Intelligent air pressure control method and system for garbage incinerator

Abstract

The invention provides an intelligent air pressure control method and system for a garbage incinerator, and relates to the technical field of garbage incineration. The method comprises the steps that air door opening degree data and air pressure control parameter data of the garbage incinerator are collected, the target air outlet amount is obtained through calculation, and a first air pressure feedback signal is generated; combustion monitoring data of a target combustion area in the garbage incinerator are obtained, a plurality of first time sequence distance parameters between any adjacent sub-areas are calculated, and the combustion state of the target combustion area is determined according to the combustion state dynamic analysis model; generating first target reference data corresponding to the combustion monitoring data based on the expected combustion reference data, and generating second target reference data corresponding to the combustion monitoring data according to the combustion scene analysis model; and the first air pressure feedback signal is corrected to generate a second air pressure feedback signal, a target fan frequency control signal is generated, and air pressure control is conducted on the garbage incinerator. According to the invention, the wind pressure of the garbage incinerator can be accurately controlled.

Description

Technical Field

[0001] The present invention relates to the technical field of waste incineration, and particularly to an intelligent air pressure control method and system for a waste incinerator. Background Art

[0002] In the control process of a waste incinerator, air pressure control is a key factor to ensure combustion efficiency and stability. Some traditional air pressure control methods rely relatively heavily on feedback control based on sensors, and adjust by monitoring parameters such as air pressure and temperature in the combustion furnace in real time.

[0003] In this process, first, the collected air pressure data may not accurately represent the actual air pressure situation in the incinerator because there may be local unevenness, and there is a certain delay in the feedback data of the sensor, resulting in a lag in air pressure adjustment during the combustion process. Second, the control process mostly relies on the air pressure value set according to empirical data to achieve air pressure regulation. In the actual combustion process, different regions of the combustion furnace may still have subtle dynamic imbalances, such as local overheating and air flow disorder, due to factors such as fuel distribution and equipment structure, even when macroscopic indicators such as temperature and flue gas composition are normal. These imbalance phenomena may be difficult to detect simply through single-sensor data in the short term, and are likely to affect the combustion efficiency over time, resulting in problems such as deterioration of emission indicators. In this case, these dynamic differences easily lead to the target value determined based on empirical data may not fully adapt to the real-time changing combustion state, thereby affecting the accuracy of the control effect. Summary of the Invention

[0004] An intelligent air pressure control method and system for a waste incinerator proposed by the present invention aims to solve at least one of the technical problems existing in the above background art.

[0005] To achieve the above object, the first aspect of the present invention provides an intelligent air pressure control method for a waste incinerator, including: Collecting the damper opening data and air pressure control parameter data of the waste incinerator and calculating the target air output, and generating a first air pressure feedback signal according to the first target air pressure data and the target air output; Obtaining the combustion monitoring data of the target combustion area in the waste incinerator, including the real-time monitoring data of each sub-area in the target combustion area regarding multiple combustion state parameters; Extract the first state time series feature vectors of each sub-region with respect to each combustion state parameter from the combustion monitoring data, calculate the first time series distance parameters between any adjacent sub-regions with respect to each combustion state parameter, analyze the multiple first time series distance parameters according to the combustion state dynamic analysis model, determine the first combustion imbalance parameters between any adjacent sub-regions, and determine the combustion state of the target combustion region according to the multiple first combustion imbalance parameters. If the combustion state of the target combustion region is in an abnormal combustion state, determine the abnormal state parameters in the combustion monitoring data, generate the first target reference data corresponding to the combustion monitoring data based on the expected combustion reference data, and analyze the first target reference data according to the combustion scenario analysis model to generate the second target reference data corresponding to the combustion monitoring data. Generate the second target air pressure data of the target combustion region according to the second target reference data, correct the first air pressure feedback signal according to the second target air pressure data to generate the second air pressure feedback signal, and generate the target fan frequency control signal based on the second air pressure feedback signal to control the air pressure of the waste incinerator.

[0006] In one embodiment, analyzing the multiple first time series distance parameters according to the combustion state dynamic analysis model, determining the first combustion imbalance parameters between any adjacent sub-regions, and determining the combustion state of the target combustion region according to the multiple first combustion imbalance parameters includes: Query the combustion state parameter transfer feature vectors between any adjacent sub-regions according to the regional combustion state association list, and perform transfer analysis on the first time series distance parameters between adjacent sub-regions with respect to each combustion state parameter according to the combustion state parameter transfer feature vectors, including weighting the first time series distance parameters between adjacent sub-regions with respect to each combustion state parameter according to the combustion state parameter transfer feature vectors, and calculating the first combustion imbalance parameters between adjacent sub-regions. Query the combustion state dynamic range between any adjacent sub-regions according to the regional combustion state association list, and determine whether the first combustion imbalance parameters between any adjacent sub-regions all conform to the corresponding combustion state dynamic range. If so, mark the target combustion region as in a normal combustion state; otherwise, mark the target combustion region as in an abnormal combustion state.

[0007] In one embodiment, for the regional combustion state association list, it further includes: The regional combustion state association list is generated by obtaining and analyzing multiple groups of historical incineration reference data of the waste incinerator. After obtaining multiple groups of historical incineration reference data, extract the combustion state feature vectors corresponding to multiple moments in each group of historical incineration reference data, calculate the state distance parameters between any adjacent sub-regions at multiple moments, and construct the state distance time series between any adjacent sub-regions. Analyze multiple combustion state parameters between adjacent sub-regions according to the state distance time series, determine the contribution ratio data of each combustion state parameter to the state distance time series, and generate a combustion state parameter transfer feature vector between adjacent sub-regions according to the contribution ratio data; Extract the second state time series feature vector of each sub-region for each combustion state parameter from each set of historical incineration reference data, calculate the second time series distance parameter between any adjacent sub-regions for each combustion state parameter, and calculate the second combustion imbalance parameter between adjacent sub-regions in each set of historical incineration reference data according to the combustion state parameter transfer feature vector. Determine the combustion state dynamic range between any adjacent sub-regions according to multiple second combustion imbalance parameters; Construct a regional combustion state association list according to the combustion state parameter transfer feature vector and the combustion state dynamic range between adjacent sub-regions.

