Real-time monitoring and automatic blocking system for hot-blast stove sparks

Abstract

The present application relates to hot blast stove combustion safety technical field, especially relates to hot blast stove spark real-time monitoring and automatic blocking system, including: spark monitoring module, the spark monitoring module is configured to be able to gather spark state information, and extract spark state feature from state information;Combustion monitoring module, the combustion monitoring module is configured to be able to obtain fuel combustion state data and fuel information;Safety control module, the safety control module is configured to be able to obtain data from spark monitoring module and combustion monitoring module, and based on spark state feature back-pushing combustion state anomaly, establish the mapping relationship of spark state feature and combustion anomaly;Self-blocking control module, the self-blocking control module is configured to be able to according to the mapping relationship of spark state feature and combustion anomaly, utilize hierarchical blocking strategy, realize hierarchical blocking, improve system security.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of hot blast stove combustion safety, in particular to a hot blast stove spark real-time monitoring and automatic blocking system. BACKGROUND

[0002] Hot blast stove combustion safety monitoring is a core link of industrial safety production, mainly used for preventing combustion out of control, explosion, backfire, flameout and fire accidents. Spark monitoring is an important part of combustion safety monitoring, because the spark appearing in high-temperature gas can directly reflect the state of combustion, which can be used as a key indicator to judge the combustion efficiency, combustion stability and safety. In the field of grain drying, hot blast stove spark monitoring and blocking play a very important role in preventing sparks from being discharged to cause grain fires.

[0003] The existing technology mainly uses the following methods to monitor sparks: tracking the temperature field distribution of sparks by an infrared thermal imager; obtaining the movement trajectory of sparks by a high-speed camera; and monitoring the concentration of sparks by laser scattering particulate matter.

[0004] After obtaining various state parameters of sparks, how to determine whether combustion abnormalities occur according to the state parameters of sparks and the characteristics of fuel, and take measures to ensure combustion safety, and how to effectively block to avoid fires, are problems that need to be solved urgently. SUMMARY

[0005] The present application establishes a mapping relationship between spark state characteristics and combustion state abnormalities by real-time monitoring of the state of sparks, judges whether the combustion state of the hot blast stove is abnormal, and takes corresponding measures to ensure combustion safety.

[0006] The technical solution of the present application is: a hot blast stove spark real-time monitoring and automatic blocking system, comprising:

[0007] a spark monitoring module, which is configured to be able to collect spark state information and extract spark state characteristics from the state information;

[0008] a combustion monitoring module, which is configured to be able to obtain fuel combustion state data and fuel information;

[0009] a safety control module, which is configured to be able to obtain data from the spark monitoring module and the combustion monitoring module, and to deduce combustion state abnormalities based on the spark state characteristics, and to establish a mapping relationship between the spark state characteristics and the combustion abnormalities;

[0010] a self-blocking control module, which is configured to be able to use a hierarchical blocking strategy to realize hierarchical blocking and improve system safety according to the mapping relationship between the spark state characteristics and the combustion abnormalities.

[0011] Preferably, the spark monitoring module is configured to collect spark state information and extract spark state features from the state information, including:

[0012] Obtain spark temperature information through infrared imaging equipment and temperature sensors;

[0013] Capture spark trajectory information through high-speed cameras;

[0014] Obtain spark concentration information online through laser scattering particle counters;

[0015] Obtain spark carbon content information through Raman spectrometers;

[0016] Extract spark surface temperature features from spark temperature information, extract spark movement speed features from spark trajectory information, and extract spark number concentration features from spark concentration information.

[0017] Preferably, the spark state feature-based combustion state abnormality is reversed, and a mapping relationship between the spark state feature and the combustion abnormality is established, including:

[0018] Conduct quantitative diagnosis of combustion abnormalities, including the following steps:

[0019] Establish a combustion abnormality index :

[0020] ;

[0021] Wherein, ( t ) represents the spark surface temperature, represents the combustion temperature reference value, ( t ) represents the spark movement speed; represents the design flue gas flow rate; ( t ) represents the spark concentration; represents the spark concentration threshold value; represents the 50% cumulative particle size of the spark, i.e. the median particle size; represents the particle size threshold value; ( t ) represents the carbon content of the spark;

[0022] If , the combustion is normal;

[0023] If , the combustion is abnormal.

