A multi-dimensional information fusion combustion monitoring system and method for a waste incinerator

Abstract

The application discloses a multi-dimensional information fusion combustion monitoring system and method for a waste incinerator, and relates to the technical field of waste incineration power generation. The system comprises a multi-dimensional basic data acquisition module, a physical characteristic fusion module, a process characteristic fusion module, a verification and secondary fusion module and an intelligent decision module. The method collects two-dimensional radiation grayscale images, sound wave flying time signals, thermal process data and flue gas chemical data in the furnace, respectively performs first-order spatial fusion and first-order time sequence fusion, obtains a first core index group representing the combustion state of the solid material layer and a second core index group representing the state of the gas phase incomplete combustion, and generates a three-dimensional comprehensive combustion state space matrix of the furnace chamber after the coupling state deduction, reverse verification and correction of the fusion boundary conditions, thereby outputting a combustion feedforward control instruction. The application realizes the advance adjustment of the air volume and the speed of the grate, effectively suppresses the temperature fluctuation of the furnace chamber, inhibits the excessive flue gas pollutants, and improves the operation stability and environmental protection of the incinerator.

Description

Technical Field

[0001] This invention relates to the field of waste incineration power generation technology, and specifically to a multi-dimensional information fusion combustion monitoring system and method for waste incinerators. Background Technology

[0002] Waste-to-energy incineration is currently the mainstream technology for reducing, rendering harmless, and recycling municipal solid waste. Its operational stability directly affects energy efficiency and environmental emission levels. As the proportion of municipal solid waste incinerated continues to increase, incinerators are developing towards larger sizes and higher parameters, placing higher demands on real-time, accurate monitoring and intelligent control of the combustion state within the furnace. A stable combustion process is a core prerequisite for ensuring the safe operation of the incinerator, reducing energy consumption, and controlling emissions of pollutants such as dioxins.

[0003] In existing technologies, traditional combustion monitoring mainly relies on contact sensors such as thermocouples and flue gas analysis systems at the chimney inlet. This approach suffers from problems such as insufficient representativeness of measurement points, sensor susceptibility to coking and failure, and significant data lag in flue gas monitoring, failing to comprehensively reflect the true distribution of solid-gas two-phase combustion within the furnace. Furthermore, existing multi-source information fusion methods lack effective solid-gas coupling verification mechanisms, resulting in low accuracy in combustion state judgment and difficulty in early intervention. This leads to large fluctuations in furnace temperature, low combustion efficiency, and easy exceedances of flue gas pollutants, failing to meet the requirements for refined operation of incinerators. Summary of the Invention

[0004] This invention provides a multi-dimensional information fusion combustion monitoring system and method for waste incinerators, which solves the problems of insufficient representativeness of traditional monitoring points, serious lag in flue gas data, low accuracy in combustion state judgment, resulting in large fluctuations in furnace temperature and easy exceedance of pollutants in the prior art.

[0005] To achieve the above objectives, one embodiment of the present invention provides a multi-dimensional information fusion combustion monitoring system for waste incinerators, comprising: a multi-dimensional basic data acquisition module for real-time acquisition of two-dimensional radiometric grayscale images, acoustic wave transit time signals, thermal process data, and flue gas chemical data at the furnace outlet; a physical feature fusion module for performing first-order spatial fusion of the two-dimensional radiometric grayscale images and acoustic wave transit time signals to calculate a first core index set to characterize the current combustion state of the solid material layer on the grate; and a process feature fusion module for performing first-order temporal fusion of the thermal process data and flue gas chemical data to calculate a second core index set to characterize incomplete gas phase combustion. Combustion state; Verification and secondary fusion module, used to perform reverse verification of the first core indicator group and the second core indicator group according to the coupling state deduction strategy; if the reverse verification is successful, a three-dimensional comprehensive combustion state space matrix of the furnace is generated according to the first core indicator group and the second core indicator group; if the reverse verification fails, the boundary conditions in the first-order space fusion are reverse corrected according to the second core indicator group, the first core indicator group is recalculated, and the reverse verification is performed again; Intelligent decision module, used to generate combustion feedforward control commands according to the three-dimensional comprehensive combustion state space matrix of the furnace to adjust the air volume and grate speed in advance, smooth furnace temperature fluctuations and suppress excessive emissions of flue gas pollutants.

[0006] Furthermore, the multidimensional basic data acquisition module includes: a spatial vision acquisition unit, including an infrared camera, for acquiring two-dimensional radiation grayscale images of the grate firebed surface; an internal acoustic acquisition unit, including a sound wave generator and an acoustic receiving sensor, for acquiring sound wave transit time signals; a thermal process acquisition unit, for acquiring thermal process data, including primary air volume, secondary air volume, wind chamber pressure, pusher frequency, grate speed, and main steam flow; and a flue gas chemistry acquisition unit, for acquiring flue gas chemistry data, including oxygen concentration and carbon monoxide concentration.

[0007] On the other hand, a multi-dimensional information fusion combustion monitoring method for waste incinerators is also provided, applied to the aforementioned multi-dimensional information fusion combustion monitoring system, including: Step S1: acquiring a two-dimensional radiometric grayscale image, acoustic wave transit time signal, thermal process data, and flue gas chemical data at the furnace outlet of the waste incinerator; Step S2: performing first-order spatial fusion of the two-dimensional radiometric grayscale image and acoustic wave transit time signal to calculate a first core index group; performing first-order temporal fusion of the thermal process data and flue gas chemical data to calculate a second core index group; Step S3: based on the coupling state deduction strategy, The first core indicator group and the second core indicator group are reverse verified. If the reverse verification is successful, a three-dimensional integrated combustion state space matrix of the furnace is generated based on the first core indicator group and the second core indicator group. If the reverse verification is unsuccessful, the boundary conditions in the first-order spatial fusion are reverse corrected based on the second core indicator group, the first core indicator group is recalculated, and the reverse verification is performed again until the verification is successful. Step S4: A combustion feedforward control command is generated based on the three-dimensional integrated combustion state space matrix of the furnace to adjust the air volume and grate speed in advance, smooth the furnace temperature fluctuations, and suppress the excessive emissions of flue gas pollutants.

[0008] Furthermore, the first-order spatial fusion includes: establishing a three-dimensional physical coordinate system for the furnace, mapping a two-dimensional radiation grayscale image onto the three-dimensional physical coordinate system for the furnace, and calculating the geometric boundary of the fire line on the firebed plane; within the three-dimensional physical coordinate system for the furnace, aligning the surface radiation features of the two-dimensional radiation grayscale image with the cross-sectional temperature field reconstructed from the acoustic time-of-flight signal at the same spatial scale; based on the spatially aligned surface radiation features and cross-sectional temperature field, and combined with the conservation relationship of radiation heat transfer, eliminating the interference of gaseous suspended matter on the radiation of the firebed surface, and calculating the layered combustion temperature distribution on the firebed surface; calculating the firebed moving speed based on the change characteristics of the fire line geometric boundary, and obtaining the solid material layer thickness and waste calorific value of each zone based on the layered combustion temperature distribution, and combining them to generate the first core index group.

[0009] Furthermore, the first-order time-series fusion includes: aligning the thermal process data and flue gas chemical data along the time axis according to a preset flue gas flow lag time model; performing multivariate nonlinear correlation analysis on the aligned main steam flow rate, oxygen concentration, and carbon monoxide concentration to calculate the combustion rate index and the chemical incomplete combustion component; and combining the furnace flow field distribution law with the combustion rate, chemical incomplete combustion component, and the time-series fluctuation characteristics of the wind chamber pressure and pusher frequency to calculate the furnace carbon monoxide concentration gradient distribution and real-time volatile matter release rate, and combining them to generate the second core index group.

[0010] Furthermore, the coupling state deduction strategy includes: matching the first core index group with a preset corresponding relationship library to obtain reference values ​​for gas phase combustion parameters; calculating the relative deviation between each reference value for gas phase combustion parameters and the corresponding measured parameters in the second core index group, and determining whether the reverse verification passes based on a preset deviation threshold.

