A shuttle kiln temperature control method and system

Abstract

The application discloses a shuttle kiln temperature control method and system, the method comprises the following steps: according to the pre-established burner heat influence weight matrix, the burner power data sequence is mapped to each temperature data subsequence, and the corresponding burner power data subsequence is obtained; the normal temperature fluctuation range of each space area is determined based on historical temperature data; it is judged whether the temperature difference degree exceeds the normal fluctuation range and the duration exceeds the preset threshold; if it exceeds, the temperature deviation vector is calculated and the power adjustment total amount and the adjustment direction are determined; the candidate burner set to be adjusted is screened according to the power adjustment total amount, the adjustment direction and the heat influence weight matrix, and the power adjustment amount of each candidate burner is calculated; the cooperative control instruction is generated and issued to the burner executing mechanism for temperature adjustment. The application realizes accurate partition control of the kiln temperature field, intelligent screening and power optimization distribution of the burner, effectively reduces energy consumption, and improves control stability and product quality.

Description

Technical Field

[0001] This invention belongs to the field of automatic control technology for industrial kilns, and particularly relates to a method and system for temperature control of shuttle kilns. Background Technology

[0002] As an intermittent firing equipment for ceramics and refractory materials, the precision of temperature control in a shuttle kiln directly determines product quality and production energy consumption. Unlike traditional continuous kilns, shuttle kilns undergo multiple stages during operation, including heating, holding, and cooling. Furthermore, the kiln's internal space is relatively large, resulting in significant temperature differences and thermal coupling effects between different areas.

[0003] Existing temperature control methods for shuttle kilns mainly employ a combination of single-point temperature feedback and PID control. This involves uniformly adjusting the power of all burners or a group of burners based on the deviation between the temperature value at a single measuring point and the target value. This method has the following technical drawbacks: First, a single measuring point cannot reflect the overall temperature distribution within the kiln, easily leading to localized over- or under-burning. Second, uniformly adjusting all burners ignores the differences in thermal impact of different burners on different areas, resulting in low adjustment efficiency and frequent temperature fluctuations. Third, it lacks an intelligent judgment mechanism for abnormal temperature events, making it susceptible to erroneous adjustments due to transient disturbances. Finally, the operating status and combustion efficiency of the burners are not comprehensively considered when determining the adjustment scheme, resulting in energy waste.

[0004] Therefore, there is an urgent need for a shuttle kiln temperature control method that can integrate multi-source data, accurately identify abnormal temperature areas, intelligently select and adjust burners, and optimize power distribution to overcome the shortcomings of existing technologies. Summary of the Invention

[0005] This invention provides a method and system for controlling the temperature of a shuttle kiln, which solves the technical problems in the prior art, such as the inability to accurately identify abnormal temperature areas, low adjustment efficiency, susceptibility to interference and misjudgment, and energy waste.

[0006] In a first aspect, the present invention provides a method for temperature control of a shuttle kiln, comprising: The temperature data of multiple temperature measuring points inside the shuttle kiln at the current moment and the power data of multiple burners are obtained to obtain the temperature data sequence and the burner power data sequence. Based on the preset spatial adjacency relationship and temperature gradient change, the temperature data sequence is divided to obtain at least one temperature data subsequence, wherein each temperature data subsequence corresponds to a continuous spatial region within the kiln. Based on the pre-established burner thermal influence weight matrix, the burner power data sequence is mapped to each temperature data subsequence to obtain the burner power data subsequence corresponding to each temperature data subsequence. The burner thermal influence weight matrix is ​​used to characterize the degree of thermal influence of each burner on each temperature data subsequence. Based on a certain historical temperature data of a certain spatial region within a certain historical time period, a certain normal temperature fluctuation range of a certain spatial region is determined. Determine whether the temperature difference of a certain temperature data subsequence exceeds a certain normal temperature fluctuation range and whether the duration exceeds a preset time threshold. If the temperature fluctuation exceeds a certain normal range and the duration exceeds a preset time threshold, the deviation between the current temperature of each temperature measurement point corresponding to the certain temperature data subsequence and the set target temperature is calculated to obtain a certain temperature deviation vector, and a certain power adjustment amount and a certain adjustment direction of the certain spatial region are determined based on the certain temperature deviation vector. Based on the total power adjustment, the adjustment direction, and the burner thermal influence weight matrix, a set of candidate burners to be adjusted is selected from a subsequence of burner power data corresponding to the subsequence of temperature data, and the power to be adjusted for each candidate burner in the set of candidate burners is calculated according to the preset power allocation rules. Based on the power to be adjusted, a coordinated control command is generated and sent to the corresponding burner actuator for temperature adjustment.

[0007] Secondly, the present invention provides a shuttle kiln temperature control system, comprising: The acquisition module is configured to acquire temperature data from multiple temperature measuring points inside the shuttle kiln at the current moment, as well as power data from multiple burners, to obtain temperature data sequences and burner power data sequences. The partitioning module is configured to partition the temperature data sequence according to a preset spatial adjacency relationship and temperature gradient change to obtain at least one temperature data subsequence, wherein each temperature data subsequence corresponds to a continuous spatial region within the kiln. The mapping module is configured to map the burner power data sequence to each temperature data subsequence according to a pre-established burner thermal influence weight matrix, thereby obtaining burner power data subsequences corresponding to each temperature data subsequence. The burner thermal influence weight matrix is ​​used to characterize the degree of thermal influence of each burner on each temperature data subsequence. The module is configured to determine a normal temperature fluctuation range for a spatial region based on historical temperature data of a spatial region within a historical time period corresponding to a certain temperature data subsequence. The judgment module is configured to determine whether the temperature difference of a certain temperature data subsequence exceeds a certain normal temperature fluctuation range and whether the duration exceeds a preset time threshold. The calculation module is configured to, if the temperature exceeds a certain normal temperature fluctuation range and the duration exceeds a preset time threshold, calculate the deviation between the current temperature of each temperature measurement point corresponding to the certain temperature data subsequence and the set target temperature, obtain a certain temperature deviation vector, and determine a certain power adjustment amount and a certain adjustment direction of the certain spatial region based on the certain temperature deviation vector. The filtering module is configured to filter out a set of candidate burners to be adjusted from a set of burner power data corresponding to a set of temperature data, based on the total power adjustment, the adjustment direction, and the burner thermal influence weight matrix, and to calculate the amount of power to be adjusted for each candidate burner in the set of candidate burners according to a preset power allocation rule. The generation module is configured to generate a collaborative control command based on the power to be adjusted, and send the collaborative control command to the corresponding burner actuator for temperature adjustment.

[0008] Thirdly, an electronic device is provided, comprising: at least one processor, and a memory communicatively connected to the at least one processor, wherein the memory stores instructions executable by the at least one processor, the instructions being executed by the at least one processor to enable the at least one processor to perform the steps of the shuttle kiln temperature control method according to any embodiment of the present invention.

[0009] Fourthly, the present invention also provides a computer-readable storage medium having a computer program stored thereon, wherein when the program instructions are executed by a processor, the processor performs the steps of the shuttle kiln temperature control method according to any embodiment of the present invention.

[0010] The shuttle kiln temperature control method and system of this application acquires temperature data and burner power data from multiple temperature measuring points within the kiln; divides the temperature data sequence according to spatial adjacency and temperature gradient changes to obtain temperature data subsequences corresponding to different spatial regions; maps the burner power data sequence to each temperature data subsequence based on a pre-established burner thermal influence weight matrix to obtain the corresponding burner power data subsequence; determines the normal temperature fluctuation range of each spatial region based on historical temperature data; judges whether the degree of temperature difference exceeds the normal fluctuation range and the duration exceeds a preset threshold; if it exceeds, calculates the temperature deviation vector and determines the total power adjustment and adjustment direction; filters the candidate burners to be adjusted based on the total power adjustment, adjustment direction, and thermal influence weight matrix, and calculates the power to be adjusted for each candidate burner; generates a collaborative control command and sends it to the burner actuator for temperature adjustment, realizing precise zonal control of the kiln temperature field, intelligent selection of burners, and optimized power allocation, effectively reducing energy consumption and improving control stability and product quality. Attached Figure Description

[0011] To more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings used in the following description of the embodiments will be briefly introduced. Obviously, the drawings described below are some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0012] Figure 1 A flowchart of a shuttle kiln temperature control method provided in an embodiment of the present invention; Figure 2 This is a structural block diagram of a shuttle kiln temperature control system according to an embodiment of the present invention; Figure 3 This is a schematic diagram of the structure of an electronic device provided in an embodiment of the present invention. Detailed Implementation

[0013] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0014] Please see Figure 1 The diagram shows a flowchart of a shuttle kiln temperature control method according to this application.

[0015] like Figure 1As shown, the temperature control method for shuttle kilns specifically includes the following steps: Step S101: Obtain temperature data of multiple temperature measuring points inside the shuttle kiln at the current moment and power data of multiple burners to obtain temperature data sequence and burner power data sequence.

[0016] In this step, Step S102: According to the preset spatial adjacency relationship and temperature gradient change, the temperature data sequence is divided to obtain at least one temperature data subsequence, wherein each temperature data subsequence corresponds to a continuous spatial region within the kiln.

[0017] In this step, the three-dimensional spatial coordinates of each temperature measuring point inside the kiln are obtained, and a heat transfer path topology map inside the kiln is constructed based on the three-dimensional spatial coordinates. The heat transfer path topology map uses the temperature measuring points as nodes and the heat conduction channels between adjacent temperature measuring points as edges. Based on the heat transfer path topology diagram, the equivalent thermal resistance between any two adjacent nodes is calculated. The equivalent thermal resistance is determined based on the spatial distance between nodes, the thermal conductivity of the medium, and the airflow disturbance coefficient. Obtain the temperature value of each node at the current moment and the temperature values ​​of the previous M historical moments to form the temperature time-series vector of each node; For each edge connecting two nodes, calculate the dynamic time-warped distance between the temperature time-series vectors of the two nodes; A comprehensive coupling coefficient between nodes is constructed based on a weighted combination of the equivalent thermal resistance and the dynamic time warping distance. Based on the edges whose comprehensive coupling coefficient is greater than a preset first threshold, the nodes are divided and clustered to obtain multiple initial connected subgraphs. Spatial interpolation is performed on the temperature values ​​inside a certain initial connected subgraph to generate a continuous temperature distribution surface, and the gradient magnitude of each point on the temperature distribution surface is calculated, wherein the certain initial connected subgraph is any one of the multiple initial connected subgraphs; The proportion of points whose gradient magnitude exceeds a preset second threshold is counted. If the proportion is lower than a preset third threshold, an initial connected subgraph is determined to be a temperature uniform region. Otherwise, the initial connected subgraph is divided into two parts to obtain at least one target connected subgraph. The initial connected subgraph determined to be a temperature uniform region and the target connected subgraph obtained after secondary segmentation are used as the final spatial region. The current temperature data of all nodes in each final spatial region are combined in order of node number to obtain the at least one temperature data subsequence.