[0024] Preferably, after judging the combustion abnormality, combustion abnormality type identification is performed to establish a mapping relationship between the spark state feature and the combustion abnormality, including:

[0025] Constructing Mars feature change combination sequence one: ;

[0026] If and , judging the combustion abnormal type as oxygen deficiency combustion; represents the collection time interval;

[0027] Constructing Mars feature change combination sequence two: ;

[0028] If and , judging the combustion abnormal type as local over-temperature;

[0029] Constructing Mars feature change combination sequence three: ;

[0030] If and , judging the combustion abnormal type as unreasonable air distribution;

[0031] Constructing Mars feature change combination sequence four: ;

[0032] If and , judging the combustion abnormal type as fuel abnormality.

[0033] Preferably, the combustion state abnormality based on the spark state feature is established, and the mapping relationship between the spark state feature and the combustion abnormality further comprises:

[0034] The phase space of the Mars surface temperature is reconstructed by using Takens embedding theorem:

[0035] ]; wherein, represents the embedding dimension, ' represents the delay time; ; wherein, represents the data collection period;

[0036] The recurrence plot is constructed, and the recurrence quantitative analysis is performed, comprising:

[0037] The state vectors in the phase space are compared , and a binary matrix, i.e. the recurrence plot, is generated;

[0038] The spark state features in the data collection period are obtained, the spark state feature time sequence is constructed, and the long-range correlation of the spark state feature time sequence is described by using the Hurst index .

[0039] Preferably, the state vector in the phase space is obtained by Comparing, a binary matrix, i.e. recurrence plot, is generated, including:

[0040] Calculating recurrence rate to quantify the spark temperature state repetition probability, reflecting the combustion stability;

[0041] ; wherein the similarity threshold ; represents the Heaviside step function;

[0042] Calculating the determinacy to show whether there is periodic oscillation in the phase space;

[0043] ; wherein, represents any element value in the recurrence plot, represents the diagonal length in the recurrence plot, represents the number of line segments with diagonal length in the recurrence plot;

[0044] The spark state feature in the data acquisition period is obtained, the spark state feature time series is constructed, and the long-range correlation of the spark state feature time series is described by the Hurst index , including:

[0045] ; wherein, represents the sequence range of the spark state feature time series; represents the standard deviation; represents the time window length; ; wherein, represents the time window adjustment amount;

[0046] The spark state nonlinear mapping is established by using a hybrid kernel support vector machine SVR, including:

[0047] A nonlinear term is added to the original combustion anomaly index, i.e.:

[0048] ;

[0049] The kernel function of the SVR is:

[0050] ; wherein, represents the Gaussian kernel weight, represents the Gaussian kernel bandwidth, represents the polynomial order; represents the SVR input variable;

[0051] By self-attention mechanism, the weight is adaptively adjusted, that is:

[0052] , =1, 2, 3, 4, 5; wherein, represents a sharpening coefficient; represents the sensitivity of the output to the input parameter.

[0053] Preferably, according to the mapping relationship between the spark state characteristics and the combustion abnormality, a hierarchical blocking strategy is used to realize hierarchical blocking and improve system safety, including:

[0054] The combustion risk level is evaluated, and the comprehensive combustion risk index is set as ; wherein, represents a safe flow rate, represents a step function; represents the slagging risk value and the volatile matter unburned risk value;

[0055] ; wherein, represents the ratio of SiO2 content and CaO content; represents the spark sphericity, >0.85;

[0056] ; wherein, represents the concentration of sparks with a particle size less than ;

[0057] The combustion risk is classified:

[0058] When , the comprehensive combustion risk is level one; an early warning signal is output; and the blocking measure is to adjust the air-fuel ratio;

[0059] When , the comprehensive combustion risk is level two; and the blocking measure is cyclone interception and water mist interception;

[0060] When , the comprehensive combustion risk is level three; and the blocking measure is to suspend combustion, perform emergency inerting and detonation blocking.