[0011] Furthermore, the generation of the three-dimensional integrated combustion state space matrix of the furnace includes: establishing a three-dimensional grid space discretization model of the furnace, dividing it into a bottom base grid and a vertically suspended grid, and determining the spatial physical coordinates of each grid node; mapping the first core index group that has passed reverse verification to the bottom base grid according to the spatial physical coordinates of the grid nodes to construct the firebed surface feature basis matrix; filling the second core index group into the vertically suspended grid to construct the gas phase combustion space feature matrix; and performing a 3D matrix superposition operation on the firebed surface feature basis matrix and the gas phase combustion space feature matrix to obtain the three-dimensional integrated combustion state space matrix of the furnace.

[0012] Furthermore, the boundary condition correction includes: locking the maximum relative deviation that is greater than or equal to a preset deviation threshold, and identifying the gas phase combustion parameters corresponding to the maximum relative deviation; calculating the radiation correction coefficient and the geometric translation of the spatial alignment boundary by combining the preset sensitivity matrix; and using the correction coefficient and the translation to quantitatively correct the first-order spatially fused radiation heat transfer boundary and spatial alignment boundary.

[0013] Furthermore, the generation of the combustion feedforward control command includes: extracting the values ​​of each grid node in the three-dimensional integrated combustion state space matrix of the furnace, obtaining the positive distribution coefficient of the firebed, the spatial coordinates of the concentrated release zone of volatile matter, and the predicted value of the total heat load of the furnace through spatial integration and deviation statistical calculation; comparing the positive distribution coefficient of the firebed and the solid material layer thickness of each zone with the preset standard firebed shape, calculating the feedforward correction increment of the grate speed of each section, and generating the grate speed adjustment command; and based on the spatial coordinates of the concentrated release zone of volatile matter, real-time... The volatile matter release rate and the amount of incomplete chemical combustion are used to calculate the local theoretical oxygen demand and secondary air supply for the corresponding region. Combined with the predicted total heat load of the furnace, multivariate decoupling calculations are performed on the coupled control parameters such as global air volume, zone air chamber pressure, and zone secondary air volume to obtain the feedforward control target values ​​for the primary air chamber pressure and the ratio of primary air volume to secondary air volume for each section. Air volume adjustment commands are generated based on the feedforward control target values. The grate speed adjustment command and the air volume adjustment command are combined into the combustion feedforward control command.

[0014] The present invention provides a multi-dimensional information fusion combustion monitoring system and method for waste incinerators. By performing first-order spatial fusion and first-order temporal fusion of in-furnace visual and acoustic data and process flue gas data, the incomplete combustion state of the solid material layer and gas phase is simultaneously acquired. Combined with a solid-gas coupling reverse verification mechanism to correct the fusion boundary conditions, the accuracy and comprehensiveness of combustion state perception are effectively improved. Based on the generated three-dimensional comprehensive combustion state space matrix of the furnace, feedforward control commands are output, realizing early prediction and precise adjustment of combustion anomalies, significantly smoothing furnace temperature fluctuations, suppressing excessive flue gas pollutants, and improving the stability and combustion efficiency of incinerator operation. Attached Figure Description

[0015] To more clearly illustrate the technical solutions in this invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are some embodiments of this invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort. In the drawings: Figure 1 This is an architecture diagram of the multi-dimensional information fusion combustion monitoring system provided in this embodiment of the invention; Figure 2 This is a general flowchart of the multi-dimensional information fusion combustion monitoring method provided in the embodiments of the present invention; Figure 3 This is a schematic diagram of the three-dimensional integrated combustion state space matrix structure of the furnace provided in an embodiment of the present invention; Figure 4 This is the operating interface of the multi-dimensional information fusion combustion monitoring system provided in this embodiment of the invention. Detailed Implementation

[0016] The specific embodiments of the present invention will be described in detail below with reference to the accompanying drawings. It should be understood that the specific embodiments described herein are for illustration and explanation only and are not intended to limit the scope of the present invention.

[0017] It should be noted that the acquisition, transmission, storage, use, and processing of data in the technical solution of this application all comply with the relevant provisions of national laws and regulations. In the embodiments of this application, certain existing industry solutions such as software, components, and models may be mentioned. These should be considered exemplary, intended only to illustrate the feasibility of implementing the technical solution of this application, and do not imply that the applicant has already used or necessarily used such solutions.

[0018] The following is combined Figures 1-4 This invention is described in detail.

[0019] like Figure 1 and Figure 4As shown, this embodiment of the invention provides a multi-dimensional information fusion combustion monitoring system for waste incinerators, comprising: a multi-dimensional basic data acquisition module for real-time acquisition of two-dimensional radiometric grayscale images, acoustic wave transit time signals, thermal process data, and flue gas chemical data at the furnace outlet; a physical feature fusion module for performing first-order spatial fusion of the two-dimensional radiometric grayscale images and acoustic wave transit time signals to calculate a first core index set to characterize the current combustion state of the solid material layer on the grate; and a process feature fusion module for performing first-order temporal fusion of the thermal process data and flue gas chemical data to calculate a second core index set to characterize the incomplete combustion state in the gas phase. The verification and secondary fusion module is used to perform reverse verification of the first core indicator group and the second core indicator group according to the coupling state inference strategy. If the reverse verification is successful, a three-dimensional comprehensive combustion state space matrix of the furnace is generated based on the first core indicator group and the second core indicator group. If the reverse verification fails, the boundary conditions in the first-order space fusion are corrected in reverse according to the second core indicator group, the first core indicator group is recalculated, and the reverse verification is performed again. The intelligent decision-making module is used to generate combustion feedforward control commands based on the three-dimensional comprehensive combustion state space matrix of the furnace to adjust the air volume and grate speed in advance, smooth furnace temperature fluctuations, and suppress excessive emissions of flue gas pollutants.

[0020] The multidimensional basic data acquisition module includes: a spatial vision acquisition unit, including an infrared camera, for acquiring two-dimensional radiation grayscale images of the grate firebed surface; an internal acoustic acquisition unit, including a sound wave generator and an acoustic receiving sensor, for acquiring sound wave transit time signals; a thermal process acquisition unit, for acquiring thermal process data, including primary air volume, secondary air volume, wind chamber pressure, pusher frequency, grate speed, and main steam flow; and a flue gas chemistry acquisition unit, for acquiring flue gas chemistry data, including oxygen concentration and carbon monoxide concentration.

[0021] The multi-dimensional basic data acquisition module is responsible for comprehensively acquiring multi-dimensional raw state data during the combustion process of the incinerator, covering the spatial characteristics of solid bed combustion, the acoustic characteristics of gas-phase combustion, the process characteristics of equipment operation, and the chemical characteristics of flue gas emissions, providing a complete data foundation for subsequent fusion analysis. Two-dimensional radiation grayscale images are generated by converting the thermal radiation signals of objects inside the furnace captured by an infrared camera. Different grayscale values ​​in the image correspond to different radiation intensities on the surface of the combustion bed, indirectly reflecting the temperature distribution and combustion intensity of the combustion bed surface. The acoustic wave transit time signal refers to the time it takes for the sound wave emitted by the acoustic wave generator to pass through the high-temperature flue gas medium inside the furnace and be captured by the acoustic receiving sensor; it is the core basis for reconstructing the temperature field of the furnace cross-section. Thermal process data reflects the real-time operating status of the incinerator body and auxiliary systems, and is a basic parameter for judging the stability and load level of the combustion process. Flue gas chemical data directly reflects the composition of combustion products at the furnace outlet, and is a key indicator for assessing the degree of combustion completeness and pollutant generation trends. The physical feature fusion module specifically processes sensing data with spatial distribution characteristics, eliminating the measurement errors and limitations of single sensors by integrating spatial data from different sources under a unified coordinate system. First-order spatial fusion is a preliminary fusion process for spatial dimension data. Its core is to transform sensor data with different spatial resolutions and measurement principles into the same physical spatial coordinate system, achieving precise alignment of spatial positions and preliminary integration of features. The first core indicator group comprehensively characterizes the combustion state of solid waste on the grate, including multiple parameters reflecting combustion intensity, material distribution, material characteristics, and combustion propulsion speed. The process feature fusion module specifically processes operational data with time-series characteristics, establishing a dynamic correlation between process parameters and flue gas parameters by eliminating time lags between different parameters. First-order temporal fusion is a preliminary fusion process for time dimension data. Its core is to align the time axis based on the differences in response time of different parameters during combustion, eliminating temporal deviations caused by data transmission and process lags. The second core indicator group comprehensively characterizes the state of gas-phase combustion within the furnace, including multiple parameters reflecting overall combustion intensity, combustion completeness, pollutant distribution, and volatile matter release characteristics. The verification and secondary fusion module is responsible for verifying the consistency of the two sets of core indicators obtained from the primary fusion, and completing the final multi-dimensional data fusion after successful verification. The coupled state extrapolation strategy is based on the inherent physical coupling between solid-phase combustion and gas-phase combustion. It extrapolates the corresponding incomplete gas-phase combustion state from the solid-phase combustion state, achieving cross-validation of two sets of independent calculation results. Reverse validation utilizes measured gas-phase combustion results to verify the accuracy of the solid-phase combustion state calculations, effectively correcting errors caused by environmental interference during spatial fusion. Boundary conditions refer to the pre-set initial parameters and constraints during spatial fusion calculations, directly affecting the accuracy of the fusion results.The three-dimensional integrated combustion state space matrix of the furnace is a three-dimensional data structure that integrates the combustion parameters of the solid and gas phases according to their actual spatial positions within the furnace. It can simultaneously reflect the combustion state distribution at different heights and locations. Based on the final generated integrated combustion state data and the control objectives of the incinerator operation, the intelligent decision-making module automatically generates optimal adjustment commands. The combustion feedforward control command is an adjustment command issued in advance based on the prediction of future combustion trends from the current combustion state. Unlike traditional feedback control, this effectively eliminates control lag.