[0018] In one specific embodiment, it is necessary to finely divide the internal space of the kiln to identify areas with relatively uniform temperature distribution, which will then serve as the basic units for subsequent independent control. Unlike traditional methods that rely solely on the current temperature value and spatial distance for division, this method incorporates multiple techniques, including heat transfer path topology diagrams, equivalent thermal resistance, time-series dynamic analysis, and gradient surface segmentation, to achieve high-precision identification of the thermal zones within the kiln. The specific implementation process is as follows: Step 1: Construct the heat transfer path topology. The system first acquires the three-dimensional spatial coordinates of each temperature measuring point within the kiln. These measuring points are pre-deployed at different heights and horizontal positions within the kiln, such as at the top, middle, and bottom, as well as near the product area and the kiln wall. Based on these three-dimensional spatial coordinates, the system constructs a heat transfer path topology map inside the kiln. This topology map uses each temperature measuring point as a node and the heat conduction channels between adjacent measuring points as edges. Here, "adjacent" is defined based on a spatial distance threshold; if the Euclidean distance between two temperature measuring points is less than a preset distance threshold (e.g., set to 0.5 meters based on the kiln size and measuring point density), then a direct heat conduction channel exists between them, and these two nodes are connected by an edge in the topology map.

[0019] Step 2: Calculate the equivalent thermal resistance between adjacent nodes For each edge in the topology graph, i.e., each pair of adjacent temperature measuring points, the system calculates the equivalent thermal resistance between them. Equivalent thermal resistance is a comprehensive index used to quantify the ease with which heat is transferred between these two nodes. Its calculation formula is: Equivalent thermal resistance = (Spatial distance × Reciprocal of medium thermal conductivity) × (1 + Airflow disturbance coefficient). Wherein, spatial distance is the straight-line distance between the two temperature measuring points; the medium thermal conductivity is determined based on the type of medium between the two temperature measuring points. For example, if the two points are connected by refractory brick masonry, the thermal conductivity of refractory brick is used; if the two points are connected by an air cavity, the thermal conductivity of air is used; the airflow disturbance coefficient is a dynamic parameter determined based on simulated airflow data or actual wind speed sensor data within the kiln. It characterizes the enhancing or weakening effect of airflow on heat conduction within the kiln. The stronger the airflow disturbance, the larger this coefficient, and the correspondingly larger the equivalent thermal resistance.

[0020] Step 3: Construct the temperature time-series vector of the node The system not only focuses on the current temperature value but also on the historical temperature trend. For each node (temperature measuring point), it acquires its current temperature value and the temperature values ​​from the previous M historical moments, forming a temperature time-series vector. The value of M is determined based on the thermal inertia characteristics of the kiln and the sampling period. For example, if the sampling period is 10 seconds, M can be set to 30, which includes historical data from the 5 minutes prior to the current moment. This vector reflects the recent temperature change trajectory at that point, containing information on thermal inertia and trend.

[0021] Step 4: Calculate the dynamic time warping distance between time series vectors For each edge connecting two nodes in the topology graph, the system calculates the dynamic time warping distance between the temperature time series vectors of these two nodes. Dynamic time warping is an algorithm used to measure the similarity between two time series. Unlike Euclidean distance, it allows for non-linear alignment of the two series on the time axis, and can better capture the similarity of change patterns caused by heat conduction delays. For example, when the burner power changes, the temperature measuring point closer to the burner heats up first, and the temperature measuring point farther away heats up later. The two temperature series have a time shift, but the change patterns are similar. Dynamic time warping distance can identify this similarity, while Euclidean distance may misjudge dissimilarity.

[0022] Step 5: Construct the comprehensive coupling coefficient between nodes The system weights and combines the equivalent thermal resistance with the dynamic time warping distance to construct a comprehensive coupling coefficient between nodes. This comprehensive coupling coefficient quantifies the thermal coupling strength between two nodes. The specific calculation formula is: Comprehensive Coupling Coefficient = α × (1 / Equivalent Thermal Resistance) + β × (1 / Dynamic Time Warping Distance), where α and β are preset weighting coefficients that can be adjusted according to the kiln characteristics. A larger value indicates a stronger thermal coupling between the two nodes, meaning they have a strong correlation in temperature changes and heat is easily transferred between them.

[0023] Step 6: Graph cutting and clustering to obtain the initial connected subgraph Nodes are clustered by graph cutting based on edges whose overall coupling coefficient is greater than a preset first threshold. Specifically, all edges in the topological graph are traversed, and if the overall coupling coefficient of an edge is less than or equal to the first threshold, the edge is deleted from the graph. After edge deletion, the original connected graph may be divided into multiple disconnected subgraphs, each of which is an initial connected subgraph. The setting of the first threshold should ensure that after deleting weakly coupled edges, the remaining strongly coupled edges can form a region division with practical thermal significance.

[0024] Step 7: Spatial Interpolation and Gradient Analysis For each initially connected subgraph, the system needs to determine whether its internal temperature is sufficiently uniform to decide whether further segmentation is necessary. To do this, spatial interpolation is performed on the temperature values ​​within the initial connected subgraph to generate a continuous temperature distribution surface. Spatial interpolation can employ methods such as Kriging interpolation or radial basis function interpolation, estimating the temperature at any point within the entire subgraph region based on the coordinates and temperature values ​​of discrete temperature measurement points. After generating the continuous surface, the gradient magnitude at each point on the surface is calculated, representing the rate of temperature change. A larger gradient magnitude indicates more drastic temperature changes near that point, potentially representing the boundary between different temperature regions.

[0025] Step 8: Uniformity assessment and secondary segmentation The proportion of the area occupied by points whose gradient magnitude exceeds a preset second threshold is calculated. If this proportion is lower than a preset third threshold, the initial connected subgraph is determined to be a temperature-uniform region and is directly retained as a final spatial region. If the proportion is higher than or equal to the third threshold, it indicates that there is significant temperature non-uniformity within the subgraph, requiring secondary segmentation.

[0026] The specific implementation of secondary segmentation is as follows: Local maxima of gradient magnitudes are searched on the temperature distribution surface. These points correspond to the locations of the most drastic temperature changes, typically the boundaries between different temperature regions. Using each local maximum as the center, a region growing algorithm is executed outwards, merging points with smaller gradient magnitudes into the same region, until a growing boundary of another local maximum is encountered. This process is similar to the watershed segmentation algorithm in image processing, dividing the temperature distribution surface into multiple "catchment basins," each basin representing a relatively uniform temperature sub-region. Each sub-region obtained after segmentation is a target connected subgraph.

[0027] To ensure sufficient segmentation, the system recalculates the internal temperature variance of each target connected subgraph obtained after the secondary segmentation. If the temperature variance of a subgraph is still greater than the preset fourth threshold, the above secondary segmentation steps are repeated for that subgraph until the temperature variance of all subgraphs is less than or equal to the fourth threshold, or the number of iterations reaches the preset upper limit.

[0028] Step 9: Generate temperature data subsequences The initial connected subgraph (temperature uniform region) directly retained after uniformity judgment, along with all target connected subgraphs obtained after iterative segmentation, are collectively used as the final spatial region. For each final spatial region, the numbers of all temperature measurement points within that region are obtained, and the temperature data of these measurement points at the current moment are arranged sequentially according to the ascending order of their numbers, forming a temperature data subsequence. All temperature data subsequences corresponding to the final spatial regions are combined according to the spatial location order of the regions (e.g., sorted according to the coordinates of the region's center) to obtain at least one temperature data subsequence, which serves as the input for subsequent steps.

[0029] In summary, the method in this embodiment captures temporal similarity through dynamic time warping distance, thus solving the problem of misclassification caused by thermal conduction delay. When there is a thermal conduction delay between two temperature measurement points, their temperature change patterns are similar but there is a time offset. Dynamic time warping distance can accurately identify this similarity, allowing nodes with strong dynamic coupling relationships to be classified into the same region, laying a good foundation for subsequent collaborative control. Furthermore, a two-level partitioning strategy is adopted: first, coarse segmentation is performed based on the coupling coefficient, and then fine segmentation is performed based on the gradient surface, taking into account both global connectivity and local uniformity. Graph cutting clustering ensures the connectivity and strong cohesion of the regions, while gradient surface analysis and secondary segmentation ensure that the temperature distribution within each final region is sufficiently uniform, avoiding the decrease in control accuracy caused by forcibly classifying points with large temperature differences into the same region.

[0030] Step S103: Based on the pre-established burner thermal influence weight matrix, the burner power data sequence is mapped to each temperature data subsequence to obtain the burner power data subsequence corresponding to each temperature data subsequence. The burner thermal influence weight matrix is ​​used to characterize the degree of thermal influence of each burner on each temperature data subsequence.

[0031] In this step, with the kiln in stable operating condition, the temperature response amplitude at each temperature measurement point is obtained when the power of each burner changes. Divide the temperature response amplitude of a certain burner to a certain temperature measuring point by the power adjustment amount of the certain burner to obtain the unit power influence coefficient of a certain burner to a certain temperature measuring point. The unit power influence coefficients of all burners on all temperature measurement points are combined into an initial weight matrix, and the initial weight matrix is ​​normalized to obtain the final burner thermal influence weight matrix. The normalization process includes making the sum of the elements in each row of the initial weight matrix equal to 1, that is, the sum of the influence coefficients of all burners on each temperature measurement point is equal to 1. Extract the weighted row vectors corresponding to all temperature measurement points within a certain spatial region from the burner thermal influence weight matrix, wherein the certain spatial region is the spatial region corresponding to a certain temperature data subsequence; The average value of all extracted weight row vectors is calculated along the column direction to obtain the comprehensive weight vector. Each element of the comprehensive weight vector represents the degree of comprehensive thermal influence of the corresponding burner on a certain spatial region. Obtain the power values ​​of all burners at the current moment and construct a power vector; The power vector is multiplied by the comprehensive weight vector, that is, the elements at corresponding positions are multiplied and then summed to obtain the equivalent power value of a certain spatial region. The equivalent power values ​​of all spatial regions are combined according to the spatial location order of the spatial regions to obtain the burner power data subsequence corresponding to each temperature data subsequence.

[0032] In one specific embodiment, the system needs to establish a burner thermal impact weight matrix and use this matrix to map the real-time power data of each burner to each spatial region, obtaining the equivalent power value corresponding to each region. This mapping process aggregates the originally scattered burner power information into a comprehensive thermal impact index for each region, providing a quantitative basis for subsequent judgment on whether temperature anomalies have occurred in that region. The specific implementation process is as follows: Step 1: Obtain the temperature response amplitude when the burner power changes. The test was conducted when the kiln was in stable operating condition. The criteria for stable operating condition were: the rate of temperature change at all measuring points was less than a preset stability threshold (e.g., no more than 0.5°C per minute), and there were no ongoing adjustments within the kiln. During the test, the power of all other burners was kept constant, and a preset power step change was applied to only one burner. The power change amplitude should be reasonably selected based on the burner's rated power, such as increasing the power by 20% or switching from the current power to a fixed value, ensuring a measurable temperature response while avoiding excessive impact on the kiln.

