the key evaluation indexes, automatically completes the data summary of production, quality and energy, automatically generates the summary report, realizes the visual analysis of the heat balance and material balance of the kiln system and the grate cooler system, and automatically generates the thermal diagnosis report, realizes the fusion of automatic diagnosis report and expert diagnosis, automatically pushes the diagnosis report function, and assists decision-making.

[0113] As Figure 5 shown, the online thermal diagnosis overview diagram has the following specific function description:

[0114] The previous day's overall operation indexes of the cement kiln include yield, process power consumption, heat consumption, kiln thermal efficiency and cooler thermal efficiency. The history optimal indexes of the cement kiln production operation process are compared and analyzed with the indexes of the previous day. The design indexes of the cement kiln are compared and analyzed with the indexes of the previous day. The clinker yield, heat consumption and cooler heat recovery efficiency are compared and analyzed in a day-on-day comparison. The coal industrial analysis, out-of-kiln clinker composition and optimal indexes of the production line are analyzed. The cooling fan air volume, out-of-kiln clinker sensible heat and cooler waste gas sensible heat core indexes are analyzed. The thermal calibration analysis report is analyzed.

[0115] As Figure 6 shown, the clinker yield trend analysis diagram has the following specific function description:

[0116] Clicking the "Home Page" clinker yield title penetrates to query the change trend of the historical yield. The baseline is the optimal yield, which realizes the comparison and analysis with the historical optimal yield. The process power consumption, heat consumption, kiln thermal efficiency and cooler thermal efficiency can realize this function, which is used for historical data trend query and traceability, and comparison and analysis with the historical optimal indexes.

[0117] As Figure 7 shown, the rotary kiln system design index comparison and analysis diagram has the following specific function description:

[0118] Since the original coal replacement cycle of the factory is 7 days, the index analysis period is 7 days in a rolling manner. The analysis of the core indexes of the clinker process power consumption, heat consumption and cooler heat recovery efficiency feeds back the power consumption, coal consumption and cooler heat recovery effect of the entire kiln system, realizes the trend comparison and analysis with the design indexes, and intuitively feeds back the running state of the current kiln system through the maximum degree of data change trend display through adaptive mode.

[0119] like Figure 8 The daily cycle analysis chart of the rotary kiln system shown below has the following detailed functional descriptions:

[0120] Daily comparison analysis of core indicators such as clinker output, heat consumption, and cooler heat recovery efficiency; using bar charts to reflect the daily trend of changes, comparative analysis can intuitively understand the operating status of the kiln system.

[0121] like Figure 9 and Figure 10 The core performance indicators of the rotary kiln system's air consumption are shown in the diagram below, with the specific functions described as follows:

[0122] The daily comparison method is used to analyze the unit air consumption of the cooler and the unit air consumption of the kiln system; the daily comparison analysis provides feedback on the trend of air consumption changes, and the qualitative and quantitative analysis helps to assist decision-making.

[0123] like Figure 11 and Figure 12 The diagram showing the core thermal balance indicators of the rotary kiln system is described in detail below:

[0124] The daily comparison method analyzes the pulverized coal combustion heat and clinker formation heat in the kiln system heat balance; it also analyzes the sensible heat of clinker exiting the kiln and the sensible heat of cooler exhaust gas in the cooler heat balance; the daily comparison analysis provides feedback on the trend changes compared to the previous day, and the qualitative and quantitative analysis helps to assist decision-making.

[0125] like Figure 13 The heat balance analysis diagram of the rotary kiln system shown below has the following specific functional descriptions:

[0126] It mainly enables the analysis of heat balance in the kiln system and the cooler; it calculates all received heat and all emitted heat in the kiln system and the cooler; it graphically displays the proportion of each sensible heat; it supports 7-day and 30-day queries of kiln heat balance and cooler heat balance trends; and it is beneficial for analyzing the trend changes of heat received and emitted by the rotary kiln and cooler during long-term use.

[0127] like Figure 14 The material balance analysis diagram of the rotary kiln system shown below has the following specific functional descriptions:

[0128] It primarily enables material balance analysis for the kiln system and cooler; the material balance analysis for the kiln system and cooler calculates all received materials and all issued materials; it graphically displays the proportion of each material; the material balance trend of the kiln system and cooler supports 7-day and 30-day queries; in long-term use, it is beneficial for analyzing the trend changes of material receipts and expenditures in the rotary kiln and cooler.

[0129] like Figure 15The online thermal diagnostic report shown below has the following detailed function description:

[0130] The core production and operation indicators of the kiln system include: output, comprehensive clinker power consumption, clinker burning heat consumption, kiln thermal efficiency, cooler heat recovery efficiency, and clinker temperature. Qualitative analysis of the kiln system is conducted from three aspects: heat consumption, cooler heat recovery efficiency, and cooler cooling efficiency. The factory's coal replacement cycle is 7 days; comparing the usage of two coal piles over the past two weeks with today's indicator data provides more guidance. Quantitative analysis prioritizes and quantifies the main influencing factors. Through both qualitative and quantitative analysis, data support is provided for production auxiliary decision-making and management. The thermal report is pushed to the production management WeChat group daily at 14:00 for easy access by users.