[0061] Preferably, the combustion state abnormality is backstepped based on the spark state characteristics, and the mapping relationship between the spark state characteristics and the combustion abnormality is established, further including:

[0062] The fuel property adaptation layer is introduced in the combustion abnormality index to adapt to the powder fuel properties, including:

[0063] Obtain fuel information, including: 90% powder fuel particle passes through the screen size, that is, the 90% cumulative particle size of fuel particles; fuel perception ash-free volatile matter; ash softening temperature and received base moisture content;

[0064] Construct a fuel characteristic vector ; wherein, represents the 90% cumulative particle size of fuel particles, represents the fuel reference particle size; represents the fuel perception ash-free volatile matter, represents the ash softening temperature, represents the received base moisture content of the fuel;

[0065] Construct a fuel-spark transfer function:

[0066] ;

[0067] Based on the corrected parameters , Calculate the combustion anomaly index to obtain the combustion anomaly index of the fused fuel characteristics ;

[0068] Add a fuel phase to the original phase space to obtain a new phase space:

[0069] ; wherein, represents the fuel characteristic time constant;

[0070] Add fuel similarity measurement to the SVR kernel function to obtain a corrected kernel function:

[0071] ; wherein, represents the Manhattan distance; when the fuel is switched, that is, ; , automatically reduce the weight of historical data, reduce the influence of fuel characteristics on the combustion anomaly index.

[0072] Preferably, according to the mapping relationship between the spark state characteristics and the combustion anomaly, the hierarchical blocking strategy is used to realize hierarchical blocking and improve the system safety, which further comprises:

[0073] Adjust the blocking measure according to the fuel characteristics to improve the blocking effect, which comprises:

[0074] Adjust the water mist concentration according to the fuel perception ash-free volatile matter; the adjusted water mist concentration ; wherein, is the reference water mist concentration;

[0075] Adjust the inert gas concentration according to the fuel particle size; the adjusted inert gas concentration is .

[0076] The application further provides a computer readable storage medium, which stores a computer program, and the computer program is executed by a processor to realize the hot blast stove spark real-time monitoring and automatic blocking system.

[0077] The application has the following beneficial effects:

[0078] 1. According to the application, the abnormal combustion state is deduced based on the monitored spark state characteristic data, and a mapping relationship between the spark state parameter and the combustion abnormality is established, specifically, a combustion abnormality index (essentially combustion entropy increase abnormality degree) is constructed to reflect the degree of deviation of the hot blast stove combustion system from the ideal combustion state.

[0079] 2. According to the application, the phase space is constructed to adapt to the nonlinear characteristics in the combustion process, and the SVR model based on the phase space and the mixed kernel function is used to capture the local mutation of the state through the Gaussian kernel and fit the global change trend of the state through the polynomial, so that the combustion abnormality index adapts to the nonlinear combustion state, more accurately reflects the relationship between the combustion state and the spark state, improves the identification accuracy of the combustion abnormality, reduces the false alarm probability, provides accurate data basis for the automatic blocking, and improves the combustion safety of the hot blast stove. BRIEF DESCRIPTION OF DRAWINGS

[0080] Figure 1 FIG. 1 is a module diagram of the hot blast stove spark real-time monitoring and automatic blocking system according to the application. DETAILED DESCRIPTION

[0081] The following description is provided to disclose the application so that those skilled in the art can implement the application. The preferred embodiments in the following description are only examples, and other obvious modifications can be made by those skilled in the art. The basic principles of the application defined in the following description can be applied to other embodiments, modifications, improvements, equivalents and other technical solutions without departing from the spirit and scope of the application.

[0082] It can be understood that the term “one” should be understood as “at least one” or “one or more”, that is, in one embodiment, the number of one element can be one, and in another embodiment, the number of the element can be multiple, and the term “one” cannot be understood as a limitation on the number.

[0083] Embodiment one:

[0084] Reference Figure 1 The technical solution provided by the application is a hot blast stove spark real-time monitoring and automatic blocking system, which comprises a spark monitoring module, a combustion monitoring module, a safety control module and a self-blocking control module.