[0033] Starting from the moment the power excitation is applied, temperature values ​​at all measurement points are continuously acquired at a fixed sampling frequency (e.g., once every 10 seconds). The acquisition duration should be sufficient to cover the entire temperature response process until the temperature at each measurement point reaches a stable state again. For each measurement point, the following values ​​are extracted: the stable temperature value before power excitation (the average of the last three sampling points before excitation), and the temperature value when the temperature stabilizes again after power excitation (the average of the last three sampling points after stabilization). The absolute value of the difference between these two stable temperatures is the temperature response amplitude at that measurement point. This amplitude reflects the magnitude of the thermal impact of the burner power change on that measurement point.

[0034] To eliminate the influence of the kiln's own slow thermal drift, one or more reference temperature measurement points can be set up during the test. These reference points are located far from the burner under test, and theoretically, are minimally affected by the burner. The net temperature response amplitude after deducting background drift is obtained by subtracting the temperature change at the reference point from the temperature response amplitude at each measurement point.

[0035] Repeat the above test process for each burner to obtain the temperature response amplitude of each burner at all temperature measurement points when the power changes.

[0036] Step 2: Calculate the influence coefficient of unit power For a given burner j and a given temperature measuring point i, the temperature response amplitude ΔT of the burner at that temperature measuring point is...ij Divide by the burner's power adjustment ΔP j The influence coefficient w of the burner on the unit power of the temperature measuring point was obtained. ij That is: w ij =ΔT ij / ΔP j This coefficient represents the change in temperature at the measuring point when the burner changes by a unit of power; the unit is usually ℃ / kW.

[0037] Step 3: Construct the initial weight matrix and normalize it. The influence coefficients of all burners on the unit power of all temperature measurement points are combined into a matrix, with rows corresponding to temperature measurement points and columns corresponding to burners, to obtain the initial weight matrix W. init The matrix has a size of m×n, where m is the total number of temperature measuring points and n is the total number of burners.

[0038] The initial weight matrix is ​​row-normalized. For each row (corresponding to a temperature measurement point), the sum of all elements in that row is calculated, and then each element in that row is divided by the row sum. The normalized matrix W satisfies the following condition: the sum of the elements in each row equals 1, meaning the sum of the influence coefficients of all burners on each temperature measurement point is 1. The purpose of normalization is to eliminate dimensional differences caused by variations in location and sensitivity between different temperature measurement points, making the weight matrix comparable. The normalized matrix is ​​the final burner thermal influence weight matrix.

[0039] Step 4: Extract the weight row vectors corresponding to the regions. Based on the spatial regions obtained in step S102, for each spatial region, the weighted row vectors corresponding to all temperature measuring points within that region are extracted from the burner thermal influence weight matrix. For example, if a region contains k temperature measuring points, then k row vectors are extracted, each with a length of n (the total number of burners).

[0040] Step 5: Calculate the comprehensive weight vector of the region For the extracted k weighted row vectors, the average value is calculated along the column direction to obtain a comprehensive weight vector of length n. Specifically, for each column (corresponding to a burner), the values ​​of that column in the k row vectors are summed, and then divided by k to obtain the comprehensive thermal influence of that burner on the region. Each element α of the comprehensive weight vector... rj This indicates the overall thermal influence of burner j on region r. The larger the value, the more significant the influence of the burner on the region.

[0041] Step 6: Obtain the current burner power vector At the current moment, acquire the real-time power values ​​of all burners, arrange them in order of burner number, and form a power vector P=[p1,p2,...,p...]. n ].

[0042] Step 7: Calculate the equivalent power value of the region The combined weight vector α of the power vector P and the region r r Performing a dot product, that is, multiplying the elements at corresponding positions and then summing them, yields the equivalent power value E of region r. r Calculate the common E r = This equivalent power value is a scalar that comprehensively reflects the total thermal impact of all burners on the region at the current moment. Since the weight matrix has been normalized, the dimensions of the equivalent power value are consistent with the dimensions of the burner power, facilitating subsequent comparisons with temperature deviations, etc.

[0043] Step 8: Combine the equivalent power values ​​of all regions For each spatial region obtained in step S102, repeat steps four through seven to obtain the equivalent power value for each region. Arrange these equivalent power values ​​according to the spatial region's positional order (e.g., according to the region's center coordinates or region numbering order) to obtain a sequence. This sequence is the burner power data subsequence corresponding to each temperature data subsequence. Each element in this sequence corresponds to the current equivalent heating power of a region, and corresponds one-to-one with the temperature data of each region in the temperature data subsequence.

[0044] In summary, the method in this embodiment directly obtains the temperature response amplitude of each temperature measuring point to the burner power change by conducting single burner excitation tests under actual operating conditions, avoiding errors caused by theoretical calculations or simulations. The testing process considers the stable operating conditions of the kiln and eliminates background drift, ensuring the accuracy of the weighting coefficients. Normalization processing makes the weight matrix have uniform dimensions and comparability, laying the foundation for subsequent applications. Furthermore, by aggregating the weight information of multiple temperature measuring points in the region through a comprehensive weight vector, and then multiplying it by the burner power, an equivalent power value reflecting the overall heating state of the region is obtained. This mapping reduces the high-dimensional burner power information to a single indicator for each region, greatly simplifying subsequent analysis and control.

[0045] Step S104: Based on a certain historical temperature data of a certain spatial region corresponding to a certain temperature data subsequence within a certain historical time period, determine a certain normal temperature fluctuation range of a certain spatial region.

[0046] In this step, historical temperature data of a certain spatial region within the same time window as the current firing stage in multiple historical firing cycles are obtained to form a historical temperature dataset. Calculate the mean and standard deviation of the historical temperature dataset; The range of the average value plus or minus three standard deviations is taken as the normal temperature fluctuation range of the certain spatial region. Obtain the process parameters for the current firing cycle, including loading amount and product type; Based on the preset correspondence table between process parameters and fluctuation range, find the correction coefficient corresponding to the current process parameters; By multiplying the upper and lower limits of the normal temperature fluctuation range by the correction coefficient, a normal temperature fluctuation range for a certain spatial region is obtained.

[0047] In one specific embodiment, the system needs to establish the temperature fluctuation range under normal operating conditions for each spatial region as the basis for subsequent judgment of whether the temperature is abnormal. Unlike traditional methods that use fixed thresholds, this method is based on historical data statistics and combined with dynamic correction of process parameters, giving the fluctuation range an adaptive capability to adapt to temperature change characteristics under different operating conditions. The specific implementation process is as follows: Step 1: Construct a historical temperature dataset For a specific spatial area, the system retrieves temperature data for that area across multiple historical firing cycles from a historical database. The selection of historical firing cycles should meet the following conditions: complete cycles, stable operation, and qualified product quality. For each historical cycle, the system determines the same time window based on the current firing stage. For example, if the current period is within the heating phase, the system selects the corresponding time window (e.g., 30 to 40 minutes after the start of heating) from the same heating phase within each historical cycle. The selection of the time window should ensure data comparability; typically, it is centered on the current moment, extending forward and backward by a certain duration (e.g., 5 minutes before and after the current moment) to form the time window.

[0048] The temperature values ​​for this time window within all selected historical periods are aggregated to form a historical temperature dataset. This dataset contains the historical temperature performance of this region under the same firing stage and operating conditions.

[0049] Step 2: Calculate statistical characteristics Statistical analysis was performed on the historical temperature dataset to calculate its mean μ and standard deviation σ. The mean reflects the typical temperature level of the region during this firing stage, and the standard deviation reflects the normal fluctuation range of the temperature.

[0050] Step 3: Determine the basic normal fluctuation range According to the 3σ principle in statistics, the range of mean plus or minus three standard deviations is taken as the baseline normal temperature fluctuation range for this region. That is, the lower limit L. base =μ-3σ, upper limit U base =μ+3σ. This range theoretically covers 99.7% of normal fluctuations and can effectively distinguish between normal and abnormal fluctuations.

[0051] Step 4: Obtain the current process parameters Obtain the process parameters for the current firing cycle from the production management system, including at least the loading volume and product type. Loading volume refers to the total amount of material loaded into the kiln, and product type refers to the type of product being fired (e.g., ceramics, refractory bricks). Different process parameters affect the kiln's thermal characteristics, causing variations in the normal temperature fluctuation range. For example, a larger loading volume results in greater thermal inertia, potentially leading to smoother temperature fluctuations; different product types have different firing curves and heat capacity characteristics, resulting in different temperature fluctuation ranges.

[0052] Step 5: Find the correction coefficients corresponding to the process parameters The system pre-stores a table showing the correspondence between process parameters and fluctuation ranges. This table, established using experimental data or historical experience, records correction coefficients for different combinations of loading capacity and product types. For example, for a loading capacity of 80% of the standard loading capacity, the correction coefficient might be 0.9, indicating that the fluctuation range should be reduced by 10%; for a loading capacity of 120%, the correction coefficient might be 1.1, indicating that the fluctuation range should be expanded by 10%. Different product types may also have their own baseline correction coefficients.

[0053] Based on the loading volume and product type of the current firing cycle, find the matching correction coefficient k in the corresponding table. If there are multiple combinations of process parameters, a weighted or multiplicative method can be used to calculate the comprehensive correction coefficient. For example, k=k 装载量 ×k 产品种类 .

[0054] Step 6: Correct to obtain the final normal fluctuation range Multiply the upper and lower limits of the basic normal fluctuation range by the correction factor k to obtain the final normal temperature fluctuation range adapted to the current operating conditions: L final =L base ×k, U final =U base ×k, This range represents the normal temperature fluctuation range of this spatial region at the current moment, and is used for subsequent anomaly detection.

[0055] For each spatial region, the above steps are repeated to obtain the normal temperature fluctuation range for each region. Since the thermal characteristics and historical data of different regions are different, their fluctuation ranges may vary, thus achieving region-adaptive threshold setting.

[0056] Step S105: Determine whether the temperature difference of a certain temperature data subsequence exceeds the normal temperature fluctuation range and whether the duration exceeds a preset time threshold.

[0057] In this step, before determining whether the temperature difference degree of a certain temperature data subsequence exceeds the normal temperature fluctuation range and whether the duration exceeds the preset time threshold, calculate the average value of the current temperatures of all temperature measurement points in the certain temperature data subsequence as the regional representative temperature; calculate the absolute value of the difference between the regional representative temperature and the central value of the normal temperature fluctuation range as the temperature difference degree.

[0058] In a specific embodiment, the system needs to determine whether the temperature in a certain spatial region is abnormal and ensure that the abnormality is not caused by instantaneous interference, so as to improve the accuracy of judgment. The specific implementation process includes three links: calculation of temperature difference degree, boundary overrun judgment, and duration monitoring.