[0022] Specifically, the multi-dimensional basic data acquisition module, serving as the system's data source, employs a multi-sensor fusion acquisition strategy to ensure data comprehensiveness and reliability. The spatial vision acquisition unit acquires combustion status information from the firebed surface through non-contact optical measurement, avoiding the vulnerability of contact sensors to high-temperature and high-corrosion environments. Infrared cameras, installed on the side walls of the furnace and at observation holes in the front arch, utilize high-temperature resistant lenses and automatic purging devices, enabling long-term stable operation under harsh conditions. The internal acoustic acquisition unit measures the temperature distribution within the furnace by utilizing the propagation characteristics of sound waves in gas, offering advantages such as a wide measurement range, fast response speed, and immunity to flue gas corrosion. Sound wave generators and acoustic receiving sensors are evenly distributed around the furnace walls of the first channel, forming a measurement network covering the entire combustion cross-section. The thermal process acquisition unit interfaces with the incinerator's existing distributed control system via a standard industrial communication interface, enabling real-time acquisition of all process parameters of the incinerator operation without the need for numerous additional sensors, thus reducing system modification costs. The acquired data, after filtering and preprocessing, is transmitted to the process feature fusion module for further processing. The flue gas chemical acquisition unit adopts an in-situ measurement method, directly measuring the flue gas composition at the furnace outlet, avoiding compositional changes and time lags during flue gas extraction and transmission. The in-situ measuring instrument is installed on the flue wall at the furnace outlet, employing heated sampling and rapid analysis technology to achieve millisecond-level response speeds.

[0023] Specifically, the physical feature fusion module and the process feature fusion module adopt a parallel processing architecture, capable of processing spatial and temporal data simultaneously, improving the system's real-time performance. The processing results from both modules are synchronously transmitted to the verification and secondary fusion module for cross-validation and final fusion. The verification and secondary fusion module employs a closed-loop verification mechanism. When the deviation between the two sets of core indicators exceeds a preset deviation threshold, a boundary condition correction process is automatically triggered, recalculating the first set of core indicators until verification is successful. This mechanism effectively improves the accuracy and robustness of the fusion results.

[0024] Specifically, the intelligent decision-making module employs a feedforward control algorithm, which can predict the combustion trend over a future period based on the three-dimensional integrated combustion state space matrix of the furnace and issue adjustment commands in advance. These adjustment commands are transmitted to the incineration plant's distributed control system via the thermal process acquisition unit, achieving automatic closed-loop control of the combustion process.

[0025] The multi-dimensional information fusion combustion monitoring system provided in this invention achieves comprehensive perception of the solid-gas two-phase combustion state in the furnace through the coordinated acquisition of infrared vision, acoustic temperature measurement, process parameters and in-situ flue gas data, combined with a parallel processing architecture of solid-phase spatial fusion and gas-phase temporal fusion, providing reliable data support for subsequent coupling verification and precise control.

[0026] like Figure 2 As shown, this embodiment of the invention also provides a multi-dimensional information fusion combustion monitoring method for waste incinerators, applied to the aforementioned multi-dimensional information fusion combustion monitoring system, including: Step S1: acquiring a two-dimensional radiometric grayscale image, acoustic wave transit time signal, thermal process data, and flue gas chemical data at the furnace outlet of the waste incinerator; Step S2: performing first-order spatial fusion of the two-dimensional radiometric grayscale image and acoustic wave transit time signal to calculate a first core index group; performing first-order temporal fusion of the thermal process data and flue gas chemical data to calculate a second core index group; Step S3: inferring strategies based on the coupling state. The reverse verification of the first core indicator group and the second core indicator group is omitted. If the reverse verification is successful, a three-dimensional integrated combustion state space matrix of the furnace is generated based on the first core indicator group and the second core indicator group. If the reverse verification fails, the boundary conditions in the first-order spatial fusion are corrected in reverse according to the second core indicator group, the first core indicator group is recalculated, and the reverse verification is performed again until the verification is successful. Step S4: A combustion feedforward control command is generated based on the three-dimensional integrated combustion state space matrix of the furnace to adjust the air volume and grate speed in advance, smooth the furnace temperature fluctuation and suppress the excessive emission of flue gas pollutants.

[0027] This method, based on the aforementioned system architecture, employs a step-by-step fusion and closed-loop verification approach to achieve precise perception and feedforward control of the combustion state. "Re-execution" refers to the cyclical process where, if reverse verification fails, the system automatically returns to the initial fusion step and recalculates using corrected boundary conditions. "Until verification passes" means the system continuously executes the cycle of reverse verification and parameter correction until the deviation between the two sets of core indicators meets preset requirements, ensuring the accuracy of the fusion results. "Smoothing furnace temperature fluctuations" refers to controlling furnace temperature fluctuations within a preset range by adjusting combustion parameters in advance, ensuring the stability of the incinerator's operation. "Suppressing excessive flue gas pollutants" refers to optimizing the combustion process to reduce the generation of pollutants such as carbon monoxide, ensuring that flue gas emissions meet environmental standards.

[0028] Specifically, step S1 employs a synchronous acquisition strategy to ensure that the timestamps of the four types of data remain consistent, providing a unified time reference for subsequent temporal and spatial fusion. The data acquisition frequency is set according to the operating characteristics of the incinerator, enabling the capture of rapid changes during the combustion process. Step S2 adopts a parallel processing approach, simultaneously performing first-order spatial fusion and first-order temporal fusion, shortening data processing time and improving the system's real-time performance. The two fusion processes are independent and do not interfere with each other, enabling the output of accurate solid-phase and gas-phase incomplete combustion state parameters respectively. Step S3 employs a closed-loop verification mechanism, effectively eliminating system errors caused by a single fusion path through cross-verification of the solid-gas two-phase combustion states. The reverse correction process is adaptive, automatically adjusting the correction amplitude according to the magnitude and direction of the deviation, avoiding over-correction or under-correction. Step S4 performs global optimization calculations based on the three-dimensional integrated combustion state space matrix of the furnace, generating feedforward control commands that can simultaneously adjust the air volume and grate speed, achieving multi-variable coordinated control. The output frequency of the control commands is consistent with the data acquisition frequency, ensuring the timeliness and continuity of control.