[0131] The data sources for online thermal diagnostics in specific implementation are shown in Tables 2 to 4:

[0132] Table 2: Chemical Analysis of Raw Materials

[0133]

[0134] Table 3: Chemical Analysis of Clinker

[0135]

[0136] Table 4: Industrial Analysis of Pulverized Coal

[0137]

[0138] Finally, diagnostic reports are automatically generated at preset time intervals (such as daily) and stored in the system database as electronic documents (such as PDF, Excel, etc.). The reports can also be pushed to the terminal devices (such as PCs, mobile phones, etc.) of relevant production managers to keep abreast of the system's operating status and make scientific decisions.

[0139] This enables real-time and continuous monitoring and diagnosis of the thermal parameters of the cement kiln firing system, greatly shortening the diagnosis cycle from the traditional 72-hour manual calibration to real-time updates. This improves the timeliness and scientific nature of production decisions. Furthermore, through precise calculations of heat balance, material balance, and thermal efficiency, it can accurately identify problems and potential optimization opportunities in the system operation process, providing a strong basis for process adjustments, effectively reducing heat consumption, improving the heat recovery efficiency of the cooler, and lowering production costs.

[0140] Meanwhile, detailed diagnosis reports are automatically generated daily and timely pushed to relevant managers, so that they can grasp the system running status at any time, take measures to solve problems in time, improve the efficiency and level of production management, and thus be suitable for various types of cement kiln burning systems, including different process types, kiln scales and fuel types, have wide application prospects and popularization value, help cement enterprises to achieve the goal of energy saving and emission reduction, reduce CO2 emissions, meet the requirements of the state for green development of building materials industry, and have good social and environmental benefits.

[0141] Step S3, comparing the diagnosis index with the historical running optimal value to determine the deviation condition, evaluating the thermal loss abnormality and running state of the cement kiln thermal system according to the deviation condition, and constructing a running cause-effect map of the cement kiln thermal system.

[0142] In one embodiment, comparing the diagnosis index with the historical running optimal value to determine the deviation condition, evaluating the thermal loss abnormality and running state of the cement kiln thermal system according to the deviation condition, and constructing a running cause-effect map of the cement kiln thermal system includes:

[0143] Based on the historical running data, the historical running optimal value and the optimal running state are determined, the deviation value between the diagnosis index and the historical running optimal value is calculated, the diagnosis index with a deviation value greater than a deviation threshold is selected, and a candidate set of abnormal evaluation variables is obtained; the candidate set of abnormal evaluation variables is detected by using correlation analysis and support vector description technology, and the thermal loss abnormality of the cement kiln thermal system is evaluated according to the detection result and enhanced local linear embedding technology; the reference sample is selected according to the candidate set of abnormal evaluation variables to establish an evaluation original matrix, and the distance between the evaluation original matrix and the optimal running state is judged, and the running state of the cement kiln thermal system is evaluated based on the distance result; the key variables affecting the thermal system state are determined based on the candidate set of abnormal evaluation variables, the thermal loss abnormality and the running state of the cement kiln thermal system, and the running cause-effect map of the cement kiln thermal system is constructed according to the causal relationship between the key variables.

[0144] In the abnormality evaluation variable candidate set is detected by correlation analysis and support vector description technology, and the thermal damage abnormality of the cement kiln thermal worker is evaluated by the detection result and enhanced local linear embedding technology, two groups of random vectors can be generated based on the abnormality evaluation variable candidate set, and the feature decomposition result of the abnormality evaluation variable candidate set is obtained by taking the maximum correlation coefficient between the two groups of random vectors as the target; the feature vector corresponding to the feature decomposition value meeting the preset requirement of the feature decomposition result is selected as the projection vector, and the optimal typical variable set is screened from the abnormality evaluation variable candidate set according to the projection vector; the minimum hypersphere is constructed according to the optimal typical variable set, the distance between the abnormality evaluation variable candidate set and the center of the minimum hypersphere is obtained, and the preliminary thermal damage abnormality set corresponding to the abnormality evaluation variable candidate set is obtained according to the distance result and the minimum hypersphere radius; the projection matrix is generated by combining the preliminary thermal damage abnormality set and the enhanced local linear embedding technology, and the direction vector representing the thermal damage state is obtained by using the projection matrix, so as to evaluate the thermal damage abnormality of the cement kiln thermal worker.

[0145] Specifically, the calculation formula of the minimum hypersphere radius is:

[0146]

[0147] In the formula, R represents the minimum hypersphere radius, a b represents the projection vector, a c represents the cth optimal typical variable set, D AB represents the abnormality evaluation variable candidate set; d c represents the optimal typical variable set a c corresponding Lagrange multiplier, a e represents the e th optimal typical variable set, d e represents the optimal typical variable set a e corresponding Lagrange multiplier.