[0085] The spark monitoring module is configured to collect spark state information and extract spark state features from the state information.

[0086] The combustion monitoring module is configured to obtain fuel combustion state data and fuel information.

[0087] The safety control module is configured to obtain data from the spark monitoring module and the combustion monitoring module, and to infer combustion state abnormalities based on the spark state features, and to establish a mapping relationship between the spark state features and the combustion abnormalities.

[0088] The self-blocking control module is configured to use a hierarchical blocking strategy to achieve hierarchical blocking and improve system safety based on the mapping relationship between the spark state features and the combustion abnormalities.

[0089] The functions of the spark monitoring module are achieved through the following steps:

[0090] Obtain spark temperature information through infrared imaging equipment and temperature sensors;

[0091] Capture spark trajectory information through a high-speed camera;

[0092] Obtain spark concentration information online through a laser scattering particle counter;

[0093] Obtain spark carbon content information through a Raman spectrometer;

[0094] Extract spark surface temperature features from the spark temperature information, extract spark motion speed features from the spark trajectory information, and extract spark number concentration features from the spark concentration information.

[0095] The inference of combustion state abnormalities based on spark state features and the establishment of a mapping relationship between spark state features and combustion abnormalities can be achieved through the following steps:

[0096] Perform combustion abnormality quantitative diagnosis, including the following steps:

[0097] Establish a combustion abnormality index :

[0098] ;

[0099] Where, t represents the spark surface temperature, represents the combustion temperature reference value, t represents the spark motion speed; represents the design flue gas flow rate; t represents the spark concentration; represents the spark concentration threshold value;​​​ This represents the 50% cumulative particle size of the spark, i.e., the median particle size. Indicates the particle size threshold; ( t () indicates the carbon content of the spark;

[0100] if If so, then combustion is considered normal;

[0101] if If so, it is determined that the combustion is abnormal.

[0102] After identifying abnormalities in the hot blast stove combustion based on the characteristics of the sparks, it is necessary to further determine the type of abnormality and establish a mapping between the characteristics of the sparks and the combustion abnormalities. Specifically:

[0103] Constructing a combined sequence of Martian feature changes: ;

[0104] if and If so, the combustion anomaly type is determined to be oxygen-deficient combustion; Indicates the time interval for data collection;

[0105] Constructing the second combination sequence of Martian feature changes: ;

[0106] if and If so, the combustion anomaly type is determined to be local overheating;

[0107] Constructing the third sequence of combined changes in Martian features: ;

[0108] if and If so, the combustion anomaly is determined to be due to improper air distribution;

[0109] Constructing the fourth combination sequence of Martian feature changes: ;

[0110] if and If so, the combustion anomaly type is determined to be a fuel anomaly.

[0111] For example, the spark characteristic monitoring data of a certain coke oven gas hot blast stove is as follows:

[0112] , ; ;

[0113] ; ; ; ; , , , , , .

[0114] The calculation obtains =0.68; since , it is judged that the combustion is abnormal.

[0115] Again and , it is judged that the air distribution is unreasonable, causing the particulate fuel to be carried away before being burned, forming sparks; causing the spark concentration to increase substantially.

[0116] By increasing the secondary air ratio from 25% to 35%, after 24 hours , the combustion returns to normal, with the value decreasing to 0.18.

[0117] Among them, according to the mapping relationship between the spark state characteristics and the combustion abnormality, the hierarchical blocking strategy is used to realize hierarchical blocking and improve the system safety, which is realized through the following steps:

[0118] The combustion risk level is evaluated, and the comprehensive combustion risk index is set as ; wherein, represents the safe flow rate, represents the step function; represents the risk value of slagging and the risk value of unburned volatile matter;

[0119] ; wherein, represents the ratio of SiO2 content and CaO content; represents the spark sphericity, >0.85;

[0120] ; wherein, represents the concentration of sparks with a particle size less than ;

[0121] The combustion risk is classified:

[0122] When , the comprehensive combustion risk is level one; output an early warning signal; the blocking measure is to adjust the air-fuel ratio;

[0123] When , the comprehensive combustion risk is level two; the blocking measure is cyclone interception and water mist interception;

[0124] When , the comprehensive combustion risk is level three; the blocking measure is to stop combustion, execute emergency inerting and detonation blocking.