[0029] Preferably, the first-order spatial fusion includes: establishing a three-dimensional physical coordinate system for the furnace, mapping a two-dimensional radiation grayscale image onto the three-dimensional physical coordinate system for the furnace, and calculating the geometric boundary of the fire line on the firebed plane; within the three-dimensional physical coordinate system for the furnace, aligning the surface radiation features of the two-dimensional radiation grayscale image with the cross-sectional temperature field reconstructed from the acoustic time-of-flight signal at the same spatial scale; based on the spatially aligned surface radiation features and cross-sectional temperature field, and combined with the conservation relationship of radiation heat transfer, eliminating the interference of gaseous suspended matter on the radiation of the firebed surface, and calculating the layered combustion temperature distribution on the firebed surface; calculating the firebed moving speed based on the change characteristics of the fire line geometric boundary, and obtaining the solid material layer thickness and waste calorific value of each zone based on the layered combustion temperature distribution, and combining them to generate the first core index group.

[0030] Further preferably, the first-order time-series fusion includes: aligning the thermal process data and flue gas chemical data along the time axis according to a preset flue gas flow lag time model; performing multivariate nonlinear correlation analysis on the aligned main steam flow rate, oxygen concentration, and carbon monoxide concentration to calculate the combustion rate index and the chemical incomplete combustion component; combining the furnace flow field distribution law, based on the combustion rate and the chemical incomplete combustion component, and in conjunction with the time-series fluctuation characteristics of the wind chamber pressure and the pusher frequency, calculating the furnace carbon monoxide concentration gradient distribution and the real-time volatile matter release rate, and combining them to generate the second core index group.

[0031] The three-dimensional physical coordinate system of the furnace is a spatial coordinate system established based on the incinerator body. It can accurately calibrate the spatial coordinates of any position within the furnace, providing a unified reference for spatial alignment of data from different sensors. The fireline geometric boundary refers to the dividing line between the most intensely burning area and the unburned area on the firebed surface, directly reflecting the combustion progression and spatial distribution of the firebed. Surface radiation characteristics refer to the grayscale distribution patterns of different regions in a two-dimensional radiation grayscale image, directly related to the temperature and combustion state of the firebed surface. The cross-sectional temperature field refers to the temperature distribution within a certain cross-section of the furnace, reflecting the spatial non-uniformity of gas-phase combustion within that section. Spatial scale overlap and alignment refers to converting data collected by different sensors to the same spatial resolution and coordinate system, ensuring accurate matching of data at corresponding positions. The conservation relationship of radiative heat transfer refers to the fundamental physical laws governing the emission, absorption, and transmission of thermal radiation from an object, serving as the theoretical basis for eliminating environmental interference. Gas-phase suspended matter refers to dust, fly ash, and unburned carbon particles suspended within the furnace, which can obstruct and scatter the propagation of infrared radiation. The stoker temperature distribution refers to the temperature distribution at different locations on the burner surface and is a core parameter characterizing the intensity of solid material combustion. The burner movement speed refers to the speed at which the geometric boundary of the fire line advances along the grate's direction of movement, and is related to the grate speed and waste combustion characteristics. The solid material layer thickness refers to the height of the waste layer accumulated on the grate, directly affecting ventilation and combustion efficiency. The calorific value of waste refers to the heat released by the complete combustion of a unit mass of waste, and is a key parameter reflecting waste combustion characteristics. The flue gas flow lag time model describes the relationship between the time it takes for flue gas to flow from the combustion zone to the furnace outlet and the furnace geometry and flue gas velocity. This flue gas flow lag time model is calculated based on the furnace geometric volume and dynamic flue gas volumetric flow rate, i.e., the lag time. ,in The effective volume from the combustion zone to the furnace outlet is determined by combining the baseline volume with dynamic correction of the main combustion zone position. The real-time flue gas volumetric flow rate is estimated based on the measured primary and secondary air volumes superimposed with the combustion volumetric expansion coefficient. Time axis alignment refers to adjusting data from different acquisition times to the same time reference, eliminating response time differences between different parameters. Multivariate nonlinear correlation analysis refers to establishing nonlinear mapping relationships between multiple input and output parameters, uncovering the intrinsic connections between parameters. The combustion rate index refers to the mass of waste burned per unit time, a core parameter reflecting overall combustion intensity. The chemically incomplete combustion component refers to the proportion of heat loss due to incomplete combustion of combustible gases, an important indicator for assessing the degree of combustion completeness. The temporal fluctuation law refers to the trend and characteristics of parameter changes over time, reflecting the dynamic changes in the combustion process. Inversion calculation refers to the method of inferring unknown internal state parameters based on known boundary parameters and physical laws. The carbon monoxide concentration gradient distribution refers to the variation of carbon monoxide concentration along the furnace height, reflecting the spatial distribution of gas-phase combustion. The real-time volatile matter release rate refers to the mass of volatile matter released from waste per unit time, an important parameter for predicting combustion trends.

[0032] Specifically, the four types of data are collected at the same time to ensure that the collection timestamps are synchronized; however, there is a physical lag in the flow of flue gas from the combustion zone to the furnace outlet. Therefore, a secondary time axis alignment is required through the flue gas flow lag time model to ensure that the data corresponds to the same combustion condition.

[0033] Specifically, the three-dimensional physical coordinate system of the furnace is established with the upper left corner of the grate front end as the origin, the grate movement direction as the X-axis, the furnace width direction as the Y-axis, and the furnace height direction as the Z-axis. The accuracy of the coordinate system can meet the spatial positioning requirements at the millimeter level. The geometric boundary of the fire line is extracted through an image edge detection algorithm. Combined with the change characteristics of continuous frame images, the boundary between the burning area and the unburned area can be accurately identified.

[0034] Specifically, the spatial scale alignment process first interpolates the discrete point data of the cross-sectional temperature field into a temperature matrix with the same resolution as the two-dimensional radiation grayscale image. Then, coordinate transformation is performed according to the sensor's installation position and viewing angle to ensure that radiation data and temperature data at the same spatial location correspond one-to-one. The conservation relationship of radiation heat transfer, combined with the composition and concentration parameters of the flue gas in the furnace, allows for the quantitative calculation of the attenuation of infrared radiation by gaseous suspended matter, thereby restoring the true temperature of the firebed surface.

[0035] Specifically, the firebed movement speed is obtained by calculating the ratio of the displacement of the geometric boundary of the fire line to the time interval in two consecutive frames. The solid material layer thickness is obtained based on the correlation between the stoker temperature distribution and the grate ventilation. The calorific value of the waste is calculated by combining the stoker temperature distribution, firebed movement speed, and main steam flow parameters.

[0036] Specifically, the flue gas flow lag time model is updated in real time based on the furnace geometry and real-time flue gas velocity parameters, accurately reflecting the flue gas flow characteristics under different loads. The effective volume is determined using a combination of baseline calibration and dynamic correction. Before system deployment, based on the three-dimensional design model of the incinerator, the internal space of the furnace between the design starting section of the main combustion zone of the grate and the furnace outlet section is defined as the baseline volume interval. A fixed baseline geometric volume is calculated through spatial integration and stored in a preset parameter library. During operation, the fire line geometric boundary output by the first-order spatial fusion is called, the axial offset of the leading edge of the main combustion zone along the grate movement direction is extracted, and the volume correction is calculated in combination with the furnace cross-sectional area at the corresponding position. The baseline geometric volume is then dynamically subtracted to obtain the real-time effective volume.

[0037] The real-time flue gas volumetric flow rate Q is estimated using a method that superimposes measured air volume with combustion expansion correction. The calculation logic is the sum of the primary air volumetric flow rate and the secondary air volumetric flow rate multiplied by the combustion volumetric expansion coefficient. Both the primary and secondary air volumetric flow rates are read in real-time from the distributed control system by the thermal process acquisition unit. The combustion volumetric expansion coefficient characterizes the relative volume increase of the air entering the furnace after the combustion reaction. It is calculated based on the measured oxygen and carbon monoxide concentrations from the flue gas chemistry acquisition unit, through carbon combustion material balance and excess air coefficient.

[0038] The time axis alignment process uses the timestamp of the thermal process data as a reference and shifts the flue gas chemical data backward by the corresponding lag time to ensure that the two sets of data reflect the combustion state at the same moment.

[0039] Specifically, the multivariate nonlinear correlation analysis, based on a large amount of historical operating data, can accurately reflect the dynamic relationship between main steam flow, oxygen concentration, carbon monoxide concentration, combustion rate, and incomplete chemical combustion components. The inversion calculation process, combined with the influence of changes in wind chamber pressure and pusher frequency on the furnace flow field, can obtain the distribution of carbon monoxide concentration and volatile matter release rate at different heights.