[0148] Specifically, in the process of obtaining the preliminary thermal damage abnormality set, first, the abnormality evaluation variable candidate set is standardized to eliminate the dimension influence, assuming that the candidate set contains m variables and n observation data, forming an n×m dimensional data matrix X, two projection vectors A and B (dimension m×1) are randomly initialized, and the vector direction is adjusted by an optimization algorithm (such as gradient ascent) to make the correlation coefficient ρ(A T X, B T X) reach the maximum. When the correlation coefficient converges, the covariance matrix X T X is decomposed, and the feature vector with a feature value greater than a preset threshold (such as 0.7) is selected as the projection base.

[0149] The selected k eigenvectors are combined to form a projection matrix W (m x k), and a low-dimensional representation is obtained by linear transformation Z = XW. The contribution of each variable is calculated, and the first p variables (p < k) whose contribution sum exceeds 85% are retained to form the optimal variable set V typical For example, in a certain cement kiln case, 3 key variables are retained after screening from the initial 7 variables: flue gas temperature, secondary air volume, and coal fineness.

[0150] In the process of minimum hypersphere modeling, the minimum hypersphere containing all normal samples is constructed in V typical space by solving the optimization problem: min R 2 +∑ξi.s.t.||Φ(vi)-a||2≤R 2 +ξi, ξi ≥ 0, where a is the center of the sphere, R is the radius, Φ is the kernel function (commonly used Gaussian kernel), and ξi is the "relaxation" amount between hyperspheres. Assuming R = 1.5 and the center coordinates a = (0.32, 0.51, 0.42).

[0151] To determine the thermal damage anomaly, the distance from the sample v to the center of the sphere is calculated, and an abnormal threshold λR (such as λ = 1.2) is set. When d > 1.8, it is determined to be abnormal. The samples outside the hypersphere are grouped into a preliminary thermal damage anomaly set D abnormal , and the degree of deviation δ = (d-R) / R is recorded. For example, if an abnormal sample d = 2.1, then δ = 0.4, indicating that the working condition deviates from the normal range by 40%.

[0152] Specifically, when evaluating the thermal damage anomaly of a cement kiln thermal worker by combining the preliminary thermal damage anomaly set with the enhanced local linear embedding technique to generate a projection matrix and using the projection matrix to obtain a direction vector representing the thermal damage state, a number of test samples can be randomly generated based on the preliminary thermal damage anomaly set, and a set of neighboring points for each test sample can be obtained. The connection weights between the test sample and the set of neighboring points are determined to obtain a weight matrix. Projection analysis is performed on the weight matrix to obtain a projection matrix in two-dimensional space, and the topological structure of the projection matrix is determined. According to the topological structure, a direction vector is obtained, which is composed of the projection matrix and the center point of the preliminary thermal damage anomaly set, to generate a test direction vector representing the thermal damage state. The optimal operating state is taken as a standard direction vector representing the ideal state, and the angle between the test direction vector and the standard direction vector is determined as the thermal damage anomaly value of the cement kiln thermal worker.

[0153] In the process of establishing the evaluation original matrix according to the reference samples selected from the abnormal evaluation variable candidate set respectively, and judging the distance between the optimal operation state, the key point reference samples and the regional reference samples can be selected from the abnormal evaluation variable candidate set according to the key point selection requirement and the partition requirement, the key point evaluation matrix and the regional evaluation matrix are constructed to obtain the evaluation original matrix; the reference samples in the key point evaluation matrix and the regional evaluation matrix are respectively subjected to fuzzy processing, and the key point evaluation membership matrix and the regional evaluation membership matrix are obtained according to the processing results; two groups of evaluation sequences with lengths meeting the preset requirement are defined from the key point evaluation membership matrix and the regional evaluation membership matrix respectively, and two groups of standard sequences corresponding to the optimal operation state are selected according to the key point selection requirement and the partition requirement; the distance matrix is constructed according to the evaluation sequence and the standard sequence, and the distance between the abnormal evaluation variable candidate set and the optimal operation state is judged; if the distance between the evaluation sequence and the standard sequence is less than a certain preset threshold, it indicates that the current operation state is ideal, and if the distance between the evaluation sequence and the standard sequence is greater than or equal to the threshold, it indicates that the current operation state is abnormal.