[0125] Example two:

[0126] The combustion anomaly index in Example 1 reflects the linear relationship between combustion state and spark state characteristics. When the hot blast stove is under constant load and the fuel-air ratio is stable, the combustion state of the hot blast stove exhibits linear characteristics. However, under turbulent mixing, abrupt changes in chemical reaction kinetics, and thermoacoustic oscillations, the combustion process exhibits strong nonlinearity, making it impossible to construct the combustion anomaly index using the method described in Example 1. A method combining nonlinear dynamics modeling and machine learning is needed to improve the accuracy of the combustion anomaly index and ultimately ensure the combustion safety of the hot blast stove. Therefore, based on Example 1, we propose the following technical solution:

[0127] Based on the characteristics of the Martian state, the anomalies in the combustion state can be inferred, and a mapping relationship between the characteristics of the spark state and the combustion anomalies can be established. This can also be achieved through the following steps:

[0128] The phase space of Martian surface temperature is reconstructed using Takens' embedding theorem:

[0129] ];in, Indicates the embedding dimension. 'Indicates a delay time; ;in, Indicates the data collection period;

[0130] Constructing a recursion graph and performing recursive quantitative analysis includes:

[0131] The state vector in phase space Compare these elements to generate a binary matrix, i.e., a recursion graph; if any element in the recursion graph... 1 ( < ), indicating that the state vectors are similar, otherwise it is 0; including:

[0132] Calculate the recursion rate This is used to quantify the probability of repeating spark temperature states and reflect combustion stability.

[0133] Among them, the similarity threshold ; Represents the Heaviside jump function;

[0134] computational determinism This is to indicate whether periodic oscillations exist in the phase space;

[0135] ;in, Represents any element value in the recursive graph. This represents the length of the diagonal in the recursive graph. The length of the diagonal in the recursive graph is... The number of line segments;

[0136] The Martian state characteristics were acquired during the data acquisition period, a time series of Martian state characteristics was constructed, and the Hurst index was used to analyze them. Long-range correlations describing time series of Martian state characteristics include:

[0137] ;in, The series range representing the time series characteristics of Mars' state; Indicates standard deviation; Indicates the length of the time window; ;in, Indicates the amount of time window adjustment;

[0138] Taking the time series of spark concentration characteristics as an example, if If the state fluctuates randomly, it indicates that combustion is normal;

[0139] if If this is the case, it indicates that the persistence of the state change is enhanced, suggesting that the state is becoming unstable.

[0140] if If the state change is determined to be non-persistent, it indicates a tendency for sudden engine shutdown.

[0141] Establishing a nonlinear mapping of spark states using a hybrid kernel support vector machine (SVR) includes:

[0142] Add a non-linear term to the original combustion anomaly index, namely:

[0143] ;

[0144] The kernel function for SVR is:

[0145] ;in, Indicates the Gaussian kernel weights. Indicates the Gaussian kernel bandwidth. Indicates the order of a polynomial; Indicates SVR input variables;

[0146] Through a self-attention mechanism, the weights are adaptively adjusted, that is:

[0147] , =1, 2, 3, 4, 5; where, Indicates the sharpening factor; This indicates the sensitivity of the output to the input parameters.