[0040] An example of first-order spatial fusion application: When an incinerator processes municipal solid waste with high moisture content, the infrared camera in the spatial vision acquisition unit can acquire high-resolution radiation images of the combustion bed surface, clearly distinguishing the boundaries of the dry zone, combustion zone, and burnt-out zone. However, due to the obstruction of fly ash and water vapor in the furnace, the gray value at the center of the combustion zone is underestimated. The internal acoustic acquisition unit can reconstruct the accurate temperature distribution of the entire combustion cross section, but its spatial resolution is low, failing to capture subtle combustion differences in the local areas of the combustion bed. First-order spatial fusion, through spatial scale alignment, interpolates the acoustic temperature field to the same resolution as the infrared image and completes coordinate matching. Then, using the accurate temperature at the center of the combustion zone obtained from acoustic thermometry, it corrects the gray value of the obstructed area in the infrared image, ultimately obtaining stratification temperature distribution data that retains the detailed distribution of combustion on the combustion bed while possessing accurate temperature measurement precision.

[0041] An example of first-order time-series fusion application: When the incinerator load increases from 70% to 90%, the thermal process acquisition unit collects real-time data on parameter changes such as increased grate speed and primary air volume. This data immediately reflects adjustments in equipment operation. However, the carbon monoxide concentration data measured by the flue gas chemistry acquisition unit at the furnace outlet requires the time it takes for the flue gas to flow from the combustion zone to the outlet to reflect the changes in combustion state caused by this load adjustment. First-order time-series fusion first calculates the flue gas lag time under the current operating condition as 28 seconds based on the current flue gas velocity and furnace geometry using a flue gas flow lag time model. Then, it shifts the time axis of the flue gas chemistry data backward by 28 seconds to perfectly align the time point of grate speed adjustment with the time point of carbon monoxide concentration change. Finally, it performs correlation analysis on the aligned grate speed, primary air volume, main steam flow, and carbon monoxide concentration data to accurately calculate the changes in combustion rate and fluctuations in the chemically incomplete combustion component corresponding to this load increase.

[0042] In a preferred embodiment of the present invention, infrared visual and acoustic temperature measurement data are integrated through first-order spatial fusion to effectively eliminate measurement errors from single sensors and interference from gaseous suspended matter, accurately acquiring key state parameters such as the layer combustion temperature distribution, layer thickness, and calorific value of the solid material layer. First-order temporal fusion aligns thermal process and flue gas chemical data, completely resolving the time lag issue caused by flue gas flow and accurately characterizing the dynamic characteristics of gas-phase combustion. Both methods employ a parallel processing architecture, ensuring real-time performance while providing a high-precision, high-reliability core indicator foundation for subsequent solid-gas coupling verification and the generation of a three-dimensional combustion state matrix.

[0043] Preferably, the coupling state deduction strategy includes: matching the first core index group with a preset corresponding relationship library to obtain reference values ​​of gas phase combustion parameters; calculating the relative deviation between each reference value of gas phase combustion parameters and the corresponding measured parameters in the second core index group, and determining whether the reverse verification is passed based on a preset deviation threshold.

[0044] The preset correspondence database is a database built based on a large amount of historical operating data and combustion theory models, storing reasonable variation ranges of gas-phase combustion parameters corresponding to different solid-phase combustion states. The reference values ​​for gas-phase combustion parameters are theoretically expected values ​​matched from the preset correspondence database based on the solid-phase combustion state, used for comparison with the actually measured gas-phase combustion parameters. The relative deviation is the ratio of the difference between the reference value and the measured value to the reference value, objectively reflecting the degree of deviation between the two. The preset deviation threshold is a pre-set maximum allowable relative deviation value used to determine whether the consistency of the two sets of fusion results meets the requirements.

[0045] Specifically, a pre-defined correlation database is established during the system debugging phase. This database collects complete operational data from the incinerator under different loads and waste calorific values, and is calibrated using a thermodynamic combustion model. It covers the correlation between all parameters in the first core indicator group and their corresponding parameters in the second core indicator group. The data in the database is categorized and stored according to grate zones, enabling independent matching and verification of different combustion zones.

[0046] Specifically, the matching process employs a partitioned matching strategy. The parameters of the grate combustion temperature, waste calorific value, firebed movement speed, and bed thickness for each grate partition in the first core indicator group are input into a pre-defined corresponding relational database to obtain reference values ​​for the gas-phase combustion parameters for each partition. The matching process uses a weighted average algorithm, assigning different weight coefficients based on the degree of influence of each parameter on gas-phase combustion.

[0047] Specifically, the relative deviation calculation process is performed separately for each gas-phase combustion parameter. The calculated relative deviation can intuitively reflect the degree of difference between the theoretical expected value and the actual measured value of the parameter. Using relative deviation rather than absolute deviation for judgment can adapt to situations where the absolute value of the parameter varies greatly under different loads, thus improving the universality of the verification results.

[0048] Specifically, the preset deviation thresholds are set comprehensively based on the incinerator's operating requirements and the sensor's measurement accuracy, with different deviation thresholds set for different parameters. For key parameters affecting combustion stability and pollutant emissions, stricter deviation thresholds are set; for secondary parameters, the deviation thresholds are appropriately relaxed to improve system operating efficiency while ensuring verification accuracy.

[0049] Specifically, when the relative deviations of all gas-phase combustion parameters are less than the corresponding preset deviation thresholds, the reverse verification is considered successful, and the two sets of fused results show good consistency. When the relative deviation of any parameter is greater than or equal to the preset deviation threshold, the reverse verification is considered unsuccessful, and the system automatically triggers the boundary condition correction process. The verification process adopts a successive approximation method, and the verification is repeated after each correction until the relative deviations of all parameters meet the requirements.

[0050] For example, when the incinerator operates at 85% of its rated load and processes mixed municipal solid waste with a moisture content of 52%, the first-order spatial fusion calculation yields an average stratification temperature of 890 degrees Celsius in the first combustion zone, a waste calorific value of 8200 kJ / kg, a firebed movement speed of 0.8 m / min, and a solid material layer thickness of 0.6 m. The system inputs these parameters into a preset correspondence database and calculates reference values ​​for the corresponding gas-phase combustion parameters for that zone through weighted matching. These parameters include a combustion rate of 12.5 kg / s, a carbon monoxide concentration of 65 mg / m³, and a volatile matter release rate of 7.2 kg / s. At this time, the measured values ​​of the second core indicator group output by the first-order temporal fusion are: a combustion rate of 12.3 kg / s, a carbon monoxide concentration of 68 mg / m³, and a volatile matter release rate of 7.1 kg / s. The system calculates the relative deviations of the three parameters: a combustion rate deviation of 1.6%, a carbon monoxide concentration deviation of 4.6%, and a volatile matter release rate deviation of 1.4%. The preset deviation thresholds for carbon monoxide concentration, combustion rate, and volatile matter release rate are 5%, 4%, and 4%, respectively. In this test, all measured parameter deviations were below their corresponding thresholds, indicating successful verification. The consistency of the two sets of core indicators meets the requirements, and the system can directly proceed to the generation step of the three-dimensional integrated combustion state space matrix in the furnace. If the measured carbon monoxide concentration is 78 mg / m³, the relative deviation reaches 20%, exceeding the preset deviation threshold, and the system will determine that the verification has failed. In this case, the system will lock the carbon monoxide parameter with the largest deviation, automatically correct the radiation attenuation coefficient of gaseous suspended matter during the first-order spatial fusion process, recalculate the first set of core indicators, and perform coupling verification again until the relative deviations of all parameters meet the requirements.

[0051] In a preferred embodiment of the present invention, a coupled state deduction strategy is used to achieve cross-verification of the solid-gas two-phase combustion state. Based on a preset correspondence database, reference values ​​of gas-phase combustion parameters are obtained. The consistency of the fusion result is judged by combining the relative deviation threshold, which effectively solves the problem that the error of a single fusion path cannot be self-checked. At the same time, the use of partitioned matching and differentiated threshold settings improves the accuracy and adaptability of the verification process, and can quickly locate deviation parameters, providing a clear basis for the accurate correction of subsequent boundary conditions, thus ensuring the reliability of the final three-dimensional combustion state matrix from the source.