[0008] In one embodiment, determine the abnormal state parameters in the combustion monitoring data, and generate the first target reference data corresponding to the combustion monitoring data based on the expected combustion reference data, including: For the process of weighting the first time series distance parameter between adjacent sub-regions for each combustion state parameter according to the combustion state parameter transfer feature vector, count the cumulative distance value of each combustion state parameter, and record the combustion state parameter with the cumulative distance value greater than the preset cumulative threshold as the abnormal state parameter; Calculate the mean value of each sub-region in the combustion monitoring data for each combustion state parameter, construct the first state matrix corresponding to the combustion monitoring data, perform parameter substitution on the abnormal state parameters in the first state matrix according to the expected combustion reference data, generate the second state matrix corresponding to the first state matrix, and obtain the first target reference data corresponding to the combustion monitoring data.

[0009] In one embodiment, analyze the first target reference data according to the combustion scenario analysis model to generate the second target reference data corresponding to the combustion monitoring data, including: Match the first target reference data with multiple combustion scenario clusters, determine the target scenario cluster associated with the first target reference data, determine the third state matrix according to the target scenario cluster, and update the data of multiple abnormal state parameters in the first target reference data according to the third state matrix to generate the second target reference data corresponding to the combustion monitoring data; For multiple combustion scenario clusters, after obtaining multiple sets of historical incineration reference data of the waste incinerator, extract multiple regional state vectors of each set of historical incineration reference data, construct a regional state matrix corresponding to each set of historical incineration reference data, perform data clustering on the multiple sets of historical incineration reference data based on the regional state matrix to generate multiple combustion scenario clusters, and generate a target combustion state matrix for each scenario cluster according to the multiple regional state matrices included in each combustion scenario cluster.

[0010] In one embodiment, generating a first wind pressure feedback signal according to the first target wind pressure data and the target air volume includes: Determine a first air volume difference parameter based on the first target wind pressure data and the target air volume, obtain the fan performance curve of the waste incinerator, process the first air volume difference parameter through the fan performance curve to generate a first wind pressure compensation parameter, and generate a first wind pressure feedback signal according to the first wind pressure compensation parameter.

[0011] In one embodiment, correcting the first wind pressure feedback signal according to the second target wind pressure data to generate a second wind pressure feedback signal includes: Determine a second air volume difference parameter according to the second target wind pressure data and the target air volume, generate wind pressure dynamic correction data by analyzing the second air volume difference parameter according to the fan performance curve of the waste incinerator, and correct the first wind pressure feedback signal according to the wind pressure dynamic correction data to generate a second wind pressure feedback signal.

[0012] The second aspect of the present invention provides an intelligent wind pressure control system for a waste incinerator, which is used to implement the above-mentioned intelligent wind pressure control method for a waste incinerator, and includes: A first wind pressure correction module, configured to collect the damper opening data and the wind pressure control parameter data of the waste incinerator and calculate the target air volume, and generate a first wind pressure feedback signal according to the first target wind pressure data and the target air volume; A combustion monitoring and analysis module, configured to obtain the combustion monitoring data of the target combustion area in the waste incinerator, including the real-time monitoring data of each sub-area in the target combustion area regarding multiple combustion state parameters, extract the first state time series feature vector of each sub-area regarding each combustion state parameter from the combustion monitoring data, calculate the first time series distance parameter between any adjacent sub-areas regarding each combustion state parameter, analyze the multiple first time series distance parameters according to the combustion state dynamic analysis model, determine the first combustion imbalance parameter between any adjacent sub-areas, and determine the combustion state of the target combustion area according to the multiple first combustion imbalance parameters; An abnormal combustion analysis module is used to determine abnormal state parameters in combustion monitoring data if the combustion state in the target combustion area is in an abnormal combustion state, generate first target reference data corresponding to the combustion monitoring data based on expected combustion reference data, and analyze the first target reference data according to a combustion scenario analysis model to generate second target reference data corresponding to the combustion monitoring data; A second air pressure correction module is used to generate second target air pressure data for the target combustion area according to the second target reference data, correct the first air pressure feedback signal according to the second target air pressure data to generate a second air pressure feedback signal, generate a target fan frequency control signal based on the second air pressure feedback signal, and perform air pressure control on the waste incinerator.

[0013] The present invention has the following beneficial effects: By combining the damper opening data and the fan performance curve, the present invention dynamically adjusts the air pressure feedback signal, effectively improves the air pressure control accuracy, and ensures the air pressure stability during the incineration process; starting from the perspective of the transfer of the combustion state between different areas, the present invention performs abnormal detection on the combustion state in the combustion area, and through a combustion state dynamic analysis model and a combustion scenario analysis model, performs real-time analysis on the combustion monitoring data, accurately identifies the imbalance in the combustion area, analyzes the combustion scenario using historical incineration reference data, accurately determines the target value of each combustion state parameter, and generates corrected target air pressure and air volume data through optimization calculation. According to the characteristics of different combustion scenarios, the present invention intelligently adjusts the control strategy to realize intelligent adjustment of the air pressure of the waste incinerator based on real-time combustion state changes, and improves the accuracy of air pressure control. Description of the Drawings

[0014] Figure 1 It is a schematic flow chart of an intelligent air pressure control method for a waste incinerator according to the present invention.

[0015] Figure 2 It is a schematic structural diagram of an intelligent air pressure control system for a waste incinerator according to the present invention. Detailed Embodiments

[0016] In order to enable those skilled in the art to better understand the technical solutions in the present invention, the following will clearly and completely describe the technical solutions in the embodiments of the present invention with reference to the accompanying drawings in the embodiments of the present invention. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not used to limit the present invention.

[0017] The embodiments of the present invention provide an intelligent air pressure control method for a waste incinerator. Please refer to Figure 1 and the method includes the following steps: Step S1: Collect the air damper opening data and the air pressure control parameter data of the waste incinerator, calculate the target air output, and generate a first air pressure feedback signal based on the first target air pressure data and the target air output.