[0148] When the fuel is powder, in order to make the combustion abnormality index applicable to the characteristics of the fuel, a combustion characteristic adaptive layer is introduced into the combustion abnormality index to adapt to the characteristics of the powder fuel, including:

[0149] Obtaining fuel information, the fuel information including: 90% powder fuel particle passing screen size, that is, 90% cumulative particle size of fuel particles; fuel perceived ash-free volatile matter; ash softening temperature and received base moisture content;

[0150] Constructing a fuel characteristic vector ; wherein, represents the 90% cumulative particle size of fuel particles, represents the fuel reference particle size; represents the fuel perceived ash-free volatile matter, represents the ash softening temperature, represents the received base moisture content of the fuel;

[0151] Constructing a fuel-spark transfer function:

[0152] ;

[0153] Based on the corrected parameters , Calculate the combustion abnormality index to obtain a combustion abnormality index that fuses fuel characteristics ;

[0154] Adding a fuel phase to the original phase space to obtain a new phase space:

[0155] ; wherein, represents the fuel characteristic time constant;

[0156] Adding fuel similarity measurement to the SVR kernel function to obtain a corrected kernel function:

[0157] ; wherein, represents the Manhattan distance; when the fuel is switched, that is, ; , automatically reducing the weight of historical data to reduce the influence of fuel characteristics on the combustion abnormality index.

[0158] When performing hierarchical blocking, adjust the blocking measures according to the characteristics of the fuel (especially powder fuel) to improve the blocking effect, specifically:

[0159] Adjusting the blocking measures according to the fuel characteristics to improve the blocking effect, including:

[0160] Adjusting the water mist concentration according to the fuel perceived ash-free volatile matter; the adjusted water mist concentration ; wherein, Reference water mist concentration;

[0161] Adjust the inert gas concentration according to the fuel particle size; the adjusted inert gas concentration is .

[0162] The application further provides a computer readable storage medium, which stores a computer program, and the computer program is executed by a processor to realize the hot blast furnace spark real-time monitoring and automatic blocking system.

[0163] The processes described above with reference to the flowcharts can be implemented as computer software programs in embodiments of the application. Embodiments of the application include a computer program product comprising a computer program carried on a computer readable medium, the computer program comprising program code for performing the methods shown in the flowcharts. In such embodiments, the computer program can be downloaded and installed from a network via a communication section, and / or installed from a detachable medium. When the computer program is executed by a central processing unit (CPU), the above-mentioned functions defined in the method of the application are performed. It should be noted that the above-mentioned computer readable medium of the application can be a computer readable signal medium or a computer readable storage medium or any combination of the two. The computer readable storage medium may, for example, but not limited to, an electrical, magnetic, optical, electromagnetic, infrared or semiconductor system, device or apparatus, or any combination of the above. More specific examples of the computer readable storage medium can include, but are not limited to, an electrical connection having one or more conductive wires, a portable computer diskette, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the above. In the present application, the computer readable storage medium can be any tangible medium containing or storing a program that can be used by or in conjunction with an instruction execution system, device or apparatus. In the present application, the computer readable signal medium can include a data signal carried in a baseband or as a part of a carrier wave, which carries the computer readable program code. Such a propagated data signal can take various forms, including but not limited to an electromagnetic signal, an optical signal or any suitable combination of the above. The computer readable signal medium can also be any computer readable medium other than the computer readable storage medium, which can send, propagate or transmit the program for use by or in conjunction with an instruction execution system, device or apparatus. The program code contained on the computer readable medium can be transmitted by any suitable medium, including but not limited to a wireless segment, a wire segment, an optical cable, an RF, etc., or any suitable combination of the above.

[0164] The computer program product of the present application can be a computer program product comprising a computer-readable medium bearing computer program code embodied therein for use with a computer. The computer program code can be code defining and / or implementing the present application. The computer program code can be written in any suitable computer readable programming language. The computer program code can be stored in a computer- readable storage medium, such as, but not limited to, any type of disk including an optical disk, a CD-ROM, a CD-R, a CD-RW, a DVD, a flash memory, a ROM, a RAM, a magnetic disk or hard drive, or any other suitable type of medium including a medium that holds the software for a particular or specialized computing purpose, or any suitable combination of media. The computer program product can be a computer program product distributed to end users, whether as a stand-alone program, as part of a physical system, or as a software download. The computer program product can be distributed on a physical medium, such as, but not limited to, a floppy disk, a CD-ROM, a CD-R, a CD-RW, a DVD, a flash memory, a ROM, a RAM, a magnetic disk or hard drive, or any other suitable type of medium, or any suitable combination of media. The computer program product can be distributed from a program distribution center, either as a tangible medium or via electronic delivery, such as from a Web site via the Internet, or from one computer to another via electronic transfer, such as by e-mail. The computer program product can be distributed in an encrypted manner, such as via encryption or via password protection.