[0154] Specifically, according to the operation characteristics of the cement kiln, the variables that have greater influence on the thermal state are determined, and for the candidate abnormal evaluation variables (such as flue gas temperature, coal fineness, secondary air volume, etc.), the following two standards are used for screening: the key point selection requirement includes selecting the variables that have the most significant influence on the thermal state, such as the variables highly related to thermal efficiency, fuel consumption, etc.; the partition requirement includes dividing the variables into different regions according to the process flow or physical region, such as the kiln head, kiln middle, and kiln tail regions; based on these requirements, the key point reference samples are selected from the candidate abnormal evaluation variables, such as selecting the variables closely related to thermal efficiency, such as kiln head temperature and coal fineness; the regional reference samples are selected, such as selecting the variables related to flue gas emission, such as kiln tail flue gas composition and flue gas temperature.

[0155] Suppose the candidate set contains multiple variables, the key point and regional reference samples are selected according to the above-mentioned standards, which are respectively used to construct the evaluation matrix, and the selected reference samples are used to construct two evaluation matrices, wherein the key point evaluation matrix is formed by the observation data of all key point reference samples, such as extracting the data of kiln head temperature and coal fineness from the data of 5 time periods; the regional evaluation matrix: similarly, the observation data of the regional reference samples are formed into a matrix to record the observation values of kiln tail flue gas composition and flue gas temperature. These two matrices contain data of different dimensions and can reflect the operation state of the key points and regions.

[0156] The key point evaluation matrix and the area evaluation matrix are combined to obtain a comprehensive evaluation original matrix, so as to cover all selected reference sample data, and the values in the evaluation matrix are subjected to fuzzy processing. The purpose of the fuzzy processing is to convert each variable value into a membership degree, indicating the degree to which the sample belongs to a certain state. In this way, a fuzzy membership matrix of the variable can be obtained, reflecting the closeness between the sample and the ideal state. For each reference sample, a membership value is assigned to indicate the similarity to the ideal state. The larger the value, the closer to the ideal state.

[0157] The evaluation sequence is extracted from the fuzzy membership matrix, each evaluation sequence containing the membership value of each reference sample. The length of the evaluation sequence is predetermined, and each sequence represents the state of the sample in a certain time period. For example, the key point and area evaluation sequences are extracted from the key point evaluation membership matrix and the area evaluation membership matrix, respectively. These sequences can show the state of the system at different time points. The standard sequence is a reference sequence based on the historical optimal running state, which represents the performance of the cement kiln in the ideal or optimal state. The standard sequence can be set based on the average or optimal value of the historical data, representing the standard performance of the system in the ideal state.

[0158] By calculating the distance between the evaluation sequence and the standard sequence, the deviation between the current state and the ideal state can be quantified. The distance is usually calculated based on the Euclidean distance (or other appropriate measurement methods). The distance value between each evaluation sequence and the standard sequence represents the difference between the current running state and the ideal state. The smaller the distance, the closer the current state is to the ideal state. According to the distance matrix, the running state of the cement kiln can be determined:

[0159] If the distance between the evaluation sequence and the standard sequence is less than a predetermined threshold, it indicates that the current running state is ideal. If the distance between the evaluation sequence and the standard sequence is greater than or equal to the threshold, it indicates that the current running state is abnormal. Thus, the running state of the cement kiln can be evaluated, and appropriate adjustment measures can be taken according to the abnormal situation.

[0160] Step S4, based on the running causal graph and the target constraint condition, the development trend of the cement kiln thermal process is predicted, and a dynamic optimization model is constructed combining model predictive control technology to obtain the self-adaptive combustion control result of the cement kiln.

[0161] In one embodiment, based on the running causal graph and the target constraint condition, the development trend of the cement kiln thermal process is predicted, and a dynamic optimization model is constructed combining model predictive control technology to obtain the self-adaptive combustion control result of the cement kiln, including:

[0162] According to the objective function and constraint condition of the cement kiln thermal control, a target constraint condition is set, and a development characteristic vector is constructed by extracting the trend elements affecting the development of the cement kiln thermal from the running causal graph. The development characteristic vector is associated and clustered by using a semantic decision tree to generate a trend target. The motion trend of the trend target is predicted in combination with the target constraint condition, and is used as the development trend of the cement kiln thermal. A mathematical model for self-adaptive combustion control of the cement kiln is established, and feedback compensation is performed on the combustion control of the cement kiln in combination with the optimal running state to output the optimal combustion control result of the cement kiln.

[0163] Specifically, when the development characteristic vector is associated and clustered by using the semantic decision tree to generate the trend target, and the motion trend of the trend target is predicted in combination with the target constraint condition, the development characteristic vector is determined according to the information to which the development characteristic vector belongs, the classification of the semantic decision tree node is determined, the development characteristic vector is placed in a two-dimensional data table based on the classification result to describe the relationship view of the development characteristic vector, the development characteristic vector is divided into a dynamic set based on the relationship view result, the node concept in the relationship view is associated, and the root node of the semantic decision tree is selected from the node concept as the trend target. The implicit state of the development characteristic vector is determined according to the trend target, the motion trend of the trend target is converted into the motion state prediction of the trend target by using the implicit state, and the observation object is represented based on the real spatial position corresponding to the trend target. The trend probability distribution of the observation object is analyzed with the target constraint condition as the boundary, and the associated region is constructed in the form of a sector wave door based on the analysis result to predict the motion trend of the trend target as the development trend of the cement kiln thermal.