[0059] Step 1: Calculate the regional representative temperature For the spatial region corresponding to a certain temperature data subsequence, obtain the current temperature values of all temperature measurement points included in the subsequence. Since this region may contain multiple temperature measurement points, it is necessary to aggregate the temperature values of these points to obtain a single value that can represent the overall temperature level of the region. This method uses the arithmetic mean method, that is, calculate the average value of the current temperature values of all temperature measurement points in the region as the regional representative temperature T rep .

[0060] Using the average value as the regional representative temperature can smooth the measurement noise or local disturbances of individual temperature measurement points and reflect the overall temperature level of the region.

[0061] Step 2: Calculate the temperature difference degree Obtain the normal temperature fluctuation range determined in step S104 for this region. This range has a lower limit L and an upper limit U, and its central value C = (L + U) / 2. The central value represents the ideal or typical temperature level of this region under the current working conditions.

[0062] Calculate the regional representative temperature T rep The absolute value of the difference from the central value C is used as the temperature difference degree D: D = |T rep - C|, This difference degree quantifies the degree to which the current regional temperature deviates from the normal center and is used for subsequent overrun judgment.

[0063] Step 3: Judge whether it exceeds the fluctuation range The system judges whether the regional representative temperature T rep exceeds the boundary of the normal temperature fluctuation range: If T rep > U, it is determined that the temperature is too high and abnormal; If T rep < L, it is determined that the temperature is too low and abnormal; If L≤T rep If the temperature is ≤U, then the temperature is considered normal and no adjustment is needed.

[0064] If the temperature is determined to be normal, the process ends and waits for the next sampling time to reassess. If the temperature is determined to be abnormal (too high or too low), the process enters the duration monitoring phase.

[0065] Step 4: Record the start time of the anomaly and start timing. When the temperature of a region is detected to exceed the normal fluctuation range for the first time, the system records the current time t. start This serves as the starting point for the abnormal event, and a timer is initiated to accumulate the duration exceeding the specified threshold.

[0066] Step 5: Continuous monitoring and timed updates At each subsequent sampling time, the system repeats steps one through three, continuously monitoring the representative temperature of the area. The following situations may occur during the monitoring process: If the temperature remains outside the range (whether too high or too low), the timer continues to accumulate, and the accumulated duration increases continuously. If the temperature returns to the normal fluctuation range, the abnormal event is considered to have ended, the system will reset the timer to zero, and wait for the next over-limit event to occur before restarting the timer. If the temperature frequently switches between out-of-range and normal states, but each out-of-range event is short-lived, the timer will reset to zero each time it returns to normal and will not accumulate.

[0067] Step 6: Determine whether the preset time threshold has been reached. The system has a preset time threshold T. threshold (e.g., 30 seconds, 1 minute, etc.) are used to distinguish between brief disturbances and persistent anomalies. During the timer's accumulation process, the current accumulated duration t is updated in real time. duration Compare with a preset time threshold: If t duration <T threshold If so, continue monitoring without triggering adjustment; If t duration ≥T threshold If the temperature anomaly in the area has been determined to have lasted for a sufficient period of time, it is confirmed to be a real anomaly rather than a transient disturbance, triggering the subsequent temperature regulation process (i.e., executing step S106 and subsequent steps).

[0068] Through this dual judgment mechanism of "exceeding the standard amplitude + meeting the standard duration", the system can effectively filter out transient interferences such as measurement noise, door opening and closing disturbances, and combustion fluctuations. Adjustments are only made after the abnormal stability is confirmed to be continuous, thus avoiding energy waste and temperature fluctuations caused by frequent adjustments.

[0069] Step S106: If the temperature fluctuation exceeds the normal range and the duration exceeds a preset time threshold, calculate the deviation between the current temperature of each temperature measurement point corresponding to the temperature data subsequence and the set target temperature to obtain a temperature deviation vector, and determine the total power adjustment amount and adjustment direction of the spatial region based on the temperature deviation vector.

[0070] In this step, the numbering information of each temperature measuring point in a certain spatial region corresponding to a certain temperature data subsequence is obtained, and the spatial location of each temperature measuring point in the kiln is determined according to the numbering information. Based on the preset firing curve and the current firing stage, determine the target temperature value corresponding to each temperature measuring point, and construct a target temperature vector based on the spatial position of each temperature measuring point in the kiln. Extract the current temperature value of each temperature measurement point at the current moment from a certain temperature data subsequence, and arrange them in the same order as the target temperature vector to form the current temperature vector; Subtract the corresponding element in the target temperature vector from each element in the current temperature vector to obtain a temperature deviation vector composed of multiple differences. Each element of the temperature deviation vector represents the deviation between the current temperature value of the corresponding temperature measurement point and the set target temperature value. A positive deviation value indicates that the current temperature value is higher than the target temperature value, and a negative deviation value indicates that the current temperature value is lower than the target temperature value. The arithmetic mean of all elements in the temperature deviation vector is used to obtain the average temperature deviation of a certain spatial region. A certain adjustment direction is determined based on the sign of the average temperature deviation. If the average temperature deviation is positive, then the adjustment direction for a certain spatial region is the cooling direction; if the average temperature deviation is negative, then the adjustment direction for a certain spatial region is the heating direction. A certain proportional coefficient is obtained for a certain spatial region. The certain proportional coefficient is pre-calibrated based on the volume of the spatial region, the heat capacity characteristics of the kiln inner wall, and historical adjustment experience data. It is used to characterize the amount of power adjustment required per unit temperature deviation. Multiplying the absolute value of the average temperature deviation by the proportional coefficient yields the total power regulation for a given spatial region.

[0071] In one specific embodiment, after step S105 determines that the temperature anomaly in a certain spatial region has persisted for more than a preset time threshold, the system enters the adjustment calculation stage. This step requires quantifying the temperature deviation at multiple points within the region into a unified total power adjustment and determining the adjustment direction, providing input for subsequent burner selection and power allocation. The specific implementation process is as follows: Step 1: Obtain the temperature measurement point number and spatial location information For a specific spatial area experiencing a temperature anomaly, the system first acquires a subsequence of temperature data corresponding to that area. This subsequence contains the current temperature data of all temperature measuring points within the area, but these data are arranged in order of their temperature measuring point numbers. To subsequently correlate this data with the target temperature, the system needs to obtain the number information for each temperature measuring point and, based on this number, query the system database for the spatial coordinates of each temperature measuring point within the kiln. The spatial location information includes the three-dimensional coordinates (x, y, z) of the temperature measuring point and a description of its functional location within the kiln (e.g., top, middle, bottom, near the product area, near the kiln wall, etc.). This information will be used to determine the target temperature value for each temperature measuring point.

[0072] Step 2: Determine the target temperature value for each temperature measurement point. Based on the preset firing curve and the current firing stage, the system determines the corresponding target temperature value for each temperature measuring point. The firing curve is a pre-set temperature-time curve that describes the target temperature requirements for each stage of the entire firing process. However, due to the different thermal characteristics of different spatial locations within the kiln, the target temperature at different locations at the same time may vary. For example, during the heating stage, the target temperature in the area closer to the burner may be slightly higher than that in the area farther from the burner; during the holding stage, all points within the kiln should maintain as consistent a temperature as possible; the area closer to the product needs to strictly adhere to the product firing temperature requirements, while the area closer to the kiln wall can tolerate a certain degree of deviation.

[0073] Therefore, the system determines the target temperature for each temperature measurement point based on the following factors: The current stage reference temperature T specified in the firing profile base ; The spatial position correction coefficient for the temperature measuring point is determined by looking up the correction value ΔT from a pre-calibrated position correction table, based on factors such as the distance of the point from the burner, the distance from the product, and the distance from the kiln wall. pos ; Due to product firing process requirements, certain key temperature measurement points may have independent temperature requirements. product .

[0074] Finally, the set target temperature value T at temperature measurement point i target-i =T base +ΔT pos-i For key points, simply take T. product-i Arrange the target temperature values ​​of all temperature measurement points in order of their numerical designation to form the target temperature vector T. target =[T target-1 ,T target-2 ,...,T target-N], where N is the number of temperature measurement points in the area.

[0075] Step 3: Construct the current temperature vector From the temperature data subsequence of this region, extract the temperature value of each temperature measurement point at the current time. Since the temperature data subsequence itself is arranged in order of the temperature measurement point numbers, it can be directly used as the current temperature vector T. current =[T current-1 ,T current-2 ,...,T current-N The order of this vector is completely consistent with the target temperature vector, ensuring the correctness of subsequent element-by-element comparisons.

[0076] Step 4: Calculate the temperature deviation vector Subtract the corresponding element in the target temperature vector from each element in the current temperature vector to obtain the temperature deviation vector ΔT = [ΔT1, ΔT2, ..., ΔT]. N ], where ΔT i =T current-i -T target-i Each element of the temperature deviation vector represents the deviation between the current temperature value and the set target temperature value at the corresponding temperature measurement point. If ΔT i A value greater than 0 indicates that the current temperature at this point is higher than the target temperature and cooling is required. If ΔT i <0 indicates that the current temperature at this point is lower than the target temperature, and the temperature needs to be increased; If ΔT i =0 indicates that the temperature at that point has just reached the target.

[0077] This vector retains detailed deviation information for each point within the region, providing a basis for subsequent fine-tuning.

[0078] Step 5: Calculate the average temperature deviation of the region The average temperature deviation ΔT for the region is obtained by taking the arithmetic mean of all elements in the temperature deviation vector. avg ; The average temperature deviation represents the degree and direction of the overall deviation of the region from the target temperature: If ΔT avg A value >0 indicates that the overall temperature in the region is relatively high; If ΔT avg A value less than 0 indicates that the overall temperature in the region is low. If ΔT avg =0 indicates that the overall temperature of the region is close to the target, but there may be a situation where positive and negative deviations cancel each other out. In this case, it is necessary to further judge based on the distribution of the deviation vector.

[0079] Step 6: Determine the adjustment direction Based on the average temperature deviation ΔT avg The sign of the value determines the adjustment direction for that region: If ΔT avg If the value is >0, the adjustment direction is cooling, and the heat input needs to be reduced by lowering the burner power. If ΔT avg If the value is less than 0, the adjustment direction is the heating direction, which requires increasing the burner power to increase the heat input.

[0080] The adjustment direction will be used for pattern matching during subsequent burner selection, as well as for assigning symbols during power distribution.

[0081] Step 7: Obtain the regional ratio coefficient Each spatial region is pre-defined with a proportionality coefficient Kr, which is used to convert temperature deviation into power regulation, representing the power regulation required per unit temperature deviation, with units of kW / ℃. The determination of the proportionality coefficient takes into account the following factors: The size of this region: the larger the volume, the greater the heat capacity, and the more power is required to regulate a unit temperature. Kiln inner wall heat capacity characteristics: The inner wall of different materials has different heat capacity and heat conduction characteristics; Historical regulation experience data: obtained by fitting the relationship between temperature changes and power regulation in historical regulation records using a system identification method; Thermal response characteristics of the region: The sensitivity of the region to power can be inferred by using the burner thermal influence weight matrix established in step S103.