[0052] like Figure 3As shown, preferably, the generation of the three-dimensional integrated combustion state space matrix of the furnace includes: establishing a three-dimensional grid space discretization model of the furnace, dividing it into a bottom base grid and a vertically suspended grid, and determining the spatial physical coordinates of each grid node; mapping the first core index group that has passed reverse verification to the bottom base grid according to the spatial physical coordinates of the grid nodes to construct the firebed surface feature basis matrix; filling the second core index group into the vertically suspended grid to construct the gas phase combustion space feature matrix; and performing a 3D matrix superposition operation on the firebed surface feature basis matrix and the gas phase combustion space feature matrix to obtain the three-dimensional integrated combustion state space matrix of the furnace.

[0053] The furnace 3D mesh spatial discretization model is a digital model that divides the continuous physical space of the furnace into multiple regular mesh units, realizing the discretization representation of the continuous combustion space and providing a unified carrier for the spatial integration of multi-source state parameters. The bottom base mesh corresponds to the two-dimensional planar region where the solid material layer above the grate is located, and is specifically used to carry state parameters related to solid combustion. The vertically suspended mesh corresponds to the gas phase space of the furnace above the solid material layer, and is divided into layers according to the height of the furnace, and is specifically used to carry state parameters related to gas phase combustion. The mesh node is the center point of each mesh unit and is the basic unit for storing combustion state parameters. The spatial physical coordinates refer to the actual position of the mesh node in the three-dimensional physical coordinate system of the furnace, ensuring that all combustion parameters correspond one-to-one with the actual spatial position in the furnace. The firebed surface feature matrix is ​​a two-dimensional data structure formed by mapping the first core index group to the bottom base mesh, which completely contains the solid combustion state information at each position on the firebed surface. The gas phase combustion space feature matrix is ​​a three-dimensional data structure formed by filling the vertically suspended mesh with the second core index group, which completely contains the combustion state information at each height and position in the gas phase space. 3D matrix superposition is a computational method that couples and fuses two-dimensional solid-phase characteristic matrices with three-dimensional gas-phase characteristic matrices according to their spatial positional relationship, thereby realizing the spatial correlation and integration of solid-gas two-phase combustion parameters.

[0054] Specifically, the three-dimensional mesh spatial discretization model of the furnace is established based on the actual geometric dimensions of the incinerator, and the mesh resolution is set according to the accuracy requirements of combustion state monitoring. The bottom base mesh is uniformly divided along the length and width of the grate, while the vertically suspended mesh is layered along the height of the furnace, with each layer maintaining the same planar resolution as the bottom base mesh. The spatial physical coordinates of the mesh nodes are directly calculated based on the three-dimensional physical coordinate system of the furnace, ensuring the accuracy of spatial positioning.

[0055] Specifically, the mapping process for the first core indicator group begins by distributing the average parameters of each grate zone to all nodes of the underlying base grid using bilinear interpolation, forming a continuous distribution of firebed surface parameters. For boundary areas such as grate edges, extrapolation is used to supplement the parameter values ​​of boundary nodes, ensuring the integrity of the firebed surface features. After mapping, the rationality of all node parameters is verified, and outliers are removed.

[0056] Specifically, the filling process of the second core indicator group first expands the one-dimensional gas phase parameters into a three-dimensional spatial distribution based on the vertical variation law of carbon monoxide concentration gradient distribution and volatile matter release rate. Combined with the basic characteristics of the furnace flow field, the gas phase parameters at different height layers are adjusted to ensure that the parameter distribution conforms to the actual flue gas flow law. After filling is completed, all node parameters of the vertically suspended grid are smoothed to eliminate abrupt changes.

[0057] Specifically, the 3D matrix superposition operation is not simply a mathematical calculation, but rather based on the physical coupling relationship of solid-gas two-phase combustion. It associates the solid-phase parameters of each bottom-level base grid node with the gas-phase parameters of all gas-phase grid nodes in the corresponding vertical direction. Each 3D grid node obtained after the operation simultaneously contains information on the solid combustion state and the incomplete gas-phase combustion state at that location, achieving an organic fusion of the solid-gas two-phase combustion states.

[0058] Specifically, each node of the three-dimensional integrated combustion state space matrix in the furnace stores seven core parameters: strata combustion temperature, solid material layer thickness, waste calorific value, combustion rate, chemically incomplete combustion component, carbon monoxide concentration, and volatile matter release rate. The matrix data is updated once per second, reflecting the dynamic changes in the combustion state inside the furnace in real time and providing comprehensive and accurate input data for the intelligent decision-making module.

[0059] Using the embodiments of the present invention, the three-dimensional mesh spatial discretization model of the furnace divides the furnace into... Spatial grids (e.g.) (Number of nodes). The bottom base grid corresponds to the firebed surface at Z=1, bearing the combustion temperature, bed thickness, and calorific value of the waste. The vertical suspended grid corresponds to the gas phase space from Z=2 to Z=30, bearing CO concentration, volatile matter release rate, etc. Due to the use of feature tensor fusion under spatial coordinate alignment, the intelligent decision module can accurately locate the local CO concentration exceeding the standard and the concentrated release of volatile matter at coordinates such as (X=5, Y=3, Z=12), thereby precisely adjusting the air volume of the local secondary air nozzles in the X=5, Y=3 region to achieve spatial targeted control.

[0060] In a preferred embodiment of the present invention, solid and gas phase parameters are carried out by grid discretization and layering, and data fusion is completed by combining 3D matrix superposition operation. The generated 3D state matrix can completely present the combustion conditions in the entire furnace, providing a reliable basis for precise control.

[0061] Preferably, the boundary condition correction includes: locking the maximum relative deviation that is greater than or equal to a preset deviation threshold, and identifying the gas phase combustion parameters corresponding to the maximum relative deviation; calculating the radiation correction coefficient and the geometric translation of the spatial alignment boundary by combining the preset sensitivity matrix; and using the correction coefficient and the translation to quantitatively correct the first-order spatially fused radiation heat transfer boundary and spatial alignment boundary.

[0062] Among these, the maximum relative deviation is the largest relative deviation among all gas-phase combustion parameters and is the primary basis for positioning calculation errors. The preset sensitivity matrix is ​​a pre-established correlation data table used to reflect the impact of gas-phase parameter fluctuations on various boundary conditions in spatial fusion. The radiation correction coefficient is a parameter specifically adjusted to adjust the radiation heat transfer calculation rules, used to compensate for radiation signal loss caused by flue gas and suspended particulate matter. The spatial alignment boundary geometric translation adjusts the position value of the coordinate matching reference, used to correct positional offsets generated during multi-source spatial data docking. The radiation heat transfer boundary refers to the constraint rules related to heat radiation propagation and energy conversion in first-order spatial fusion. The spatial alignment boundary refers to the position range and reference rules set when matching coordinates of different spatial data. Quantitative correction adjusts the original constraint rules based on calculated precise values, ensuring the standardization and accuracy of the adjustment process.

[0063] Specifically, the system automatically traverses all gas phase parameter deviation data involved in reverse verification, quickly filters out the largest relative deviation, and simultaneously locks the gas phase combustion parameters and corresponding furnace area corresponding to the deviation, completing the rapid location of the error source, saving the step of manual investigation one by one, and improving the overall processing efficiency.

[0064] Specifically, the preset sensitivity matrix is ​​a partial derivative matrix, which stores the rate of change of the partial derivatives of specific gas-phase combustion parameters, such as CO concentration deviation, with respect to various boundary conditions (such as radiation attenuation coefficient and geometric translation). This is achieved through matrix multiplication. The quantitative correction value was calculated.

[0065] Specifically, the system uses a radiation correction coefficient to adjust the radiation heat transfer boundary within the first-order spatial fusion, recalibrates the thermal radiation attenuation ratio and energy conversion standard, offsets the radiation measurement deviation caused by changes in gaseous suspended matter and flue gas composition inside the furnace, and restores the true radiation and temperature characteristics of the firebed surface.

[0066] Specifically, the system adjusts the spatial alignment boundary according to the calculated geometric translation amount, corrects the positional deviation between infrared imaging data and acoustic temperature measurement data in the coordinate mapping and spatial registration process, and enables the two types of spatial data to achieve accurate matching across the entire domain again.