[0164] It should be explained that in the process of obtaining the development trend of the cement kiln thermal, each development characteristic vector is mapped to the corresponding classification node through the construction of the semantic decision tree. The semantic decision tree is a tree structure constructed by analyzing the internal relationship and characteristics of the data. Each node represents a feature or attribute, and its child nodes represent different values or classifications of the feature.

[0165] Suppose there are the following characteristic vectors: coal fineness, fuel input, kiln temperature, etc. According to the distribution and characteristic relationship of the data, these vectors are classified into different nodes when the semantic decision tree is constructed, for example, node 1 is coal fineness (high / low); node 2 is fuel input (normal / excessive / insufficient); node 3 is kiln temperature (normal / high / low). Each development characteristic vector is placed in a two-dimensional data table according to its category based on the classification result of the semantic decision tree. This data table is used to describe the relationship view between the development characteristic vectors, as shown in Table 5:

[0166] Table 5: Relationship table between development characteristic vectors

[0167] Features Classification 1 (high / low) Classification 2 (normal / excessive / insufficient) Classification 3 (normal / higher / lower) Coal fineness High Fuel input Normal Kiln temperature Normal

[0168] On the basis of the above two-dimensional data table, the development feature vector is divided into multiple dynamic sets according to the relationship view, and the dynamic set reflects the change trend of the feature vector under different classification conditions. Specifically, according to the relationship between the classification results and the features, the feature vector is divided into multiple sets, for example, set 1 is that the coal powder fineness is high and the fuel input is normal, and the kiln temperature is normal; set 2 is that the coal powder fineness is low and the fuel input is excessive, and the kiln temperature is high; set 3 is that the coal powder fineness is high and the fuel input is insufficient, and the kiln temperature is low; according to the dynamic set and the node concept in the relationship view, a semantic decision tree root node is selected as a trend target, and the trend target is a key node describing the change of the target state, representing the starting point of the future trend prediction, for example, if the coal powder fineness and the fuel input are the key factors affecting the thermal development trend of the cement kiln, the corresponding decision tree root node (such as "coal powder fineness") can be selected as the trend target.

[0169] The implicit state reflects the change trend or state of the development feature vector in the system, for example, implicit state 1 is that the coal powder fineness is high, the fuel input is normal, and the kiln temperature is stable; implicit state 2 is that the coal powder fineness is low, the fuel input is excessive, and the kiln temperature is high; the implicit state is determined by analyzing the historical data, the current data and the change trend thereof, which can help predict the future movement trend. Once the trend target and the implicit state are determined, the change trend of the trend target is converted into an actual movement state prediction by using the historical trend and the current implicit state, which can be realized by time series analysis, regression model or machine learning algorithm. Assuming that the trend target is "coal powder fineness", the change of the future coal powder fineness can be predicted by trend analysis of the historical data, and the change is converted into a specific numerical value to represent the expected state of the coal powder fineness in the future few hours.

[0170] The target constraint condition defines the boundary of the change of the trend target, for example, some operations limit the upper and lower limits of the coal powder fineness, or the kiln temperature has a certain safety range. Therefore, these constraint conditions need to be analyzed to ensure that the predicted trend is within an acceptable range. In combination with the target constraint condition, the probability distribution of the movement trend of the trend target is analyzed, and by using a model (such as Monte Carlo simulation, Bayesian network, etc.), the probability distribution of the possible change of the target in the future period of time can be obtained. According to the analyzed trend probability distribution, a correlation area is constructed. The correlation area is defined by a specific geometric shape (such as a sector wave door form) to represent the area in which the trend target can move in a specific time, for example, assuming that the coal powder fineness has a 60% probability of falling within a certain range in the future few hours, then an expected change range can be constructed by using this probability information.

[0171] Further, the development trend prediction of the cement kiln thermal system is obtained, which includes the change trend of key variables such as coal fineness and fuel input in the future; whether the predicted change state meets the target constraint condition (such as temperature, coal fineness limit, etc.); finally, based on the trend target and the prediction result of the spatial position, decision support can be provided for the operation optimization of the cement kiln, thereby improving the thermal efficiency and production stability of the kiln.

[0172] In the process of establishing a mathematical model for adaptive combustion control of a cement kiln, feedback compensation is performed on the combustion control of the cement kiln in combination with the optimal operating state, and the optimal combustion control result of the cement kiln is output. Based on the control process of the cement kiln, the composition structure of the combustion control can be determined, and the structure of the combustion control system of the cement kiln can be determined according to the working principle corresponding to the composition structure, and the performance loss of each composition structure in the control process is analyzed. The performance loss is used as a constraint condition to establish a mathematical model for adaptive combustion control of a cement kiln, and is converted into a state space equation, which is processed in combination with the development trend and model predictive control technology. According to the processing result, the optimal controller of the adaptive combustion control of the cement kiln is obtained, and feedback compensation is performed on the combustion control of the cement kiln based on the optimal controller and the optimal operating state, and the optimal combustion control result of the cement kiln is output.