[0082] The proportional coefficient can be determined by combining offline calibration with online correction, and dynamically adjusted according to the actual adjustment effect during kiln operation.

[0083] Step 8: Calculate the total power regulation Multiplying the absolute value of the average temperature deviation by the regional proportionality coefficient yields the total power regulation ΔPr for that region. ΔPr=ΔT avg ×K_r The total power adjustment ΔP_r is a positive number, representing the absolute value of the power that needs to be increased or decreased. Combined with the adjustment direction determined in step six, a complete adjustment command can be obtained: if it is a heating direction, the power needs to be increased by ΔP_r; if it is a cooling direction, the power needs to be decreased by ΔP_r.

[0084] At this point, the system has identified clear adjustment targets and quantitative indicators for the areas where temperature anomalies occur: the direction of adjustment (heating or cooling) and the total amount of power adjustment (the amount of power that needs to be adjusted), laying a solid foundation for subsequent burner selection and power allocation.

[0085] Step S107: Based on the total power adjustment, the adjustment direction, and the burner thermal influence weight matrix, select a set of candidate burners to be adjusted from a subsequence of burner power data corresponding to the subsequence of the temperature data, and calculate the amount of power to be adjusted for each candidate burner in the set of candidate burners according to the preset power allocation rules.

[0086] In this step, the weight row vectors corresponding to all temperature measurement points within a certain temperature data subsequence are obtained in the burner thermal influence weight matrix to form a thermal influence submatrix; The thermal influence submatrix is ​​decomposed into a nonnegative matrix decomposition into a product of a basis matrix and a coefficient matrix. Each column of the basis matrix represents a potential thermal influence mode, and each row of the coefficient matrix represents the contribution of a burner to each thermal influence mode. Based on a certain adjustment direction, target feature patterns that match the heating or cooling requirements are selected from the base matrix; Extract the column vectors corresponding to the target feature patterns from the coefficient matrix to obtain the contribution coefficient of each burner to the target feature patterns; Obtain the current operating status parameters of each burner, including current power value, cumulative running time, and fuel supply pressure; A comprehensive evaluation of burner adjustment priority is constructed based on the contribution coefficient of each burner, the ratio of current power value to rated power, the ratio of cumulative running time to design life, and the ratio of fuel supply pressure to rated pressure. The burners are sorted from highest to lowest according to the comprehensive score value; Select burners sequentially from the sorted list, calculate the sum of the contribution coefficients of the selected burners, and continue until the sum of the contribution coefficients of the selected burners reaches the preset coverage threshold as a proportion of the total contribution coefficients of all burners. Then, select the burners as a candidate burner set.

[0087] In one specific embodiment, after step S106 determines the total power adjustment amount and direction for a certain spatial region, this step needs to address two core issues: which burners to adjust, and how much power each selected burner should be adjusted by. Unlike traditional methods that uniformly adjust all burners, this solution extracts the thermal influence mode through non-negative matrix decomposition and combines this with a comprehensive score based on the burner's operating status, achieving precise burner selection and power allocation. The specific implementation process is as follows: Step 1: Construct the thermal influence submatrix For the spatial region where temperature anomalies occur, weighted row vectors corresponding to all temperature measuring points within that region are extracted from the burner thermal influence weight matrix established in step S103. Assuming the region contains m temperature measuring points and the total number of burners is n, the extracted m row vectors constitute a thermal influence submatrix W of size m × n. sub Each row of this matrix corresponds to a temperature measuring point, and each column corresponds to a burner. The element w... ij This represents the weight of the thermal influence of burner j on temperature measuring point i.

[0088] The thermal influence matrix centrally reflects the thermal influence relationship between all temperature measuring points and each burner in the region, and is a mathematical expression of the thermal characteristics of the region.

[0089] Step 2: Extracting the thermal effect mode through nonnegative matrix decomposition thermal influence submatrix W sub Perform nonnegative matrix decomposition, decomposing it into the product of two nonnegative matrices: W sub =B×C. Where: B is an m×k basis matrix, where k is the rank of the decomposition (a preset value, usually an integer less than m and n, such as 3-5). Each column of the basis matrix represents a potential thermal effect pattern, i.e., a typical heat distribution pattern. For example, one column might correspond to a "top heating pattern," meaning that this pattern has a greater impact on the temperature measurement points at the top of the area and a smaller impact on the bottom; another column might correspond to a "center heating pattern," meaning that it has a greater impact on the temperature measurement points in the center of the area.

[0090] C is a k×n coefficient matrix, where each row corresponds to a heat-affected zone and each column corresponds to a burner. The element c... pj This represents the contribution coefficient of burner j to the p-th heat-affected mode. The larger the contribution coefficient, the more significant the burner's performance in that mode.

[0091] The nonnegative matrix factorization can be solved using a multiplicative iterative algorithm, obtaining B and C by minimizing the reconstruction error. The decomposition result ensures that all elements are nonnegative, which is consistent with the physical meaning that the thermal influence weights are all nonnegative, and the decomposition patterns and contributions have clear physical interpretability.

[0092] Step 3: Filtering and adjusting target feature patterns to match direction Based on the adjustment direction (heating or cooling) determined in step S106, target feature patterns that match the current adjustment requirements are selected from the k column vectors (thermal influence patterns) of the basis matrix B. The selection is based on the correlation between the spatial distribution characteristics of the patterns and the temperature deviation vector.

[0093] Specifically, the system obtains the temperature deviation vector ΔT calculated in step S106 (this vector has a length of m, representing the temperature deviation at each temperature measurement point). For each column (pattern) of the basis matrix B, the cosine similarity or correlation coefficient between this column vector and the temperature deviation vector ΔT is calculated. The higher the cosine similarity, the more consistent the spatial distribution of the pattern is with the current temperature deviation distribution.

[0094] For example, if the temperature deviation vector shows that the temperature at the top of the area is significantly higher than the bottom, while the temperature at the bottom is normal, then the mode that is most similar to the top heating mode will be selected. If the adjustment direction is heating, then the mode that is similar to the negative deviation (the area that needs to be heated) distribution will be selected; if it is cooling, then the mode that is similar to the positive deviation distribution will be selected.

[0095] Select one or more patterns with the highest similarity as the target feature patterns. Typically, you can choose a single pattern with the highest similarity, or combine multiple patterns by weighted similarity.

[0096] Step 4: Extract the burner's contribution coefficient to the target mode From the coefficient matrix C, extract the row vectors corresponding to the target feature patterns (if multiple patterns are selected, use a weighted combination). The length of this row vector is n, and each element corresponds to the contribution coefficient α of a burner to that target pattern. j Contribution coefficient α j The larger the value, the stronger the burner j is in achieving the heat-affected zone, meaning that adjusting the burner can more effectively correct the current temperature deviation distribution.

[0097] Step 5: Obtain the current operating status parameters of the burner. To ensure that the selected burners have good adjustability and safety, the system acquires the current operating status parameters of each burner, including: Current power value p j : Read directly from the burner control system; Rated power p rated-j The burner's maximum design power; Cumulative runtime t j Total operating time since the burner was put into use; Design life t life-j The burner's designed service life; Fuel supply pressure f j Current fuel pipeline supply pressure; Rated pressure f rated-j Rated pressure required for the burner to operate normally.

[0098] These parameters reflect the burner's current load status, remaining lifespan, and operational health.

[0099] Step 6: Construct a comprehensive score for burner adjustment priorities. Build a comprehensive score S for each burner j This is used to quantify its priority as a candidate for regulation. The comprehensive score takes into account the burner's contribution to the target mode, current load margin, remaining life, and fuel supply status. An example scoring function is: S j =α j ×w α +(1-p j / p rated-j )×w p +(1-t j / t life-j )×w t +(f j / f rated-j )×w f Where: α j The contribution coefficient of the burner to the target mode; (1-p j / p rated-j The value represents the power margin; a larger value indicates that the burner has more room for power adjustment (for heating) or more room for power adjustment (for cooling). (1-t j / t life-j The value represents the percentage of the burner's remaining lifespan; a higher value indicates a newer burner that should be used preferentially. (f j / f rated-j This indicates the fuel pressure adequacy level; a higher value indicates a more sufficient gas supply. w α w p w t w f The weighting coefficients for each factor can be adjusted according to the control strategy (e.g., increasing w when emphasizing the regulatory effect). α It is emphasized that when protecting equipment, the w can be increased t ).

[0100] This scoring function ensures that burners with high contribution, large load margin, long remaining lifespan, and sufficient gas supply receive higher priority.

[0101] Step 7: Sort by overall score Calculate the overall score S for all burners. j The burners are sorted according to their scores from highest to lowest to obtain a burner priority sequence. The higher the score, the more suitable the burner is for this adjustment.

[0102] Step 8: Determine the set of candidate burners Burners are selected in order of ranking, and the contribution coefficient α of each selected burner is accumulated. j The sum of α selected When α selected α, representing the total contribution coefficient of all burners total When the proportion of burners reaches or exceeds the preset coverage threshold θ (e.g., θ=80%) for the first time, selection stops. All burners selected at this point are then used as a candidate burner set.

[0103] The setting of the coverage threshold θ balances the adjustment effect and the adjustment cost: a higher threshold allows for the selection of more burners, ensuring a more reliable adjustment effect, but may cause unnecessary burner operation; a lower threshold allows for the selection of fewer burners, resulting in higher adjustment efficiency, but may lead to insufficient contribution. Typically, θ is set between 70% and 90%, and can be adjusted according to kiln characteristics and control requirements.

[0104] Thus, the system has selected a minimal and effective set of candidate burners from a large pool of burners. This set can cover most of the thermal impact capabilities in terms of contribution coefficients, while also taking into account the operating status of the burners.

[0105] Step 9: Power Allocation (followed by subsequent steps) After determining the candidate burner set, the system will enter the power allocation stage, which involves calculating the amount of power to be adjusted for each candidate burner according to the preset power allocation rules. This part will be described in detail in step S108, but it should be noted that the power allocation will be based on the candidate burner set and the comprehensive thermal influence weight will be used as the allocation basis to ensure that the total adjustment amount is reasonably allocated to each burner.

[0106] Step S108: Generate a coordinated control command based on the power to be adjusted, and send the coordinated control command to the corresponding burner actuator for temperature adjustment.

[0107] In this step, after step S107 determines the candidate burner set and the adjustable power of each candidate burner, this step needs to convert these adjustable power values ​​into executable control commands and ensure that multiple burners work in coordination to smoothly complete the power adjustment. The specific implementation process is as follows: Step 1: Calculate the target power value for each candidate burner For each burner j in the candidate burner set, obtain its current power value p. current-j (Can be read from the burner power data sequence in step S101), and the power to be adjusted Δp calculated in step S107. j (The signs are positive and negative; positive indicates that power needs to be increased, and negative indicates that power needs to be decreased.) Calculate the target power value p for this burner. target-j : p target-j =p current-j+Δp j , Perform boundary checks on the target power value to ensure it does not exceed the burner's allowable operating range. If p target-j Exceeding the burner's maximum power p max-j Then it is truncated to p max-j The excess power value is then proportionally redistributed to other candidate burners (the logic for this redistribution can be found in the allocation rules in step S107, and will not be repeated here); if p target-j Below the minimum stable power p of the burner min-j Then it will be promoted to p min-j Alternatively, the burner can be shut off directly (depending on process requirements).