[0165] Those skilled in the art will understand that the application described above and illustrated in the accompanying drawings is presented by way of example only and is not intended to limit the application. The present application thus extends to any and all embodiments within the scope of the following claims.

Claims

1. A hot blast stove spark real-time monitoring and automatic blocking system, characterized in that, Comprise: Spark monitoring module, the spark monitoring module is configured to be able to collect spark state information, and extract spark state features from state information;Comprise: Obtain spark temperature information through infrared imaging equipment and temperature sensor; Capture spark trajectory information through high-speed camera; Obtain spark concentration information online through laser scattering particle counter; Obtain spark carbon content information in Raman spectrometer; Extract spark surface temperature features from spark temperature information, extract spark motion speed features from spark trajectory information, and extract spark number concentration features from spark concentration information; Combustion monitoring module, the combustion monitoring module is configured to be able to obtain fuel combustion state data and fuel information; Safety control module, the safety control module is configured to be able to obtain data from spark monitoring module and combustion monitoring module, and based on spark state features, the combustion state abnormality is deduced, and the mapping relationship between spark state features and combustion abnormality is established; Self-blocking control module, the self-blocking control module is configured to be able to use hierarchical blocking strategy according to the mapping relationship between spark state features and combustion abnormality, realize hierarchical blocking, and improve system safety.

2. The hot blast stove spark real time monitoring and automatic blocking system as claimed in claim 1, wherein, The mapping relationship between spark state features and combustion abnormality is established based on the combustion state abnormality deduced from spark state features, comprising: Carrying out combustion abnormality quantitative diagnosis, comprising the following steps: establishing a combustion anomaly index wherein, represents a surface temperature of the Mars, represents a combustion temperature reference value, represents a Mars movement speed; represents the design flue gas flow rate; represents the spark concentration; represents the spark concentration threshold; represents the 50% cumulative particle size, i.e. the median particle size, of the sparks; represents the particle size threshold; represents the carbon content of the sparks; If then the combustion is judged to be normal; If abnormal combustion is determined.

3. The hot blast stove spark real time monitoring and automatic blocking system as claimed in claim 2, wherein, After judging the combustion abnormality, combustion abnormality type identification is carried out to establish the mapping relationship between spark state features and combustion abnormality, comprising: Constructing a sequence of combinations of changes in features of Mars: ; If and then determine the type of combustion abnormality as oxygen deficiency combustion; denotes the acquisition time interval; Constructing a sequence of combinations of changes in features of Mars: ; If and then the combustion abnormality type is judged to be local over-temperature; Constructing a sequence of combinations of changes in features of Mars: ; If and then the combustion abnormality type is judged to be unreasonable air distribution; Constructing a sequence of combinations of changes in features of Mars ; If and then the combustion abnormality type is judged to be a fuel abnormality.