[0173] It needs to be explained that when the optimal combustion control result is output, the specific control process of the cement kiln combustion system needs to be understood first. The combustion control system includes but is not limited to: fuel input subsystem (such as coal powder, oil gas), air supply subsystem (such as primary air, secondary air), temperature detection and feedback subsystem, control execution mechanism (such as valve adjustment, motor drive) and the like. These subsystems together constitute the entire combustion control system, and the data interaction and physical action between them directly affect the thermal efficiency and stability. For each subsystem, the working principle and interaction need to be sorted out: the fuel supply system adjusts the heat supply, the air system adjusts the combustion efficiency, the feedback system monitors the temperature and smoke composition parameters in real time, and the execution system adjusts the proportion of fuel and air dynamically to achieve combustion balance.

[0174] In actual control, each subsystem has problems such as response lag, transmission delay, energy loss, etc., which need to be identified and quantified one by one: for example, the valve adjustment lags 1-2 seconds, the fan has energy transmission efficiency decline, the fuel combustion process is not complete, etc. These performance losses will be used as constraint conditions for subsequent mathematical models to ensure that the model is closer to the actual working condition, and then a mathematical model for describing the combustion control of the cement kiln is established. The core elements include system input fuel flow, air volume and other control variables; system output temperature, oxygen content in flue gas and other observation variables; internal state temperature distribution, pressure fluctuation, air-fuel ratio change, etc.; the constraint condition is determined by the performance loss, such as the maximum air volume, the minimum fuel passage response time, etc.

[0175] In order to realize system control, the mathematical model of the previous step needs to be converted into a state space form. The state space model takes the "state variable" of the system as the core and can dynamically describe the evolution process of the system under the action of different inputs. The mapping relationship between the state and the output is described in sequence with time, continuously or discretely. Combined with model predictive control (MPC) technology, optimization decision is made.

[0176] The historical operation data is introduced to model the trend of system variables (such as temperature and fuel consumption) for predicting future operation state. The trend model can help to predict potential deviations in advance. MPC is a control strategy based on model prediction of future behavior and rolling optimization. Its core is to predict the state evolution path in a certain period of time at each control time. Based on the optimization target (such as minimum fuel consumption and maximum temperature stability), the calculation is carried out. Considering the system constraints and performance loss, the optimal control quantity sequence is solved. Only the current control output is executed, and the subsequent prediction and correction are continued to generate a control strategy that can respond to dynamic changes in real time.

[0177] On the basis of the above optimization, combined with the development trend and constraint conditions, a controller that can adapt to the dynamic changes of the system is designed. The controller has learning and self-adjusting ability and can adjust the strategy according to the feedback. It can continuously output effective control signals under different working conditions (such as raw material change and temperature fluctuation). The closed-loop control process of "control-feedback-optimization-update" is realized. Finally, based on the deviation between the optimal controller and the optimal operation state of the cement kiln, feedback compensation is realized. The current operation state is compared with the optimal state to evaluate the gap. Intelligent compensation adjustment is carried out by the controller, such as fine-tuning of fuel ratio and air volume intensity. Real-time adjustment of control instructions is carried out to output the current optimal combustion control parameters. The controller is automatically updated to ensure that the operation continuously approaches the optimal state.

[0178] As shown in Figure 2 According to another embodiment of the present application, a cement kiln thermal diagnosis analysis system based on heat balance calculation is also proposed. The analysis system comprises:

[0179] A balance relationship acquisition module 1 is used to acquire the heat balance state corresponding to the cement kiln thermal parameter and material composition data calculation, and analyze the material balance relationship between the input and output of the cement kiln material according to the law of conservation of mass;

[0180] A diagnosis index generation module 2 is used to take the ratio between the clinker formation heat of the cement kiln and the total input heat as a thermal efficiency index, evaluate the energy utilization efficiency of the burning system of the cement kiln, and generate a diagnosis index in combination with the heat balance state and the material balance relationship;

[0181] A cause-effect map construction module 3 is configured to compare the diagnostic indexes with the historical optimal values to determine deviation conditions, evaluate thermal loss abnormalities and operation states of the cement kiln thermal system according to the deviation conditions, and construct a cause-effect map of the cement kiln thermal system operation;

[0182] A dynamic optimization control module 4 is configured to predict the development trend of the cement kiln thermal system based on the operation cause-effect map and target constraint conditions, construct a dynamic optimization model in combination with a model predictive control technology, and obtain a self-adaptive combustion control result of the cement kiln.

[0183] The above merely describes the preferred embodiments of the present application and is not intended to limit the present application. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present application shall be included in the protection scope of the present application.