[0067] Specifically, after all boundary conditions are corrected, the system automatically reverts to the first-order spatial fusion process to recalculate the first core indicator group. Then, it restarts the coupled state deduction and reverse verification process, forming a complete closed-loop correction mechanism. This process iterates repeatedly until the verification results meet the requirements. This scheme prioritizes correcting the boundary conditions of first-order spatial fusion. If the deviation still exceeds the threshold after multiple closed-loop iterations, the system automatically triggers the self-check process of the sensor and temporal fusion algorithm and outputs an anomaly alarm.

[0068] For example, in the reverse verification, the system statistically determined that the relative deviation of the combustion rate was 2.1%, the relative deviation of the volatile matter release rate was 3.2%, and the relative deviation of the carbon monoxide concentration was 18.6%. The system locked the maximum relative deviation of 18.6% and determined the corresponding specific gas-phase combustion parameter as carbon monoxide concentration. Subsequently, the system retrieved the preset sensitivity matrix for calculation, obtaining a radiation correction coefficient of 0.92 and a spatial alignment boundary geometric translation of 2.4 cm in the horizontal direction. The system automatically adopted this set of values ​​to perform overall correction of the radiation heat transfer boundary parameters in the first-order spatial fusion, compensating for the radiation attenuation effect caused by fly ash in the furnace. At the same time, it adjusted the coordinate matching benchmark according to the 2.4 cm translation to correct the positional offset between the infrared image and the acoustic temperature measurement data. After the boundary conditions were corrected, the system re-run the first-order spatial fusion process to calculate the first core index group and carried out the coupled reverse verification again. Finally, the relative deviation of the carbon monoxide concentration was reduced to 3.7%, and the deviations of other parameters were also less than the preset deviation thresholds, and the verification was successfully passed.

[0069] In a preferred embodiment of the present invention, the gas phase parameter with the largest positioning deviation is used in combination with the sensitivity matrix to quantitatively correct the radiation and spatial alignment boundary, thereby accurately eliminating measurement and registration errors, and relying on a closed-loop iterative mechanism to ensure that the fused data meets the verification standards.

[0070] Preferably, the generation of the combustion feedforward control command includes: extracting the values ​​of each grid node in the three-dimensional integrated combustion state space matrix of the furnace, obtaining the positive distribution coefficient of the firebed, the spatial coordinates of the concentrated release zone of volatile matter, and the predicted value of the total heat load of the furnace through spatial integration and deviation statistical calculation; comparing the positive distribution coefficient of the firebed and the solid material layer thickness of each zone with the preset standard firebed shape, calculating the feedforward correction increment of the grate speed of each section, so as to generate the grate speed adjustment command; calculating the local theoretical oxygen demand and secondary air replenishment of the corresponding area based on the spatial coordinates of the concentrated release zone of volatile matter, the real-time volatile matter release rate, and the chemical incomplete combustion component, and combining the predicted value of the overall total heat load of the furnace, performing multivariate decoupling calculation on the mutually coupled control parameters such as the global air volume, the zone air chamber pressure, and the zone secondary air volume, respectively obtaining the feedforward control target values ​​of the primary air chamber pressure, the ratio of primary air volume to secondary air volume of each section; and combining the grate speed adjustment command and the air volume adjustment command into the combustion feedforward control command.

[0071] Among them, the grid node values ​​are all the combustion state parameters stored at each spatial point within the three-dimensional integrated combustion state space matrix of the furnace, serving as the original basis for data analysis and control calculations. Spatial integration is the integration calculation of parameters from all grid nodes within a defined spatial range, thereby characterizing the overall combustion characteristics of a local area or even the entire furnace. Deviation statistics is a statistical method that compares the combustion parameters obtained from real-time acquisition and calculation with the system's preset standard parameters to quantify the degree of difference between the two. The firebed skew distribution coefficient is an indicator that quantifies the deviation of the firebed combustion state on the grate from the ideal state, and can intuitively reflect the degree of uneven combustion distribution. The spatial coordinates of the concentrated volatile matter release zone are the three-dimensional position information corresponding to the area where a large amount of volatile matter is released inside the furnace, serving as an important basis for achieving precise air distribution in zones. The predicted total heat load of the furnace is the overall heat release capacity of the furnace obtained by combining real-time combustion conditions, reflecting the overall load level of the equipment in subsequent stages. The standard firebed morphology is the ideal distribution state formed by parameters such as grate material layer thickness and combustion temperature during stable and efficient operation of the incinerator, serving as a reference benchmark for grate control. The grate speed feedforward correction increment is the value that needs to be adjusted based on the original grate operating speed, used to correct the waste residence time and combustion rhythm in each section. Local theoretical oxygen demand (RDD) is the total amount of air required to ensure the full oxidation and combustion of combustible gases in areas where volatile matter is concentrated. Secondary air supplement is the additional air volume added to enhance local combustion on top of the existing secondary air supply. Multivariable decoupling calculation separates and independently calculates multiple interrelated control parameters, eliminating mutual interference between parameters to obtain precise and controllable control values. The feedforward control target value is the standard control value for the air pressure, primary air volume, and secondary air volume of each air chamber, determined through calculation.

[0072] Specifically, the system extracts parameter data from all grid nodes in the three-dimensional integrated combustion state space matrix of the furnace. It then divides the data into different statistical ranges according to grate zones and furnace height, performing spatial integration and deviation statistics to calculate the positive distribution coefficient of the firebed, the spatial coordinates of the concentrated volatile matter release zone, and the predicted total furnace heat load. These three calculations are performed simultaneously to ensure the real-time performance of the control logic. The system compares parameters such as the solid material layer thickness and combustion temperature of each zone with the standard firebed morphology, and uses the positive distribution coefficient of the firebed to determine whether combustion is lagging or excessive in each area. Based on this, it calculates the feedforward correction increment corresponding to each grate speed segment, and generates corresponding grate speed adjustment commands based on these increments to specifically adjust the waste travel speed and material residence time.

[0073] Specifically, the system calculates the local theoretical oxygen demand of the corresponding area based on the spatial coordinates of the concentrated volatile release zone, the real-time volatile release rate, and the amount of incomplete chemical combustion. Combined with the furnace flow field distribution, it determines a reasonable amount of secondary air replenishment to ensure that local combustible gases can be fully combusted and reduce the generation of incomplete combustion products.

[0074] Specifically, because primary and secondary air volumes, air chamber pressure, and total furnace heat load are interconnected, the system performs multivariate decoupling calculations to eliminate the correlation interference between various parameters, obtaining the feedforward control target values ​​for air chamber pressure, primary air volume ratio, and secondary air volume ratio for each section, ensuring that the regulation of each air path parameter is independent and effective. The system integrates the generated grate speed adjustment command and air volume adjustment command to form a complete combustion feedforward control command and outputs it outward. Relying on the pre-issued control commands to change the equipment operating state, it smooths combustion fluctuations from the source and reduces the risk of pollutant emissions.

[0075] In summary, the multi-dimensional information fusion combustion monitoring system and method for waste incinerators provided by this invention simultaneously acquires multi-dimensional data from vision, acoustics, thermal processes, and flue gas chemistry through multiple acquisition units. First-order spatial fusion and first-order temporal fusion are used to calculate the solid-phase and gas-phase incomplete combustion state parameters. Coupled state deduction enables cross-reverse verification of the two sets of parameters, and closed-loop correction of fusion boundary conditions can be performed based on deviations, effectively eliminating interference from single sensors, data temporal misalignment, and spatial registration errors. Furthermore, a three-dimensional comprehensive combustion state space matrix of the furnace is constructed using grid discretization and dimensional matrix superposition operations, fully restoring the solid-gas two-phase combustion distribution characteristics throughout the furnace. Finally, feedforward control commands are generated based on matrix data calculations to coordinate the adjustment of grate speed, primary and secondary air volumes, and wind chamber pressure. This not only achieves comprehensive, high-precision real-time perception of the combustion state of the waste incinerator but also intervenes in combustion anomalies in advance, significantly smoothing furnace temperature fluctuations, improving combustion completeness, effectively suppressing excessive emissions of flue gas pollutants, and enhancing the overall operational stability and combustion efficiency of the incineration equipment.