[0018] In this embodiment, the air damper opening data is used to indicate the opening degree of the intake air damper, which directly affects the magnitude of the air volume. The air pressure control parameter data specifically refers to the specific working parameters of the fan, such as the operating state data, etc. Based on the air damper opening data and the air pressure control parameter data, the target air output can be calculated, that is, the ideal air volume theoretically output by the fan under the current operating conditions.

[0019] On the basis of obtaining the target air output, combined with the first target air pressure data, a first air pressure feedback signal can be generated. Specifically, first determine a first air volume difference parameter based on the first target air pressure data and the target air output. The first target air pressure data can specifically be the air volume data required for waste incineration determined based on empirical knowledge. The first air volume difference parameter characterizes the difference between the predicted theoretical data and the expected required data. Then, obtain the fan performance curve of the waste incinerator, and process the first air volume difference parameter through the fan performance curve to generate a first air pressure compensation parameter. The fan performance curve is usually provided by the fan manufacturer and can reflect the relationship between the air volume and the air pressure output by the fan under different working conditions. Through the fan performance curve, the required air pressure adjustment amount of the fan under the given air volume difference condition can be understood, that is, calculate the first air pressure compensation parameter, which is used to adjust the air pressure of the system to compensate for the air volume difference, so as to generate a first air pressure feedback signal as the input of the fan frequency control signal. Through this method, the fan can adjust the air pressure according to the gap between the target air volume and the actual air volume, ensuring more accurate air pressure control of the system.

[0020] It should be noted that some air pressure control systems are mostly based on the air volume data collected by sensors and combined with the set target value for PID control. However, in this process, the air pressure data collected by the sensors may not accurately represent the actual air pressure situation in the incinerator because the air pressure distribution may be uneven during the waste incineration process. For the sensors arranged at non-air damper outlet positions, although their data can better reflect the air volume in the actual combustion area, there may be a certain delay and they may be affected. In this case, by collecting the air damper switch signal, obtaining the air damper opening data and the working data of the fan, predicting the theoretically achievable air volume in advance, and combining it with the conventional PID control signal as a feedforward signal to improve the stability of the air pressure control.

[0021] Step S2: Obtain the combustion monitoring data of the target combustion area in the waste incinerator, extract the first state time-series feature vectors of each sub-area regarding each combustion state parameter from the combustion monitoring data, and calculate the first time-series distance parameters between any two adjacent sub-areas regarding each combustion state parameter.

[0022] In this embodiment, the combustion monitoring data of the target combustion area includes the real-time monitoring data of each sub-area in the target combustion area regarding multiple combustion state parameters. The target combustion area refers to the key area inside the waste incinerator. The inside of the waste incinerator can be divided into a drying area, a combustion area, and an afterburning area according to the stages of waste combustion. Among them, the drying area is located at the front section of the waste incinerator, where the waste is heated and the moisture is evaporated to prepare conditions for the subsequent combustion stage. The combustion area is located in the middle of the incinerator and is the area where the waste burns most violently. A large amount of oxygen supply is required in this area, and the temperature is relatively high and it is the main action point of the primary air. The afterburning area is located at the rear section of the incinerator and is used for secondary combustion of the unburned solid residues and flue gas to ensure that the combustible components in the waste and flue gas are completely burned. The target combustion area specifically refers to the combustion area. As the key area of waste incineration, in order to more precisely control the combustion efficiency and reduce emissions, it is more common to conduct sub-area monitoring on the combustion area. According to the actual layout positions of the sensors and the combustion characteristics at different positions in the combustion area, it can be further divided into multiple sub-areas. For example, the front section area, that is, the waste ignition stage, is sensitive to temperature and oxygen concentration; the middle section area, with high-temperature and intense combustion, has the highest CO concentration; the rear section area, where the residual residues at the tail end are burned out and the oxygen concentration increases. During the actual monitoring process, for example, for temperature monitoring, multiple monitoring points are generally arranged along the grate, covering the front, middle, and tail end areas.

[0023] The combustion state of each sub-area is crucial for the control of the overall incineration process. The collected combustion monitoring data includes the real-time monitoring data of each sub-area regarding multiple combustion state parameters such as temperature, oxygen concentration, air volume, CO / CO2 concentration, etc. The first state time-series feature vectors of each sub-area regarding each combustion state parameter are extracted from these data, indicating the relationship between the combustion state parameters in the sub-area changing with time. Then, the first time-series distance parameters between any two adjacent sub-areas are calculated. This parameter reflects the difference and transmission relationship of the combustion states between regions by measuring the time-series difference of the combustion state parameters between two adjacent regions. Among them, the calculation of the first time-series distance parameter can adopt distance calculation methods such as dynamic time warping (DTW), etc., which can effectively compare the combustion state characteristics at different time steps and help discover potential anomalies in the combustion process.

[0024] Step S3: Analyze the multiple first time-series distance parameters according to the combustion state dynamic analysis model to determine the first combustion imbalance parameters between any two adjacent sub-areas, and determine the combustion state of the target combustion area according to the multiple first combustion imbalance parameters.

[0025] In this embodiment, for the calculated multiple first timing distance parameters, further analysis is performed through the combustion state dynamic analysis model. The combustion state dynamic analysis model is constructed by analyzing multiple groups of historical incineration reference data of the waste incinerator. By analyzing and capturing the dynamic transfer change law of the combustion state between different sub-regions in the historical data under an ideal combustion condition, it can be used to identify potential combustion imbalance situations during the transfer process of the combustion state between regions.

[0026] In this process, the first combustion imbalance parameter between any two adjacent sub-regions is calculated through the combustion state dynamic analysis model. These parameters reflect the inconsistency or imbalance degree of the combustion state between adjacent sub-regions and can quantify the combustion stability of different regions. Under an ideal combustion state, there is a relatively stable transfer law of the combustion state between different regions. If a potential abnormality occurs in one of the links, such as a low temperature, it may cause changes in the gas composition of the next sub-region. The combustion state of different sub-regions in the target combustion region is determined through the first combustion imbalance parameter, thereby determining whether there is an abnormality in the overall target combustion region.