Claims

1. A method for thermal diagnostic analysis of cement kilns based on heat balance calculations, characterized in that, The analytical method includes: The thermal parameters and material composition data of the cement kiln are obtained to calculate the corresponding thermal balance state, and the material balance relationship between the material input and material output of the cement kiln is analyzed based on the law of conservation of mass. The ratio between the clinker formation heat and the total input heat in the cement kiln is used as a thermal efficiency index to evaluate the energy utilization efficiency of the cement kiln firing system, and diagnostic indicators are generated by combining the thermal balance state and the material balance relationship. The diagnostic indicators are compared with the historical best values ​​to determine the deviation. Based on the deviation, the abnormal heat loss and operating status of the cement kiln are assessed, and a cause-and-effect diagram of the cement kiln's thermal operation is constructed. Based on the objective function and constraints of cement kiln thermal control, target constraints are set, and dynamic elements affecting the development of cement kiln thermal performance are extracted from the causal graph to construct a development feature vector. The classification of semantic decision tree nodes is determined based on the information belonging to the development feature vectors. Based on the classification results, the development feature vectors are placed into a two-dimensional data table to describe the relationship view of the development feature vectors. Based on the relationship view results, the development feature vectors are divided into dynamic sets, and the node concepts in the relationship view are associated. The root node of the semantic decision tree is selected from the node concepts as the dynamic target. The implicit state of the development feature vector is determined based on the dynamic target. The implicit state is used to convert the dynamic target's motion trend into a motion state prediction, and the observed object is represented based on the corresponding real spatial location of the dynamic target. The trend probability distribution of the observed object is analyzed using the target constraints as boundaries, and an associated region is constructed using a fan-shaped gate based on the analysis results to predict the dynamic target's motion trend as the development trend of cement kiln thermal performance. A mathematical model for adaptive combustion control of cement kilns is established, and combined with the optimal operating state, feedback compensation is performed on the cement kiln combustion control to output the optimal combustion control result of the cement kiln.

2. The method for thermal diagnosis and analysis of cement kilns based on heat balance calculation according to claim 1, characterized in that, The process of comparing diagnostic indicators with historical best operating values ​​to determine deviations, assessing abnormal heat loss and operating status of the cement kiln based on deviations, and constructing a cause-and-effect graph of cement kiln thermal operation includes: Based on historical operating data, determine the historical optimal operating value and optimal operating state, calculate the deviation between the diagnostic indicators and the historical optimal operating value, select the diagnostic indicators with deviation values ​​greater than the deviation threshold, and obtain the corresponding candidate set of abnormal assessment variables. Anomaly detection was performed on the candidate set of anomaly assessment variables using correlation analysis and support vector description techniques. Based on the detection results and enhanced local linear embedding techniques, the thermal loss anomalies of cement kiln thermal performance were evaluated. Based on the candidate set of abnormal assessment variables, reference samples are selected to establish the original assessment matrix, and the distance between the matrix and the optimal operating state is determined. The operating state of the cement kiln thermal system is then assessed based on the distance results. Based on the candidate set of abnormal assessment variables, the abnormal heat loss of cement kiln thermal system and operating status, the key variables affecting the thermal state are identified, and a causal graph of cement kiln thermal operation is constructed according to the causal relationship between the key variables.

3. The method for thermal diagnosis and analysis of cement kilns based on heat balance calculation according to claim 2, characterized in that, The process of using correlation analysis and support vector description techniques to detect anomalies in the candidate set of anomaly assessment variables, and then evaluating the thermal loss anomalies in cement kilns based on the detection results and enhanced local linear embedding techniques, includes: Two sets of random vectors are randomly generated based on the candidate set of anomaly assessment variables, and the feature decomposition results of the candidate set of anomaly assessment variables are obtained with the objective of maximizing the correlation coefficient between the two sets of random vectors. The feature vectors corresponding to the feature decomposition values ​​that meet the preset requirements are selected as projection vectors, and the optimal set of typical variables is selected from the candidate set of anomaly evaluation variables based on the projection vectors. Construct a minimum hypersphere based on the optimal set of typical variables, obtain the distance between the candidate set of anomaly assessment variables and the center of the minimum hypersphere, and obtain the preliminary heat loss anomaly set corresponding to the candidate set of anomaly assessment variables based on the distance result and the radius of the minimum hypersphere. The preliminary heat loss anomaly set is combined with the enhanced local linear embedding technique to generate a projection matrix, and the projection matrix is ​​used to obtain the direction vector characterizing the heat loss state to evaluate the heat loss anomaly of cement kiln thermal operation.