[0108] Step 2: Obtain the dynamic response characteristics of the burner actuator Each burner's actuator (such as a gas valve, solenoid valve, frequency converter, etc.) has specific dynamic response characteristics, and the system pre-stores these characteristic parameters, including: Adjustment rate v j The power value that the actuator can change per unit time, expressed in kW / s. For different types of actuators, the adjustment rate may be constant or nonlinear; this scheme uses an average rate or a segmented rate for approximation.

[0109] Pure time delay τ delay-j The time delay from the issuance of an instruction to the start of a response by the actuator is usually caused by communication delays, mechanical inertia of the actuator, etc.

[0110] Inertial time constant τ inertia-j The time constant of the actuator during the transition from the start of response to reaching the target value is used to describe the smoothness of the regulation.

[0111] These parameters can be obtained from the actuator's technical manual or identified through on-site testing.

[0112] Step 3: Generate the power adjustment command sequence for each burner. Based on the target power value and dynamic response characteristics of each burner, corresponding adjustment commands are generated. The form of the commands depends on the control method of the actuator; there are two common types: Direct target value command: For actuators that support setting target power (such as combustion controllers with PID controllers), the target power value p can be directly issued. target-j The actuator will automatically adjust to this value. At this point, lag time needs to be considered, but there is no need to generate a ramp sequence.

[0113] Rate limiting ramp command: For actuators requiring an externally given adjustment rate, or for achieving fine control of the adjustment process, the system generates a time-varying power ramp command. Specifically, based on the current power p... current-j Target power p target-j and adjustment rate v j Calculate the required settling time t adj-j =|p target-j -p current-j | / v j The adjustment time is discretized into multiple control cycles (e.g., one point every 1 second), generating a linear interpolation sequence from the current power to the target power, with each time point corresponding to a power setpoint.

[0114] The second method is preferred in this solution because it allows for more precise control of the adjustment process and avoids temperature fluctuations caused by excessively fast or slow adjustments within the actuator.

[0115] Step 4: Calculate instruction delays to achieve coordinated adjustment When multiple burners operate simultaneously, if their adjustment rates and amounts differ, the time it takes to reach the target power may be inconsistent, easily causing an imbalance in heat injection and resulting in temperature fluctuations. To solve this problem, the system introduces a command delay mechanism to ensure that all burners reach a stable state simultaneously.

[0116] Let the longest settling time among all candidate burners be t. max =max(t adj-j For each burner j, calculate its instruction delay time t. delay-j =t max -t adj-j That is, burners with shorter adjustment times need to wait for a period of time before starting to adjust, so that they complete the adjustment at the same time as burners with the longest adjustment times.

[0117] For the existence of a pure time delay τ delay-j In cases where the actual response time begins is the command issuance time plus τ, this needs to be taken into account. delay-j To accurately align the settling time, the delay time can be corrected, but the pure delay time is usually short and the difference is not significant, so it can be ignored or treated as a safety margin.

[0118] Step 5: Encapsulate the collaborative control instruction package The adjustment commands for each burner (including delay time, power sequence, or target value) are encapsulated in a standardized format to form a coordinated control command package. The command package includes: Command header: contains command type identifier, timestamp, command packet length, checksum, etc. Burner entry list: Each entry contains burner ID and delay time τ. delay-j , Adjustment mode (ramp or direct target), power sequence data (if ramp mode) or target power value (if direct target mode).

[0119] The instruction packets are encapsulated using standardized communication protocols (such as Modbus TCP, Profinet, EtherCAT, etc.) to ensure compatibility with fieldbus.

[0120] Step 6: Issue instructions to the burner actuator The system distributes collaborative control command packets to each burner actuator via an industrial fieldbus. The distribution method can be broadcast (all burners simultaneously receive the same command packet and parse the corresponding command based on their own ID) or point-to-point (each burner sends its own command). Considering the real-time requirements of industrial environments, broadcasting is typically used to improve efficiency.

[0121] After the command is issued, each burner actuator parses the information in the command packet: Read the entry corresponding to its own ID; Based on the delay time τ delay-j Start a local timer; once the timer finishes, begin power regulation. In ramp mode, the output is adjusted point by point according to the preset power sequence; in direct target mode, the target power is set directly and adjusted by the internal controller.

[0122] Step 7: Real-time monitoring and closed-loop correction After the command is issued, the system continues to collect the actual power feedback values ​​of each burner at a high frequency (e.g., once every second), forming a closed-loop monitoring system. The monitoring content includes: Does the actual power change according to the command sequence? Does the actual time it takes for the power to reach the target value match the expectation? Whether abnormal fluctuations or overshoot occur during the adjustment process.

[0123] If the detected deviation between the actual power and the command value exceeds the preset tracking error threshold (e.g., 5%), or the adjustment time exceeds the expected range, the system determines that the adjustment is abnormal and triggers the command correction mechanism. Recalculate the remaining power to be adjusted based on the current actual power value; Regenerate correction instructions based on the remaining adjustment time; Immediately issue correction instructions to dynamically adjust the process.

[0124] This closed-loop correction mechanism ensures that even if there is nonlinearity in the actuator or external disturbances, the target power value can still be accurately achieved.

[0125] Step 8: Confirm adjustment complete When the actual power feedback values ​​of all candidate burners stabilize within the allowable error range of the target power value (e.g., ±2%) for a certain period of time (e.g., 10 seconds), the system confirms that the adjustment is complete, records the adjustment log, and returns control to the normal monitoring process (returning to step S101). If a serious abnormality that cannot be corrected occurs during the adjustment process (such as actuator failure), the system triggers an alarm and notifies the operator.

[0126] In summary, the method of this application, firstly, in the spatial region division stage (step S102), abandons the traditional control mode based on single points or fixed partitions. By constructing a heat transfer path topology map, calculating equivalent thermal resistance and time-series dynamic time warping distance, and introducing graph cutting clustering and gradient surface secondary segmentation, it achieves refined and adaptive partitioning of the complex temperature field inside the kiln. This process ensures that the temperature inside each final spatial region is uniform and the boundaries are clear, providing precise spatial units for subsequent independent control, fundamentally solving the problem of local over-firing or under-firing caused by the coarse region division in traditional methods.

[0127] Secondly, in the burner thermal impact modeling and power mapping stage (step S103), this invention establishes a normalized burner thermal impact weight matrix by measuring the temperature response amplitude of each temperature measurement point to the actual burner power change, and uses this matrix to map the dispersed burner power to the equivalent power value of each region. This mapping not only quantifies the degree of thermal impact of each burner on different regions, but also reduces the high-dimensional burner power information to a single thermal impact index for each region, providing an input dimension directly related to temperature deviation for judging regional temperature anomalies. This realizes the transformation from "burner perspective" to "regional perspective," greatly simplifying subsequent analysis and control.

[0128] In the anomaly detection stage (steps S104-S105), this invention first establishes an adaptive normal temperature fluctuation range for each region based on historical data statistics and dynamic correction of process parameters, overcoming the shortcomings of fixed thresholds that cannot adapt to changes in operating conditions. Furthermore, it employs a dual detection mechanism of "amplitude exceeding the limit + duration meeting the limit," effectively filtering out instantaneous interference and measurement noise, ensuring that adjustment is triggered only after confirming that the anomaly is continuously stable. This avoids energy waste and temperature oscillations caused by frequent ineffective adjustments, significantly improving the robustness and accuracy of control.

[0129] Once a true anomaly is confirmed, in the adjustment calculation stage (step S106), this invention constructs a temperature deviation vector, calculates the regional average deviation, and combines it with a regional proportional coefficient to accurately quantify the multi-point temperature deviation into a unified total power adjustment and a clear adjustment direction. This process not only preserves the distribution information of the deviation at each point within the region but also achieves a scientific mapping from the temperature domain to the power domain, providing a clear quantitative target for subsequent burner selection.

[0130] In the burner selection stage (step S107), this invention innovatively introduces non-negative matrix factorization technology to automatically extract physically meaningful thermal influence patterns from the thermal influence sub-matrix. By calculating the similarity between the temperature deviation vector and each pattern, the target pattern that best matches the current adjustment requirements is intelligently matched. Furthermore, by combining multi-dimensional operating status parameters such as the burner's contribution coefficient, current load, remaining lifespan, and fuel pressure, a comprehensive priority score is constructed. Finally, the optimal set of candidate burners is selected based on the contribution coverage rate as a constraint. This mechanism ensures that the selected burners can effectively correct the current temperature deviation while also considering equipment health management and adjustment efficiency, achieving an optimal balance between adjustment effect and operating cost.

[0131] Finally, in the instruction generation and execution stage (step S108), the present invention generates a ramp adjustment instruction with a delay based on the dynamic response characteristics of each candidate burner, and distributes it collaboratively via the industrial bus to ensure that all burners involved in the adjustment simultaneously and smoothly reach the target power, avoiding asynchronous heat injection and temperature fluctuations caused by inconsistent adjustment speeds. Simultaneously, a real-time feedback monitoring and closed-loop correction mechanism is introduced to dynamically correct deviations during execution, further ensuring adjustment accuracy and reliability.

[0132] In summary, this invention comprehensively solves the technical problems of traditional shuttle kiln temperature control, such as coarse zone division, frequent interference misjudgments, low adjustment efficiency, high energy consumption, and poor stability, through a series of interconnected innovative steps including spatial partitioning, influence modeling, adaptive thresholding, intelligent screening, and collaborative control. This method achieves precise zoned control of the kiln temperature field, intelligent scheduling of burner resources, and optimized power allocation. While improving temperature control accuracy, it effectively reduces energy consumption, extends equipment life, enhances the automation level of the production process, and improves product quality stability, demonstrating extremely high industrial application value and significant economic and social benefits.

[0133] Please see Figure 2 The diagram shows a structural block diagram of a shuttle kiln temperature control system according to this application.

[0134] like Figure 2As shown, the shuttle kiln temperature control system 200 includes an acquisition module 210, a division module 220, a mapping module 230, a determination module 240, a judgment module 250, a calculation module 260, a filtering module 270, and a generation module 280.