[0027] Step S4: If the combustion state of the target combustion region is in an abnormal combustion state, determine the abnormal state parameters in the combustion monitoring data, generate the first target reference data corresponding to the combustion monitoring data based on the expected combustion reference data, and analyze the first target reference data according to the combustion scenario analysis model to generate the second target reference data corresponding to the combustion monitoring data.

[0028] In this embodiment, if it is determined that the combustion state of the target combustion area is in an abnormal combustion state, it indicates that although the target combustion area does not show abnormalities in the macroscopic sensor monitoring data for the time being, there may be local abnormalities therein. During the transmission and accumulation in multiple sub-areas, this abnormality is detected, and it is necessary to further analyze the combustion state of the target combustion area in detail. In this process, first, multiple abnormal state parameters in the combustion monitoring data will be determined, that is, multiple key combustion state parameters that cause the first combustion imbalance parameter to be too large. Then, based on the expected combustion reference data, that is, the reference values expected under the ideal combustion state pre-determined based on empirical data, the first target reference data corresponding to the combustion monitoring data will be generated. The first target reference data is relatively close to the combustion state data under the expected state in the current scenario. Then, the first target reference data will be deeply analyzed according to the combustion scenario analysis model. The combustion scenario analysis model is also constructed by analyzing multiple groups of historical incineration reference data of the waste incinerator. By analyzing and capturing the combustion states of different sub-areas in the historical data, the waste incineration is divided into multiple scenarios, and the combustion state-related data representative of each combustion scenario is determined. After obtaining the first target reference data, the combustion scenario corresponding to the first target reference data will be matched, and the reference data of each combustion parameter under the ideal combustion state in the actual combustion scenario corresponding to the combustion monitoring data will be determined according to the representative combustion state of the combustion scenario, and the second target reference data corresponding to the combustion monitoring data will be generated. This process can be regarded as a secondary correction of the current combustion state, correcting the data from the perspective of the combustion scenario to ensure that the data is closer to the ideal state.

[0029] Step S5: Generate the second target air pressure data of the target combustion area according to the second target reference data, correct the first air pressure feedback signal according to the second target air pressure data to generate the second air pressure feedback signal, generate the target fan frequency control signal based on the second air pressure feedback signal, and perform air pressure control on the waste incinerator.

[0030] In this embodiment, for the second target reference data, which involves the air pressure data of the target combustion area under ideal conditions, that is, the second target air pressure data determined by the second target reference data, represents the ideal air pressure required for the target combustion area under the corrected combustion state. The second target air pressure data can be used to correct the first air pressure feedback signal to generate a second air pressure feedback signal. The second air pressure feedback signal will be used as a new input. Its generation process takes into account the situation of local combustion state imbalance existing inside the waste incinerator. Starting from the combustion state transfer law between different sub-regions, it analyzes and mines the potential anomalies in the data, dynamically determines the new air pressure demand data instead of based on fixed static target parameters, and further adjusts the fan frequency through the second air pressure feedback signal, including generating a target fan frequency control signal to control the air pressure of the waste incinerator, so as to ensure that the system air pressure adjustment is more accurate and achieve a more stable combustion state.

[0031] In this process, first determine the second air volume difference parameter according to the second target air pressure data and the target air output, that is, characterize the difference between the dynamically determined air volume demand and the statically set threshold, and incorporate this part of the difference into the correction process of air pressure control. The air pressure dynamic correction data can be generated by analyzing the second air volume difference parameter according to the fan performance curve of the waste incinerator, that is, the relevant correction data for controlling the fan for this part of the difference, and fuse it into the first air pressure feedback signal to correct the first air pressure feedback signal to generate a second air pressure feedback signal. Finally, based on the corrected second air pressure feedback signal, the air pressure of the waste incinerator is controlled. The frequency control signal of the fan will be adjusted according to the combined action of the second air pressure feedback signal and the PID control signal, so as to ensure the stability of the air pressure and the efficiency of the combustion process.

[0032] In one of the embodiments, for step S3, the combustion state dynamic analysis model is used to deeply analyze multiple first time series distance parameters, so as to judge whether the combustion states of each region are balanced and finally determine the combustion state of the target combustion area. The specific implementation process is as follows: Query the combustion state parameter transfer eigenvectors between any adjacent sub-regions according to the regional combustion state association list, and perform transfer analysis on the first time series distance parameters of each combustion state parameter between adjacent sub-regions according to the combustion state parameter transfer eigenvectors, and calculate the first combustion imbalance parameter between adjacent sub-regions.

[0033] In this embodiment, the dynamic combustion state analysis model generates a regional combustion state association list by mining the combustion state transfer rules in multiple groups of historical incineration reference data of the waste incinerator, and conducts transfer analysis on the first time series distance parameters of each combustion state parameter between adjacent sub-regions. The regional combustion state association list first records the combustion state parameter transfer feature vectors between any adjacent sub-regions. The combustion state parameter transfer feature vector is used to represent the importance degree of each parameter in the combustion state parameter transfer process between two adjacent sub-regions. For example, the influence degree of parameters such as the temperature, oxygen concentration, and gas composition of each sub-region on the combustion state of adjacent sub-regions. This is used to assist in analyzing the combustion state transfer characteristics between each sub-region.

[0034] For the calculation of the first combustion imbalance parameter, specifically, the first time series distance parameters of each combustion state parameter between adjacent sub-regions are weighted according to the combustion state parameter transfer feature vector, that is, the eigenvalue corresponding to each combustion state parameter in the combustion state parameter transfer feature vector is used as the weight for weighting the first time series distance parameters of each combustion state parameter. Thus, for the multiple first time series distance parameters of the combustion state parameter between adjacent sub-regions, the first combustion imbalance parameter characterizing the overall combustion state difference is calculated.

[0035] Then, according to the regional combustion state association list, the combustion state dynamic range between any adjacent sub-regions is queried, and it is judged whether the first combustion imbalance parameters between any adjacent sub-regions all conform to the corresponding combustion state dynamic range. If so, the target combustion region is marked as being in a normal combustion state; otherwise, the target combustion region is marked as being in an abnormal combustion state.