[0076] The above description is merely a preferred embodiment of the technical solution of the present invention and is not intended to limit the scope of protection of the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the scope of protection of the present invention.

Claims

1. A multi-dimensional information fusion combustion monitoring system for waste incinerators, characterized in that, include: The multidimensional basic data acquisition module is used to acquire in real time two-dimensional radiometric grayscale images, acoustic wave transit time signals, thermal process data, and flue gas chemical data at the furnace outlet of the waste incinerator. The physical feature fusion module is used to perform first-order spatial fusion of the two-dimensional radiation grayscale image and the acoustic wave transit time signal to calculate the first core index group to characterize the current combustion state of the solid material layer on the grate. The process feature fusion module is used to perform first-order time-series fusion of the thermal process data and flue gas chemical data to calculate the second core index group to characterize the gas phase incomplete combustion state. The verification and secondary fusion module is used to perform reverse verification on the first core indicator group and the second core indicator group according to the coupling state inference strategy. If the reverse verification passes, a three-dimensional integrated combustion state space matrix of the furnace is generated based on the first core indicator group and the second core indicator group. If the reverse verification fails, the boundary conditions in the first-order spatial fusion are corrected in reverse according to the second core indicator group, the first core indicator group is recalculated, and the reverse verification is performed again. The intelligent decision-making module is used to generate combustion feedforward control commands based on the three-dimensional integrated combustion state space matrix of the furnace to adjust the air volume and grate speed in advance.

2. The multi-dimensional information fusion combustion monitoring system according to claim 1, characterized in that, The multidimensional basic data acquisition module includes: The spatial vision acquisition unit includes an infrared camera for acquiring two-dimensional radiometric grayscale images of the grate firebed surface; The internal acoustic acquisition unit includes a sound wave generator and an acoustic receiving sensor, used to acquire the sound wave transit time signal; The thermal process data acquisition unit is used to collect thermal process data, including primary air volume, secondary air volume, wind chamber air pressure, pusher frequency, grate speed and main steam flow. The flue gas chemistry acquisition unit is used to collect flue gas chemistry data, including oxygen concentration and carbon monoxide concentration.

3. A multi-dimensional information fusion combustion monitoring method for waste incinerators, applied to the multi-dimensional information fusion combustion monitoring system according to any one of claims 1 to 2, characterized in that, include: Step S1: Acquire two-dimensional radiometric grayscale images, acoustic wave transit time signals, thermal process data, and flue gas chemical data at the furnace outlet of the waste incinerator. Step S2: Perform first-order spatial fusion of the two-dimensional radiometric grayscale image and the acoustic wave transit time signal to calculate the first core index group; The thermal process data and flue gas chemical data are fused in a first-order time series to calculate the second core index group; Step S3: Based on the coupling state deduction strategy, perform reverse verification on the first core indicator group and the second core indicator group; If the reverse verification passes, a three-dimensional integrated combustion state space matrix of the furnace is generated based on the first core indicator group and the second core indicator group; if the reverse verification fails, the boundary conditions in the first-order spatial fusion are corrected in reverse based on the second core indicator group, the first core indicator group is recalculated, and the reverse verification is performed again until the verification passes. Step S4: Generate combustion feedforward control commands based on the three-dimensional integrated combustion state space matrix of the furnace to adjust the air volume and grate speed in advance, smooth furnace temperature fluctuations, and suppress excessive emissions of flue gas pollutants.

4. The multi-dimensional information fusion combustion monitoring method according to claim 3, characterized in that, The first-order spatial fusion includes: A three-dimensional physical coordinate system for the furnace is established, and a two-dimensional radiation grayscale image is mapped onto the three-dimensional physical coordinate system for the furnace to calculate the geometric boundary of the fire line on the firebed plane. Within the three-dimensional physical coordinate system of the furnace, the surface radiation features of the two-dimensional radiation grayscale image are aligned with the cross-sectional temperature field reconstructed from the acoustic wave time-of-flight signal at the same spatial scale. Based on the surface radiation characteristics and cross-sectional temperature field after spatial alignment, combined with the conservation relationship of radiation heat transfer, the interference of gaseous suspended matter on the surface radiation of the firebed is eliminated, and the layered combustion temperature distribution on the surface of the firebed is calculated. The firebed movement speed is calculated based on the changing characteristics of the fire line geometric boundary, and the solid material layer thickness and waste calorific value of each zone are obtained based on the layer combustion temperature distribution. These are then combined to generate the first core index group.

5. The multi-dimensional information fusion combustion monitoring method according to claim 3, characterized in that, The first-order temporal fusion includes: Based on the preset flue gas flow lag time model, the thermal process data and flue gas chemical data are aligned on the time axis. Multivariate nonlinear correlation analysis was performed on the aligned main steam flow rate, oxygen concentration, and carbon monoxide concentration to calculate the combustion rate index and the chemically incomplete combustion component. Combining the furnace flow field distribution pattern, based on the combustion rate, the chemical incomplete combustion component, and the temporal fluctuation characteristics of the wind chamber pressure and the pusher frequency, the furnace carbon monoxide concentration gradient distribution and real-time volatile matter release rate are calculated, and the second core index group is generated.

6. The multi-dimensional information fusion combustion monitoring method according to claim 3, characterized in that, The coupling state deduction strategy includes: Match the first core indicator group with the preset corresponding relationship library to obtain reference values ​​for gas phase combustion parameters; Calculate the relative deviation between the reference values ​​of each gas phase combustion parameter and the corresponding measured parameters in the second core index group, and determine whether the reverse verification passes based on the preset deviation threshold.

7. The multi-dimensional information fusion combustion monitoring method according to claim 3, characterized in that, The generation of the three-dimensional integrated combustion state space matrix of the furnace includes: A three-dimensional mesh spatial discretization model of the furnace is established, dividing the bottom base mesh and the vertical floating mesh, and determining the spatial physical coordinates of each mesh node; The first core indicator group that has passed reverse verification is mapped to the bottom base grid according to the spatial physical coordinates of the grid nodes to construct the firebed surface feature basis matrix. The second core indicator group is filled into the vertical suspended grid to construct the gas phase combustion spatial feature matrix; The three-dimensional integrated combustion state space matrix of the furnace is obtained by performing a 3D matrix superposition operation on the characteristic basis matrix of the firebed surface and the characteristic matrix of the gas phase combustion space.

8. The multi-dimensional information fusion combustion monitoring method according to claim 3, characterized in that, The correction of the boundary conditions includes: Lock the maximum relative deviation that is greater than or equal to a preset deviation threshold, and identify the gas phase combustion parameters corresponding to the maximum relative deviation; By combining the preset sensitivity matrix, the radiation correction coefficient and the geometric translation of the spatial alignment boundary are calculated; The correction coefficients and translation amounts are used to quantitatively correct the radiation heat transfer boundary and spatial alignment boundary of the first-order spatial fusion.

9. The multi-dimensional information fusion combustion monitoring method according to claim 3, characterized in that, The generation of the combustion feedforward control command includes: The values ​​of each grid node in the three-dimensional integrated combustion state space matrix of the furnace are extracted, and the positive distribution coefficient of the fire bed, the spatial coordinates of the concentrated release zone of volatile matter, and the predicted value of the total heat load of the furnace are obtained through spatial integration and deviation statistical calculation. The positive distribution coefficient of the firebed and the thickness of the solid material layer in each zone are compared with the preset standard firebed shape to calculate the feedforward correction increment of the grate speed in each section, so as to generate the grate speed adjustment command. Based on the spatial coordinates of the concentrated volatile release zone, the real-time volatile release rate, and the chemically incomplete combustion component, the local theoretical oxygen demand and secondary air replenishment of the corresponding region are calculated. Combined with the predicted value of the overall furnace heat load, multivariate decoupling calculations are performed on the coupled control parameters such as global air volume, zone air chamber pressure, and zone secondary air volume to obtain the feedforward control target values ​​of primary air chamber pressure and the ratio of primary air volume to secondary air volume for each section. Air volume adjustment commands are generated based on the feedforward control target values. The grate speed adjustment command and the air volume adjustment command are combined into the combustion feedforward control command.

Images