[0135] The system includes the following modules: Acquisition module 210, configured to acquire temperature data from multiple temperature measuring points inside the shuttle kiln at the current moment and power data from multiple burners, resulting in a temperature data sequence and a burner power data sequence; Division module 220, configured to divide the temperature data sequence according to a preset spatial adjacency relationship and temperature gradient change, resulting in at least one temperature data subsequence, where each temperature data subsequence corresponds to a continuous spatial region within the kiln; Mapping module 230, configured to map the burner power data sequence to each temperature data subsequence according to a pre-established burner thermal influence weight matrix, resulting in burner power data subsequences corresponding to each temperature data subsequence, where the burner thermal influence weight matrix characterizes the degree of thermal influence of each burner on each temperature data subsequence; Determination module 240, configured to determine a normal temperature fluctuation range for a spatial region based on historical temperature data of a spatial region corresponding to a certain temperature data subsequence within a historical time period; and Judgment module 250, configured to judge the temperature data... The calculation module 260 is configured to: determine whether the temperature difference of the subsequence exceeds a certain normal temperature fluctuation range and whether the duration exceeds a preset time threshold; calculate the deviation between the current temperature and the set target temperature of each temperature measurement point corresponding to the certain temperature data subsequence if the temperature difference exceeds the certain normal temperature fluctuation range and the duration exceeds the preset time threshold, obtain a certain temperature deviation vector, and determine a certain power adjustment amount and a certain adjustment direction of the certain spatial region based on the certain temperature deviation vector; filter module 270 is configured to: filter out a set of candidate burners to be adjusted from a certain burner power data subsequence corresponding to the certain temperature data subsequence based on the certain power adjustment amount, the certain adjustment direction, and the burner thermal influence weight matrix, and calculate the amount of power to be adjusted for each candidate burner in the set of candidate burners according to a preset power allocation rule; and generate module 280 is configured to: generate a cooperative control command based on the amount of power to be adjusted, and send the cooperative control command to the corresponding burner actuator for temperature adjustment.

[0136] It should be understood that Figure 2 The modules and references described in the document Figure 1 The steps described in the text correspond to those in the method described above. Therefore, the operations, features, and corresponding technical effects described above also apply to the method described in the text. Figure 2 The various modules in the document will not be described in detail here.

[0137] In other embodiments, the present invention also provides a computer-readable storage medium having a computer program stored thereon, wherein when the program instructions are executed by a processor, the processor performs the shuttle kiln temperature control method in any of the above method embodiments. In one embodiment, the computer-readable storage medium of the present invention stores computer-executable instructions, which are configured as follows: The temperature data of multiple temperature measuring points inside the shuttle kiln at the current moment and the power data of multiple burners are obtained to obtain the temperature data sequence and the burner power data sequence. Based on the preset spatial adjacency relationship and temperature gradient change, the temperature data sequence is divided to obtain at least one temperature data subsequence, wherein each temperature data subsequence corresponds to a continuous spatial region within the kiln. Based on the pre-established burner thermal influence weight matrix, the burner power data sequence is mapped to each temperature data subsequence to obtain the burner power data subsequence corresponding to each temperature data subsequence. The burner thermal influence weight matrix is ​​used to characterize the degree of thermal influence of each burner on each temperature data subsequence. Based on a certain historical temperature data of a certain spatial region within a certain historical time period, a certain normal temperature fluctuation range of a certain spatial region is determined. Determine whether the temperature difference of a certain temperature data subsequence exceeds a certain normal temperature fluctuation range and whether the duration exceeds a preset time threshold. If the temperature fluctuation exceeds a certain normal range and the duration exceeds a preset time threshold, the deviation between the current temperature of each temperature measurement point corresponding to the certain temperature data subsequence and the set target temperature is calculated to obtain a certain temperature deviation vector, and a certain power adjustment amount and a certain adjustment direction of the certain spatial region are determined based on the certain temperature deviation vector. Based on the total power adjustment, the adjustment direction, and the burner thermal influence weight matrix, a set of candidate burners to be adjusted is selected from a subsequence of burner power data corresponding to the subsequence of temperature data, and the power to be adjusted for each candidate burner in the set of candidate burners is calculated according to the preset power allocation rules. Based on the power to be adjusted, a coordinated control command is generated and sent to the corresponding burner actuator for temperature adjustment.

[0138] Computer-readable storage media may include a stored program area and a stored data area, wherein the stored program area may store an operating system and an application program required for at least one function; the stored data area may store data created based on the use of the shuttle kiln temperature control system, etc. Furthermore, the computer-readable storage medium may include high-speed random access memory, and may also include memory, such as at least one disk storage device, flash memory device, or other non-volatile solid-state storage device. In some embodiments, the computer-readable storage medium may optionally include memory remotely configured relative to a processor, and this remote memory may be connected to the shuttle kiln temperature control system via a network. Examples of such networks include, but are not limited to, the Internet, corporate intranets, local area networks, mobile communication networks, and combinations thereof.

[0139] Figure 3 This is a schematic diagram of the structure of the electronic device provided in the embodiment of the present invention, such as... Figure 3 As shown, the device includes a processor 310 and a memory 320. The electronic device may also include an input device 330 and an output device 340. The processor 310, memory 320, input device 330, and output device 340 can be connected via a bus or other means. Figure 3 Taking a bus connection as an example, the memory 320 is the computer-readable storage medium described above. The processor 310 executes various server functions and data processing by running non-volatile software programs, instructions, and modules stored in the memory 320, thereby implementing the shuttle kiln temperature control method described in the above embodiment. The input device 330 can receive input digital or character information and generate key signal inputs related to user settings and function control of the shuttle kiln temperature control system. The output device 340 may include a display screen or other display device.

[0140] The aforementioned electronic device can execute the method provided in the embodiments of the present invention, and has the corresponding functional modules and beneficial effects for executing the method. Technical details not described in detail in this embodiment can be found in the method provided in the embodiments of the present invention.

[0141] In one implementation, the aforementioned electronic device is used in a shuttle kiln temperature control system as a client, comprising: at least one processor; and a memory communicatively connected to the at least one processor; wherein the memory stores instructions executable by the at least one processor, the instructions being executed by the at least one processor to enable the at least one processor to: The temperature data of multiple temperature measuring points inside the shuttle kiln at the current moment and the power data of multiple burners are obtained to obtain the temperature data sequence and the burner power data sequence. Based on the preset spatial adjacency relationship and temperature gradient change, the temperature data sequence is divided to obtain at least one temperature data subsequence, wherein each temperature data subsequence corresponds to a continuous spatial region within the kiln. Based on the pre-established burner thermal influence weight matrix, the burner power data sequence is mapped to each temperature data subsequence to obtain the burner power data subsequence corresponding to each temperature data subsequence. The burner thermal influence weight matrix is ​​used to characterize the degree of thermal influence of each burner on each temperature data subsequence. Based on a certain historical temperature data of a certain spatial region within a certain historical time period, a certain normal temperature fluctuation range of a certain spatial region is determined. Determine whether the temperature difference of a certain temperature data subsequence exceeds a certain normal temperature fluctuation range and whether the duration exceeds a preset time threshold. If the temperature fluctuation exceeds a certain normal range and the duration exceeds a preset time threshold, the deviation between the current temperature of each temperature measurement point corresponding to the certain temperature data subsequence and the set target temperature is calculated to obtain a certain temperature deviation vector, and a certain power adjustment amount and a certain adjustment direction of the certain spatial region are determined based on the certain temperature deviation vector. Based on the total power adjustment, the adjustment direction, and the burner thermal influence weight matrix, a set of candidate burners to be adjusted is selected from a subsequence of burner power data corresponding to the subsequence of temperature data, and the power to be adjusted for each candidate burner in the set of candidate burners is calculated according to the preset power allocation rules. Based on the power to be adjusted, a coordinated control command is generated and sent to the corresponding burner actuator for temperature adjustment.

[0142] Through the above description of the embodiments, those skilled in the art can clearly understand that each embodiment can be implemented by means of software plus necessary general-purpose hardware platforms, and of course, it can also be implemented by hardware. Based on this understanding, the above technical solutions, in essence or the part that contributes to the prior art, can be embodied in the form of a software product. This computer software product can be stored in a computer-readable storage medium, such as ROM / RAM, magnetic disk, optical disk, etc., including several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute the methods of various embodiments or some parts of embodiments.

[0143] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, and not to limit them; although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features; and these modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A method for temperature control in a shuttle kiln, characterized in that, include: The temperature data of multiple temperature measuring points inside the shuttle kiln at the current moment and the power data of multiple burners are obtained to obtain the temperature data sequence and the burner power data sequence. Based on the preset spatial adjacency relationship and temperature gradient change, the temperature data sequence is divided to obtain at least one temperature data subsequence, wherein each temperature data subsequence corresponds to a continuous spatial region within the kiln. Based on the pre-established burner thermal influence weight matrix, the burner power data sequence is mapped to each temperature data subsequence to obtain the burner power data subsequence corresponding to each temperature data subsequence. The burner thermal influence weight matrix is ​​used to characterize the degree of thermal influence of each burner on each temperature data subsequence. Based on a certain historical temperature data of a certain spatial region within a certain historical time period, a certain normal temperature fluctuation range of a certain spatial region is determined. Determine whether the temperature difference of a certain temperature data subsequence exceeds a certain normal temperature fluctuation range and whether the duration exceeds a preset time threshold. If the temperature fluctuation exceeds a certain normal range and the duration exceeds a preset time threshold, the deviation between the current temperature of each temperature measurement point corresponding to the certain temperature data subsequence and the set target temperature is calculated to obtain a certain temperature deviation vector, and a certain power adjustment amount and a certain adjustment direction of the certain spatial region are determined based on the certain temperature deviation vector. Based on the total power adjustment, the adjustment direction, and the burner thermal influence weight matrix, a set of candidate burners to be adjusted is selected from a subsequence of burner power data corresponding to the subsequence of temperature data, and the power to be adjusted for each candidate burner in the set of candidate burners is calculated according to the preset power allocation rules. Based on the power to be adjusted, a coordinated control command is generated and sent to the corresponding burner actuator for temperature adjustment.

2. The method for controlling the temperature of a shuttle kiln according to claim 1, characterized in that, The step of dividing the temperature data sequence according to a preset spatial adjacency relationship and temperature gradient change to obtain at least one temperature data subsequence includes: The three-dimensional spatial coordinates of each temperature measuring point inside the kiln are obtained, and a heat transfer path topology map inside the kiln is constructed based on the three-dimensional spatial coordinates. The heat transfer path topology map uses the temperature measuring points as nodes and the heat conduction channels between adjacent temperature measuring points as edges. Based on the heat transfer path topology diagram, the equivalent thermal resistance between any two adjacent nodes is calculated. The equivalent thermal resistance is determined based on the spatial distance between nodes, the thermal conductivity of the medium, and the airflow disturbance coefficient. Obtain the temperature value of each node at the current moment and the temperature values ​​of the previous M historical moments to form the temperature time-series vector of each node; For each edge connecting two nodes, calculate the dynamic time-warped distance between the temperature time-series vectors of the two nodes; A comprehensive coupling coefficient between nodes is constructed based on a weighted combination of the equivalent thermal resistance and the dynamic time warping distance. Based on the edges whose comprehensive coupling coefficient is greater than a preset first threshold, the nodes are divided and clustered to obtain multiple initial connected subgraphs. Spatial interpolation is performed on the temperature values ​​inside a certain initial connected subgraph to generate a continuous temperature distribution surface, and the gradient magnitude of each point on the temperature distribution surface is calculated, wherein the certain initial connected subgraph is any one of the multiple initial connected subgraphs; The proportion of points whose gradient magnitude exceeds a preset second threshold is counted. If the proportion is lower than a preset third threshold, an initial connected subgraph is determined to be a temperature uniform region. Otherwise, the initial connected subgraph is divided into two parts to obtain at least one target connected subgraph. The initial connected subgraph determined to be a temperature uniform region and the target connected subgraph obtained after secondary segmentation are used as the final spatial region. The current temperature data of all nodes in each final spatial region are combined in order of node number to obtain the at least one temperature data subsequence.