[0036] In this embodiment, after calculating the first combustion imbalance parameters between multiple adjacent sub-regions, the combustion state dynamic range of any adjacent sub-regions is further queried according to the regional combustion state association list. The combustion state dynamic range defines the normal change range of the first combustion imbalance parameter between each adjacent sub-region. The first combustion imbalance parameter characterizes the difference in the combustion state transfer between adjacent sub-regions. For each pair of adjacent sub-regions, according to its first combustion imbalance parameter, it is judged whether this parameter is within the corresponding combustion state dynamic range. If the first combustion imbalance parameter is within the allowed dynamic range, it indicates that the combustion state between this pair of adjacent sub-regions still remains within the normal range. Finally, according to the combustion imbalance situation of adjacent sub-regions, the overall combustion state of the target combustion region is comprehensively judged. If the combustion states between all adjacent sub-regions conform to the predetermined dynamic range, the target combustion region can be marked as being in a normal combustion state. Otherwise, if the imbalance parameter of any pair of regions exceeds the dynamic range, it is considered that the target combustion region is in an abnormal combustion state. This indicates that an abnormality has occurred in the combustion state transfer process, and there may be potential abnormal situations that need to be further analyzed.

[0037] In one embodiment, for the above-mentioned regional combustion state association list, it is specifically used to characterize the association of combustion states between different sub-regions in the target combustion region, and can be used to assist in understanding the combustion state of the target combustion region in the waste incinerator. The method for generating the regional combustion state association list by analyzing the combustion state transfer law in multiple sets of historical incineration reference data of the waste incinerator is as follows: After obtaining multiple sets of historical incineration reference data, extract the combustion state feature vectors corresponding to multiple moments in each set of historical incineration reference data, calculate the state distance parameters between any adjacent sub-regions for multiple moments, and construct the state distance time series between any adjacent sub-regions; In this embodiment, multiple sets of historical incineration reference data can specifically be multiple sets of data with a good combustion state. For example, some reference data that are considered to be combusted sufficiently and do not cause excessive waste of combustion resources. For these data, first extract the combustion state feature vectors corresponding to multiple moments in each set of historical incineration reference data to represent the combustion state parameters at each time point, such as temperature, oxygen concentration, CO / CO2 concentration, etc. Then, for each pair of adjacent sub-regions, calculate the state distance parameters corresponding to multiple moments, which represent the difference in combustion states between adjacent sub-regions at the same moment. For example, the Euclidean distance between vectors is used to calculate the state distance parameters. Combining the state distance parameters at multiple moments can construct the state distance time series between each pair of adjacent sub-regions, reflecting the difference in the combustion state change trend between adjacent sub-regions.

[0038] Analyze multiple combustion state parameters between adjacent sub-regions according to the state distance time series, determine the contribution ratio data of each combustion state parameter to the state distance time series, and generate the combustion state parameter transfer feature vector between adjacent sub-regions according to the contribution ratio data; In this embodiment, for the contribution ratio data of combustion state parameters to the state distance time series, it is extracted according to the state distance parameters corresponding to the combustion state parameters at multiple moments. Specifically, for the state distance parameter of the combustion state parameter at any moment, calculate the difference between the state values corresponding to the combustion state parameters in adjacent sub-regions at this moment, and take the ratio of the absolute value of the difference to the state distance parameter between adjacent sub-regions at this moment as the local contribution ratio of the combustion state parameter to the state distance time series at this moment. Then, take the mean value of the local contribution ratios at multiple moments to obtain the contribution ratio data of the combustion state parameter to the state distance time series. After normalizing the contribution ratio data of multiple combustion state parameters to the state distance time series, the corresponding combustion state parameter transfer feature vector can be constructed. By the above method, the role of each combustion state parameter in the diffusion process between combustion regions is accurately evaluated, and the influence of each combustion state parameter is optimized. The combustion state parameter transfer feature vector quantitatively characterizes the relative importance of each combustion state parameter in the process of analyzing transfer differences.

[0039] Extract the second state time series feature vector of each sub-region regarding each combustion state parameter from each group of historical incineration reference data, calculate the second time series distance parameter regarding each combustion state parameter between any adjacent sub-regions, calculate the second combustion imbalance parameter between adjacent sub-regions in each group of historical incineration reference data according to the combustion state parameter transfer feature vector, and determine the combustion state dynamic range between any adjacent sub-regions according to multiple second combustion imbalance parameters.

[0040] In this embodiment, the second state time series feature vector is similar to the previous first state time series feature vector and characterizes the time variation relationship of a certain combustion state parameter. After calculating the second time series distance parameter regarding each combustion state parameter between any adjacent sub-regions, the second combustion imbalance parameter between adjacent sub-regions in each group of historical incineration reference data can be calculated according to the combustion state parameter transfer feature vector determined in the previous steps. These multiple second combustion imbalance parameters between each group of adjacent sub-regions represent the fluctuation characteristics of the transfer law of the combustion state under a relatively perfect combustion state. In this case, the mean value and standard deviation of multiple second combustion imbalance parameters can be calculated, and the combustion state dynamic range for measuring the combustion state imbalance can be determined according to the mean value and standard deviation of the second combustion imbalance parameters. For example, taking the mean value as the core, the range within 1.5 times the standard deviation is recorded as the corresponding combustion state dynamic range. If it exceeds this range, it indicates that in the target combustion region, an abnormality occurs in the transfer process of the combustion state between sub-regions.

[0041] Finally, according to the combustion state parameter transfer eigenvector and the combustion state dynamic range between adjacent sub-regions, a regional combustion state association list is constructed. This list comprehensively reflects the combustion state transfer relationship between different sub-regions, and can be used to analyze the association of combustion states between sub-regions, as well as whether there is an imbalance phenomenon during the transfer process, to assist in analyzing the combustion state differences between each sub-region, and to detect and locate local combustion state anomalies in the target combustion region. Through the above method, an accurate combustion state dynamic analysis framework is constructed. It can accurately evaluate the combustion imbalance between adjacent sub-regions, and judge whether the combustion process is in a normal state through the combustion state dynamic range, providing a solid data support and decision-making basis for subsequent air volume correction.