4. The hot blast stove spark real time monitoring and automatic blocking system as claimed in claim 3, wherein, The mapping relationship between spark state features and combustion abnormality is established based on the combustion state abnormality deduced from spark state features, further comprising: Reconstructing the phase space of the spark surface temperature by using Takens embedding theorem: ； wherein, denotes the embedding dimension, denotes the delay time; ; wherein, denotes the data acquisition period; Constructing recurrence plot and carrying out recursive quantitative analysis, comprising: The state vector in the phase space is compared A binary matrix, the recurrence plot, is generated by comparing The spark state feature in the data acquisition period is acquired, a spark state feature time sequence is constructed, and long-range correlation of the spark state feature time sequence is described by a Hurst index The spark state feature time sequence is constructed, and long-range correlation of the spark state feature time sequence is described by a Hurst index 5. The hot blast stove spark real time monitoring and automatic blocking system as claimed in claim 4, wherein, said state vector in phase space comparing, generating a binary matrix, i.e. a recurrence plot, comprising: Computing the recurrence rate To quantify the recurrence probability of spark temperature state, reflecting the combustion stability; ； wherein the similarity threshold ; denotes the Heaviside step function; Computational determinism to indicate whether periodic oscillations exist in the phase space; wherein, represents any one element value in the recursive graph, represents the diagonal length in the recursive graph, represents the number of line segments in the recursive graph with diagonal length of The spark state feature in the data acquisition period is acquired, a spark state feature time sequence is constructed, and the long-range correlation of the spark state feature time sequence is described by using a Hurst index The spark state feature time sequence is described by using a Hurst index wherein, denotes a sequence range of the spark state feature time series; denotes a standard deviation; denotes a time window length; ; wherein, denotes a time window adjustment amount; Establishing spark state nonlinear mapping by using hybrid kernel support vector machine SVR, comprising: In the original combustion anomaly index, a non-linear term is added, i.e.: ; The kernel function of SVR is: ; wherein, denotes a Gaussian kernel weight, denotes a Gaussian kernel bandwidth, denotes a polynomial order; denotes an SVR input variable; Through the self-attention mechanism, the weight is adjusted adaptively, that is: ； wherein, represents a sharpening coefficient; represents the sensitivity of the output to the input parameter.

6. The hot blast stove spark real time monitoring and automatic blocking system as claimed in claim 5, wherein, The mapping relationship between spark state features and combustion abnormality is established based on the combustion state abnormality deduced from spark state features, further comprising: A combustion risk level assessment is performed, and a comprehensive combustion risk index is set as ; wherein, represents a safe flow rate, represents a step function; represents a risk value of slagging and a risk value of unburnt volatile matter; wherein, represents the ratio of the Si02 content and the CaO content; represents the spark sphericity, ; wherein, represents the concentration of sparks having a particle size of less than 1 μm. The combustion risk is classified: When the integrated combustion risk is level one, an early warning signal is outputted, and the blocking measure is adjusting the air-fuel ratio; When the integrated combustion risk is secondary; the blocking measures are cyclone interception and water mist interception; When the integrated combustion risk is three; the blocking measure is to suspend combustion, perform emergency inerting and detonation blocking.

7. The hot blast stove spark real time monitoring and automatic blocking system as claimed in claim 6, wherein, Introducing fuel characteristic adaptation layer into combustion abnormality index to adapt to powder fuel characteristics, comprising: Obtaining fuel information, the fuel information includes: 90% powder fuel particle passes through the screen hole size, that is, the cumulative particle size of 90% fuel particles;Fuel perception ashless machine volatile matter;Ash softening temperature and received base water content; Constructing fuel-spark transfer function: Constructing a fuel property vector ; wherein, represents the 90% cumulative particle size of the fuel particles, represents the reference particle size of the fuel; represents the ash-free volatile matter of the fuel as perceived; represents the ash softening temperature, represents the received base moisture content of the fuel; Increase fuel phase in the original phase space to obtain new phase space: based on the corrected parameters calculating a combustion abnormality index, obtaining a combustion abnormality index that fuses fuel characteristics ; Add fuel similarity measurement in SVR kernel function to obtain corrected kernel function: wherein, represents a fuel characteristic time constant; The mapping relationship between spark state features and combustion abnormality is established based on the combustion state abnormality deduced from spark state features, further comprising: wherein, denotes the Manhattan distance; When the fuel is switched, i.e. Automatically reduces the historical data weight, reducing the influence of fuel properties on the combustion abnormality index.

8. The hot blast stove spark real time monitoring and automatic blocking system as claimed in claim 7, wherein, According to the fuel characteristics, adjust the blocking measures to improve the blocking effect, comprising: ​ Adjusting the water mist concentration according to fuel sensed ash-free based volatile matter; the adjusted water mist concentration wherein, is the reference water mist concentration; Adjusting the inert gas concentration according to the fuel particle size; the adjusted inert gas concentration is 9. A computer-readable storage medium, characterized in that, The computer readable storage medium stores a computer program, and the computer program is executed by the processor to realize the hot blast furnace spark real-time monitoring and automatic blocking system in any one of claims 1-8.

Images