3. The method for controlling the temperature of a shuttle kiln according to claim 1, characterized in that, The step of mapping the burner power data sequence to various temperature data subsequences based on a pre-established burner thermal influence weight matrix, to obtain burner power data subsequences corresponding to each temperature data subsequence, includes: When the kiln is in stable operating condition, the temperature response amplitude of each temperature measuring point is obtained when the power of each burner changes. Divide the temperature response amplitude of a certain burner to a certain temperature measuring point by the power adjustment amount of the certain burner to obtain the unit power influence coefficient of a certain burner to a certain temperature measuring point. The unit power influence coefficients of all burners on all temperature measurement points are combined into an initial weight matrix, and the initial weight matrix is ​​normalized to obtain the final burner thermal influence weight matrix. The normalization process includes making the sum of the elements in each row of the initial weight matrix equal to 1, that is, the sum of the influence coefficients of all burners on each temperature measurement point is equal to 1. Extract the weighted row vectors corresponding to all temperature measurement points within a certain spatial region from the burner thermal influence weight matrix, wherein the certain spatial region is the spatial region corresponding to a certain temperature data subsequence; The average value of all extracted weight row vectors is calculated along the column direction to obtain the comprehensive weight vector. Each element of the comprehensive weight vector represents the degree of comprehensive thermal influence of the corresponding burner on a certain spatial region. Obtain the power values ​​of all burners at the current moment and construct a power vector; The power vector is multiplied by the comprehensive weight vector, that is, the elements at corresponding positions are multiplied and then summed to obtain the equivalent power value of a certain spatial region. The equivalent power values ​​of all spatial regions are combined according to the spatial location order of the spatial regions to obtain the burner power data subsequence corresponding to each temperature data subsequence.

4. The method for controlling the temperature of a shuttle kiln according to claim 1, characterized in that, The step of determining a normal temperature fluctuation range for a spatial region based on historical temperature data of a spatial region within a historical time period, corresponding to a certain temperature data subsequence, includes: Historical temperature data of a certain spatial region within the same time window as the current firing stage are obtained over multiple historical firing cycles to form a historical temperature dataset. Calculate the mean and standard deviation of the historical temperature dataset; The range of the average value plus or minus three standard deviations is taken as the normal temperature fluctuation range of the certain spatial region. Obtain the process parameters for the current firing cycle, including loading amount and product type; Based on the preset correspondence table between process parameters and fluctuation range, find the correction coefficient corresponding to the current process parameters; By multiplying the upper and lower limits of the normal temperature fluctuation range by the correction coefficient, a normal temperature fluctuation range for a certain spatial region is obtained.

5. The method for controlling the temperature of a shuttle kiln according to claim 1, characterized in that, Before determining whether the temperature difference of a certain temperature data subsequence exceeds a certain normal temperature fluctuation range and whether the duration exceeds a preset time threshold, the method further includes: Calculate the average current temperature of all temperature measurement points in a certain temperature data subsequence, and use it as the representative temperature of the region; The absolute value of the difference between the temperature represented by the region and the center value of the normal temperature fluctuation range is calculated as the degree of temperature difference.

6. The method for controlling the temperature of a shuttle kiln according to claim 1, characterized in that, The calculation of the deviation between the current temperature and the set target temperature of each temperature measurement point corresponding to the certain temperature data subsequence, to obtain a certain temperature deviation vector, and the determination of a certain power adjustment amount and a certain adjustment direction for a certain spatial region based on the certain temperature deviation vector, includes: Obtain the numbering information of each temperature measuring point within a certain spatial region corresponding to a certain temperature data subsequence, and determine the spatial location of each temperature measuring point within the kiln based on the numbering information; Based on the preset firing curve and the current firing stage, determine the target temperature value corresponding to each temperature measuring point, and construct a target temperature vector based on the spatial position of each temperature measuring point in the kiln. Extract the current temperature value of each temperature measurement point at the current moment from a certain temperature data subsequence, and arrange them in the same order as the target temperature vector to form the current temperature vector; Subtract the corresponding element in the target temperature vector from each element in the current temperature vector to obtain a temperature deviation vector composed of multiple differences. Each element of the temperature deviation vector represents the deviation between the current temperature value of the corresponding temperature measurement point and the set target temperature value. A positive deviation value indicates that the current temperature value is higher than the target temperature value, and a negative deviation value indicates that the current temperature value is lower than the target temperature value. The arithmetic mean of all elements in the temperature deviation vector is used to obtain the average temperature deviation of a certain spatial region. A certain adjustment direction is determined based on the sign of the average temperature deviation. If the average temperature deviation is positive, then the adjustment direction for a certain spatial region is the cooling direction; if the average temperature deviation is negative, then the adjustment direction for a certain spatial region is the heating direction. A certain proportional coefficient is obtained for a certain spatial region. The certain proportional coefficient is pre-calibrated based on the volume of the spatial region, the heat capacity characteristics of the kiln inner wall, and historical adjustment experience data. It is used to characterize the amount of power adjustment required per unit temperature deviation. Multiplying the absolute value of the average temperature deviation by the proportional coefficient yields the total power regulation for a given spatial region.

7. The method for controlling the temperature of a shuttle kiln according to claim 1, characterized in that, The step of selecting a set of candidate burners to be adjusted from a burner power data subsequence corresponding to a certain temperature data subsequence based on the certain total power adjustment, the certain adjustment direction, and the burner thermal influence weight matrix includes: Obtain the weight row vectors corresponding to all temperature measurement points within a certain temperature data subsequence in the burner thermal influence weight matrix, and construct the thermal influence submatrix; The thermal influence submatrix is ​​decomposed into a nonnegative matrix decomposition into a product of a basis matrix and a coefficient matrix. Each column of the basis matrix represents a potential thermal influence mode, and each row of the coefficient matrix represents the contribution of a burner to each thermal influence mode. Based on a certain adjustment direction, target feature patterns that match the heating or cooling requirements are selected from the base matrix; Extract the column vectors corresponding to the target feature patterns from the coefficient matrix to obtain the contribution coefficient of each burner to the target feature patterns; Obtain the current operating status parameters of each burner, including current power value, cumulative running time, and fuel supply pressure; A comprehensive evaluation of burner adjustment priority is constructed based on the contribution coefficient of each burner, the ratio of current power value to rated power, the ratio of cumulative running time to design life, and the ratio of fuel supply pressure to rated pressure. The burners are sorted from highest to lowest according to the comprehensive score value; Select burners sequentially from the sorted list, calculate the sum of the contribution coefficients of the selected burners, and continue until the sum of the contribution coefficients of the selected burners reaches the preset coverage threshold as a proportion of the total contribution coefficients of all burners. Then, select the burners as a candidate burner set.

8. The method for controlling the temperature of a shuttle kiln according to claim 1, characterized in that, The calculation of the adjustable power for each candidate burner in the candidate burner set according to the preset power allocation rules includes: Obtain the comprehensive thermal influence weight of each candidate burner in the candidate burner set on the certain spatial region, and the combustion efficiency curve of each candidate burner under the current operating conditions. The combustion efficiency curve represents the conversion efficiency between fuel calorific value and actual effective calorific value under different power outputs. Obtain the power regulation response characteristic parameters for each candidate burner, including minimum adjustable power, maximum adjustable power, power regulation step size, and the shortest time required to adjust from the current power to the target power; A multi-objective optimization allocation model is constructed with the goal of minimizing total fuel consumption, the power adjustment range of each burner as a constraint, and the sum of the products of the comprehensive thermal influence weight of each burner and the power adjustment amount equal to the total power adjustment amount as an equality constraint. The multi-objective optimization allocation model is transformed into a single-objective optimization problem. The total fuel consumption and the regulation response time are weighted and combined using the weighted summation method to obtain the comprehensive optimization objective function. The particle swarm optimization algorithm is used to solve the comprehensive optimization objective function, and the optimal power allocation scheme is searched in the solution space. In the process of particle swarm optimization, a simulated annealing mechanism is introduced. In the early stage of iteration, a certain probability of accepting a deteriorated solution is allowed to avoid getting trapped in a local optimum. In the later stage of iteration, the acceptance probability is gradually reduced to accelerate convergence. When the maximum number of iterations is reached or the change in the solution is less than the preset convergence threshold, the iteration stops and the current optimal solution is output as the initial adjustable power of each candidate burner. Post-processing verification is performed on each power quantity to be adjusted to check whether it meets the minimum and maximum adjustable power constraints of each burner. If it does not meet the constraints, the nearest integer or boundary truncation is performed to obtain the power quantity to be adjusted.

9. A shuttle kiln temperature control system, characterized in that, include: The acquisition module is configured to acquire temperature data from multiple temperature measuring points inside the shuttle kiln at the current moment, as well as power data from multiple burners, to obtain temperature data sequences and burner power data sequences. The partitioning module is configured to partition the temperature data sequence according to a preset spatial adjacency relationship and temperature gradient change to obtain at least one temperature data subsequence, wherein each temperature data subsequence corresponds to a continuous spatial region within the kiln. The mapping module is configured to map the burner power data sequence to each temperature data subsequence according to a pre-established burner thermal influence weight matrix, thereby obtaining burner power data subsequences corresponding to each temperature data subsequence. The burner thermal influence weight matrix is ​​used to characterize the degree of thermal influence of each burner on each temperature data subsequence. The module is configured to determine a normal temperature fluctuation range for a spatial region based on historical temperature data of a spatial region within a historical time period corresponding to a certain temperature data subsequence. The judgment module is configured to determine whether the temperature difference of a certain temperature data subsequence exceeds a certain normal temperature fluctuation range and whether the duration exceeds a preset time threshold. The calculation module is configured to, if the temperature exceeds a certain normal temperature fluctuation range and the duration exceeds a preset time threshold, calculate the deviation between the current temperature of each temperature measurement point corresponding to the certain temperature data subsequence and the set target temperature, obtain a certain temperature deviation vector, and determine a certain power adjustment amount and a certain adjustment direction of the certain spatial region based on the certain temperature deviation vector. The filtering module is configured to filter out a set of candidate burners to be adjusted from a set of burner power data corresponding to a set of temperature data, based on the total power adjustment, the adjustment direction, and the burner thermal influence weight matrix, and to calculate the amount of power to be adjusted for each candidate burner in the set of candidate burners according to a preset power allocation rule. The generation module is configured to generate a collaborative control command based on the power to be adjusted, and send the collaborative control command to the corresponding burner actuator for temperature adjustment.

Images