[0042] In one embodiment, for step S4, the abnormality of the combustion state is identified and corrected by analyzing the difference between the combustion monitoring data and the expected combustion reference data to generate the first target reference data corresponding to the combustion monitoring data. This process specifically includes: For the process of weighting the first time series distance parameter of each combustion state parameter between adjacent sub-regions according to the combustion state parameter transfer eigenvector, the cumulative distance value of each combustion state parameter is statistically calculated, and the combustion state parameter with the cumulative distance value greater than the preset cumulative threshold is recorded as an abnormal state parameter.

[0043] Specifically, the larger the first time series distance parameter of the combustion state parameter, the greater the difference between the two. During the process of weighting according to the combustion state parameter transfer eigenvector, the importance of the time series distance parameter is further optimized according to the importance of the parameter. The first time series distance parameters weighted between multiple adjacent sub-regions of the combustion state parameter are added together to finally obtain the cumulative distance value of each combustion state parameter, which reflects the overall difference degree of the parameter in the data. The process according to the combustion state parameter transfer eigenvector standardizes the single distance measurement index according to the importance, avoiding that a parameter with a lower importance but a larger distance measurement index is misjudged as an abnormal state parameter, and also preventing that an abnormal parameter with a smaller distance measurement index but a higher importance is misjudged as a normal state parameter. In this way, multiple combustion state parameters with anomalies in the combustion monitoring data can be more accurately determined.

[0044] Calculate the mean value of each sub-region in the combustion monitoring data for each combustion state parameter, construct the first state matrix corresponding to the combustion monitoring data, and perform parameter replacement on the abnormal state parameters in the first state matrix according to the expected combustion reference data to generate the second state matrix corresponding to the first state matrix, and obtain the first target reference data corresponding to the combustion monitoring data.

[0045] Specifically, the first state matrix represents the average level of each combustion state parameter in the combustion monitoring data. For the data of the abnormally detected state parameters, parameter replacement is performed on the abnormally detected state parameters in the first state matrix according to the expected combustion reference data. The expected combustion reference data is usually provided by historical data or empirical knowledge and represents the expected values of each combustion state parameter in each region under the ideal combustion state. After replacing the abnormal parameters with the expected values, the second state matrix characterizes the characteristics that the combustion monitoring data should exhibit under the expected state, which is denoted as the first target reference data corresponding to the combustion monitoring data. The extraction of the first target reference data combines the data that appears normal in practice and the expected part of the data as the basis for subsequent air volume correction and air pressure regulation.

[0046] In one embodiment, for step S4, the first target reference data is further analyzed according to the combustion scenario analysis model, and finally the second target reference data corresponding to the combustion monitoring data is generated. This process specifically includes: The first target reference data is matched with multiple combustion scenario clusters to determine the target scenario cluster associated with the first target reference data. According to the target scenario cluster, the third state matrix is determined, and the data of multiple abnormally detected state parameters in the first target reference data is updated to generate the second target reference data corresponding to the combustion monitoring data.

[0047] In this embodiment, the combustion scenario analysis model first matches the first target reference data with multiple combustion scenario clusters. These combustion scenario clusters are generated based on the analysis of multiple groups of historical incineration reference data of the waste incinerator and represent different combustion scenario types. By clustering the data of multiple combustion processes according to the characteristics representing the overall combustion state, categories representing different combustion working conditions are formed, reflecting the relative relationship between the parameters in each region under different combustion states. The first target reference data is matched with multiple combustion scenario clusters to determine the target scenario cluster associated with the reference data. The matching process can be completed by using a similarity measurement method. By calculating the similarity between the first target reference data and the centroid of each combustion scenario cluster, that is, the center point of each cluster, and selecting the most similar cluster as the target scenario cluster. At the same time, for the centroid of the target scenario cluster, as a representative feature, we can update the data of multiple abnormally detected state parameters in the first target reference data, that is, regard the level shown by the abnormally detected state parameters at the centroid of the target scenario cluster as the more conforming expected state under the current combustion state. According to the third state matrix corresponding to the centroid of the target scenario cluster, the data of multiple abnormally detected state parameters in the first target reference data is updated, so as to obtain the second target reference data that more conforms to the combustion scenario currently matched by the combustion monitoring data.

[0048] For multiple combustion scenario clusters, after obtaining multiple sets of historical incineration reference data of the waste incinerator, multiple regional state vectors of each set of historical incineration reference data are extracted. Specifically, the mean value of each combustion state parameter in each sub-region can be used as the average level to construct the regional state vector. Then, the regional state vectors of multiple sub-regions are combined to construct the regional state matrix corresponding to each set of historical incineration reference data. Based on the regional state matrix, data clustering is performed on multiple sets of historical incineration reference data to generate multiple combustion scenario clusters. For example, clustering algorithms such as K-means and DBSCAN are used. After clustering to obtain multiple combustion scenario clusters, a corresponding target combustion state matrix is generated based on the multiple regional state matrices included in the cluster, that is, the third state matrix corresponding to the centroid of the above-mentioned combustion scenario cluster, which is used to characterize the overall level of multiple sets of data in the cluster.

[0049] It should be noted that during each combustion process, due to possible differences in the composition and state of the waste put in, etc., the specific values of the combustion states such as temperature and product gas concentration between sub-regions in the combustion area may be different. Therefore, there are differences in scenarios between different combustion processes. However, these differences usually do not change the transfer law between adjacent regions, that is, the combustion scenario differences, such as generally higher temperature or lower oxygen concentration in a certain process, will affect the specific state of each region. But in the case of complete combustion or a relatively perfect combustion state, the transfer mode of the combustion state, that is, the transfer process and the mutual influence law between adjacent sub-regions, is roughly similar. That is, there are differences in some initial parameters. For example, certain components in the waste are relatively high. But in relatively perfect combustion, such as complete combustion and without wasting too many resources, the gas concentration of some gas products in practice will also be relatively high, and the combustion state is affected, or it causes unreasonable resource allocation, such as providing too much air or incomplete combustion, then the transfer of the combustion state parameters between sub-regions will also be affected. Against this background, considering the complexity of the combustion scenario, preliminary combustion detection is directly carried out based on the transfer law of the combustion state. In the case of abnormalities, it is further located to the specific combustion scenario, and then based on the laws mined from historical data, the relatively ideal combustion state in the actual situation is determined, so as to realize the dynamic control of the wind pressure during the waste incineration process, achieve better control effects, and more accurately adjust the air volume and wind pressure to ensure the combustion efficiency and the stability of the combustion process.