4. The method for thermal diagnosis and analysis of cement kilns based on heat balance calculation according to claim 3, characterized in that, The formula for calculating the minimum hypersphere radius is as follows: ； In the formula, R represents the minimum hypersphere radius, and a b Let a represent the projection vector. c Let D represent the c-th optimal canonical variable set. AB d represents the candidate set of variables for anomaly evaluation; c Represents the optimal canonical variable set a c The corresponding Lagrange multiplier, a e Let d represent the e-th optimal canonical variable set. e Represents the optimal canonical variable set a e The corresponding Lagrange multipliers.

5. The method for thermal diagnosis and analysis of cement kilns based on heat balance calculation according to claim 4, characterized in that, The process of combining the initial heat loss anomaly set with enhanced local linear embedding technology to generate a projection matrix, and using the projection matrix to obtain the direction vector characterizing the heat loss state, to evaluate the heat loss anomalies in the thermal operation of the cement kiln includes: Several test samples are randomly generated based on the initial thermal loss anomaly set, and the set of nearest points for each test sample is obtained. The connection weight between the test sample and the set of nearest points is determined to obtain the weight matrix. Projection analysis is performed on the weight matrix to obtain the projection matrix in two-dimensional space, and the topological structure of the projection matrix is ​​determined. Based on the topological structure, the direction vector formed by the projection matrix and the center point of the preliminary thermal loss anomaly set is obtained, and a test direction vector characterizing the thermal loss state is generated. The optimal operating state is used as the standard direction vector to characterize the ideal state, and the angle between the test direction vector and the standard direction vector is determined as the heat loss anomaly value for evaluating the thermal performance of the cement kiln.

6. The method for thermal diagnosis and analysis of cement kilns based on heat balance calculation according to claim 5, characterized in that, The process of selecting reference samples from the candidate set of abnormal evaluation variables to establish the original evaluation matrix, determining the distance between the matrix and the optimal operating state, and evaluating the thermal operating state of the cement kiln based on the distance results includes: Based on the key point selection requirements and zoning requirements, key point reference samples and regional reference samples are selected from the candidate set of abnormal evaluation variables to construct the key point evaluation matrix and regional evaluation matrix to obtain the original evaluation matrix. The reference samples in the key point evaluation matrix and the regional evaluation matrix are fuzzed respectively, and the key point evaluation membership matrix and the regional evaluation membership matrix are obtained based on the processing results. Two sets of evaluation sequences with the defined lengths that meet the preset requirements are evaluated from the membership matrix of key points and the membership matrix of regions, respectively. At the same time, two sets of standard sequences corresponding to the optimal operating state are selected according to the key point selection requirements and the zoning requirements. Based on the evaluation sequence and the standard sequence, construct distance matrices respectively to determine the distance between the candidate set of abnormal evaluation variables and the optimal operating state; If the distance between the evaluation sequence and the standard sequence is less than a preset threshold, it indicates that the current operating state is ideal. If the distance between the evaluation sequence and the standard sequence is greater than or equal to the threshold, it indicates that the current operating state is abnormal.

7. The method for thermal diagnosis and analysis of cement kilns based on heat balance calculation according to claim 6, characterized in that, The mathematical model established for adaptive combustion control of cement kilns, combined with the optimal operating state, provides feedback compensation for the combustion control of cement kilns, and outputs the optimal combustion control results of the cement kiln, including: The composition structure of combustion control is determined based on the control process of cement kiln, and the structure of cement kiln combustion control system is determined according to the working principle corresponding to the composition structure. The performance loss of each composition structure during the control process is analyzed. A mathematical model for adaptive combustion control of cement kilns is established using performance loss as a constraint, and then converted into a state-space equation. The state-space equation is then processed using development trend-based and model predictive control techniques. The optimal controller for adaptive combustion control of cement kiln is obtained based on the processing results. Then, feedback compensation is performed on the combustion control of cement kiln based on the optimal controller and the optimal operating state, and the optimal combustion control result of cement kiln is output.

8. A cement kiln thermal diagnostic analysis system based on heat balance calculation, used to implement the cement kiln thermal diagnostic analysis method based on heat balance calculation as described in any one of claims 1-7, characterized in that, The analysis system includes: The equilibrium relationship acquisition module is used to acquire the thermal parameters and material composition data of the cement kiln to calculate the corresponding thermal equilibrium state, and to analyze the material balance relationship between the material input and material output of the cement kiln based on the law of conservation of mass. The diagnostic index generation module is used to use the ratio between the clinker formation heat and the total input heat in the cement kiln as a thermal efficiency index to evaluate the energy utilization efficiency of the cement kiln firing system and generate diagnostic indexes by combining the thermal balance state and material balance relationship. The causal graph construction module is used to compare diagnostic indicators with historical best values ​​to determine deviations, assess the thermal loss anomalies and operating status of cement kilns based on the deviations, and construct a causal graph of cement kiln thermal operation. The dynamic optimization control module is used to predict the development trend of cement kiln thermal performance based on the operating causal graph and target constraints, and to construct a dynamic optimization model by combining model predictive control technology to obtain the adaptive combustion control results of cement kiln.

Images