[0050] The embodiment of the present invention also provides an intelligent wind pressure control system for a waste incinerator, which is used to implement the above-mentioned intelligent wind pressure control method for a waste incinerator. Please refer to Figure 2 , and the system includes: The first wind pressure correction module is used to collect the damper opening data and wind pressure control parameter data of the waste incinerator, calculate the target air output, and generate a first wind pressure feedback signal according to the first target wind pressure data and the target air output; Specifically, based on the first target wind pressure data and the target air output, determine the first air volume difference parameter, obtain the fan performance curve of the waste incinerator, process the first air volume difference parameter through the fan performance curve to generate a first wind pressure compensation parameter, and generate a first wind pressure feedback signal according to the first wind pressure compensation parameter.

[0051] The combustion monitoring and analysis module is used to obtain the combustion monitoring data of the target combustion area in the waste incinerator, including the real-time monitoring data of each sub-area in the target combustion area regarding multiple combustion state parameters, extract the first state time series feature vector of each sub-area regarding each combustion state parameter from the combustion monitoring data, calculate the first time series distance parameter between any adjacent sub-areas regarding each combustion state parameter, analyze the multiple first time series distance parameters according to the combustion state dynamic analysis model, determine the first combustion imbalance parameter between any adjacent sub-areas, and determine the combustion state of the target combustion area according to the multiple first combustion imbalance parameters; Among them, analyzing the multiple first time series distance parameters according to the combustion state dynamic analysis model, determining the first combustion imbalance parameter between any adjacent sub-areas, and determining the combustion state of the target combustion area according to the multiple first combustion imbalance parameters includes: Query the combustion state parameter transfer feature vector between any adjacent sub-areas according to the regional combustion state association list, and perform transfer analysis on the first time series distance parameter between adjacent sub-areas regarding each combustion state parameter according to the combustion state parameter transfer feature vector, including weighting the first time series distance parameter between adjacent sub-areas regarding each combustion state parameter according to the combustion state parameter transfer feature vector, and calculating the first combustion imbalance parameter between adjacent sub-areas; Query the combustion state dynamic range between any adjacent sub-areas according to the regional combustion state association list, and judge whether the first combustion imbalance parameters between any adjacent sub-areas all conform to the corresponding combustion state dynamic range. If so, mark that the target combustion area is in a normal combustion state, otherwise mark that the target combustion area is in an abnormal combustion state.

[0052] The combustion anomaly analysis module is used to determine the abnormal state parameters in the combustion monitoring data if the combustion state of the target combustion area is in an abnormal combustion state, generate the first target reference data corresponding to the combustion monitoring data based on the expected combustion reference data, and analyze the first target reference data according to the combustion scenario analysis model to generate the second target reference data corresponding to the combustion monitoring data; The second wind pressure correction module is used to generate the second target wind pressure data of the target combustion area according to the second target reference data, correct the first wind pressure feedback signal according to the second target wind pressure data to generate a second wind pressure feedback signal, generate a target fan frequency control signal based on the second wind pressure feedback signal, and perform wind pressure control on the waste incinerator.

[0053] Specifically, a second air volume difference parameter is determined according to the second target wind pressure data and the target air output, and the wind pressure dynamic correction data is generated by analyzing the second air volume difference parameter according to the fan performance curve of the waste incinerator. The first wind pressure feedback signal is corrected according to the wind pressure dynamic correction data to generate a second wind pressure feedback signal.

[0054] The above are only specific embodiments of the present invention, enabling those skilled in the art to understand or implement the present invention. Various modifications to these embodiments will be obvious to those skilled in the art. The parts not described in detail in this specification belong to the prior art well-known to those skilled in the art.

Claims

1. An intelligent air pressure control method for a waste incinerator, characterized in that Including: Collecting the air damper opening data and the air pressure control parameter data of the waste incinerator and calculating to obtain the target air output, and generating a first air pressure feedback signal according to the first target air pressure data and the target air output; Obtaining the combustion monitoring data of the target combustion area in the waste incinerator, including the real-time monitoring data of each sub-area in the target combustion area regarding multiple combustion state parameters; Extracting the first state time series feature vectors of each sub-area regarding each combustion state parameter from the combustion monitoring data, calculating the first time series distance parameters between any adjacent sub-areas regarding each combustion state parameter, analyzing the multiple first time series distance parameters according to the combustion state dynamic analysis model, determining the first combustion imbalance parameters between any adjacent sub-areas, and determining the combustion state of the target combustion area according to the multiple first combustion imbalance parameters; If the combustion state of the target combustion area is in an abnormal combustion state, determining the abnormal state parameters in the combustion monitoring data, generating the first target reference data corresponding to the combustion monitoring data based on the expected combustion reference data, and analyzing the first target reference data according to the combustion scenario analysis model to generate the second target reference data corresponding to the combustion monitoring data; Generating the second target air pressure data of the target combustion area according to the second target reference data, correcting the first air pressure feedback signal according to the second target air pressure data to generate a second air pressure feedback signal, generating a target fan frequency control signal based on the second air pressure feedback signal and performing air pressure control on the waste incinerator.

2. The intelligent air pressure control method for a waste incinerator according to claim 1, wherein, Analyzing the multiple first time series distance parameters according to the combustion state dynamic analysis model, determining the first combustion imbalance parameters between any adjacent sub-areas, and determining the combustion state of the target combustion area according to the multiple first combustion imbalance parameters, including: Querying the combustion state parameter transfer feature vectors between any adjacent sub-areas according to the area combustion state association list, and performing transfer analysis on the first time series distance parameters between adjacent sub-areas regarding each combustion state parameter according to the combustion state parameter transfer feature vector, including weighting the first time series distance parameters between adjacent sub-areas regarding each combustion state parameter according to the combustion state parameter transfer feature vector, and calculating to obtain the first combustion imbalance parameters between adjacent sub-areas; Querying the combustion state dynamic range between any adjacent sub-areas according to the area combustion state association list, and judging whether the first combustion imbalance parameters between any adjacent sub-areas all conform to the corresponding combustion state dynamic range. If so, marking the target combustion area as being in a normal combustion state, otherwise marking the target combustion area as being in an abnormal combustion state.

3. The intelligent air pressure control method for a waste incinerator according to claim 2, wherein For the area combustion state association list, it also includes: The area combustion state association list is generated by obtaining and analyzing multiple groups of historical incineration reference data of the waste incinerator. After obtaining multiple groups of historical incineration reference data, extracting the combustion state feature vectors corresponding to multiple moments in each group of historical incineration reference data, calculating the state distance parameters between any adjacent sub-areas regarding multiple moments, and constructing the state distance time series between any adjacent sub-areas; Analyze multiple combustion state parameters between adjacent sub-regions according to the state distance time series, determine the contribution ratio data of each combustion state parameter to the state distance time series, and generate a combustion state parameter transfer feature vector between adjacent sub-regions according to the contribution ratio data; Extract the second state time series feature vector of each sub-region for each combustion state parameter from each set of historical incineration reference data, calculate the second time series distance parameter between any adjacent sub-regions for each combustion state parameter, calculate the second combustion imbalance parameter between adjacent sub-regions in each set of historical incineration reference data according to the combustion state parameter transfer feature vector, and determine the combustion state dynamic range between any adjacent sub-regions according to multiple second combustion imbalance parameters; Construct a regional combustion state association list according to the combustion state parameter transfer feature vector and the combustion state dynamic range between adjacent sub-regions.

4. The intelligent air pressure control method for a waste incinerator according to claim 3, characterized in that, Determine the abnormal state parameters in the combustion monitoring data, and generate the first target reference data corresponding to the combustion monitoring data based on the expected combustion reference data, including: For the process of weighting the first time series distance parameter between adjacent sub-regions for each combustion state parameter according to the combustion state parameter transfer feature vector, count the cumulative distance value of each combustion state parameter, and record the combustion state parameter with the cumulative distance value greater than the preset cumulative threshold as the abnormal state parameter; Calculate the mean value of each sub-region in the combustion monitoring data for each combustion state parameter, construct the first state matrix corresponding to the combustion monitoring data, perform parameter replacement on the abnormal state parameters in the first state matrix according to the expected combustion reference data, generate the second state matrix corresponding to the first state matrix, and obtain the first target reference data corresponding to the combustion monitoring data.

5. The intelligent air pressure control method for a waste incinerator according to claim 4, characterized in that, Analyze the first target reference data according to the combustion scenario analysis model to generate the second target reference data corresponding to the combustion monitoring data, including: Match the first target reference data with multiple combustion scenario clusters, determine the target scenario cluster associated with the first target reference data, determine the third state matrix according to the target scenario cluster, and update the data of multiple abnormal state parameters in the first target reference data according to the third state matrix to generate the second target reference data corresponding to the combustion monitoring data; For multiple combustion scenario clusters, after obtaining multiple sets of historical incineration reference data of the waste incinerator, extract multiple regional state vectors of each set of historical incineration reference data, construct the regional state matrix corresponding to each set of historical incineration reference data, perform data clustering on multiple sets of historical incineration reference data based on the regional state matrix to generate multiple combustion scenario clusters, and generate the target combustion state matrix of each scenario cluster according to the multiple regional state matrices included in each combustion scenario cluster.

6. The intelligent air pressure control method for a waste incinerator according to claim 5, characterized in that, Generate the first wind pressure feedback signal according to the first target wind pressure data and the target air volume, including: Determine the first air volume difference parameter based on the first target wind pressure data and the target air volume, obtain the fan performance curve of the waste incinerator, process the first air volume difference parameter through the fan performance curve to generate the first wind pressure compensation parameter, and generate the first wind pressure feedback signal according to the first wind pressure compensation parameter.

7. The intelligent air pressure control method for a waste incinerator according to claim 6, characterized in that, Correcting the first wind pressure feedback signal according to the second target wind pressure data to generate a second wind pressure feedback signal, including: Determining a second air volume difference parameter according to the second target wind pressure data and the target air output, generating wind pressure dynamic correction data by analyzing the second air volume difference parameter according to the fan performance curve of the waste incinerator, and correcting the first wind pressure feedback signal according to the wind pressure dynamic correction data to generate a second wind pressure feedback signal.

8. An intelligent air pressure control system for a waste incinerator, characterized in that, The system is used to implement the intelligent wind pressure control method for a waste incinerator described in any one of claims 1-7 above, including: A first wind pressure correction module, configured to collect the damper opening data and wind pressure control parameter data of the waste incinerator and calculate the target air output, and generate a first wind pressure feedback signal according to the first target wind pressure data and the target air output; A combustion monitoring and analysis module, configured to obtain the combustion monitoring data of the target combustion area in the waste incinerator, including the real-time monitoring data of each sub-area in the target combustion area regarding multiple combustion state parameters, extract the first state time series feature vector of each sub-area regarding each combustion state parameter from the combustion monitoring data, calculate the first time series distance parameter between any adjacent sub-areas regarding each combustion state parameter, analyze the multiple first time series distance parameters according to the combustion state dynamic analysis model, determine the first combustion imbalance parameter between any adjacent sub-areas, and determine the combustion state of the target combustion area according to the multiple first combustion imbalance parameters; A combustion anomaly analysis module, configured to, if the combustion state of the target combustion area is in an abnormal combustion state, determine the abnormal state parameters in the combustion monitoring data, generate the first target reference data corresponding to the combustion monitoring data based on the expected combustion reference data, and generate the second target reference data corresponding to the combustion monitoring data by analyzing the first target reference data according to the combustion scenario analysis model; A second wind pressure correction module, configured to generate the second target wind pressure data of the target combustion area according to the second target reference data, correct the first wind pressure feedback signal according to the second target wind pressure data to generate a second wind pressure feedback signal, generate a target fan frequency control signal based on the second wind pressure feedback signal, and perform wind pressure control on the waste incinerator.

Images