Incinerator temperature field optimization method and system based on CFD numerical simulation

Abstract

The invention provides an incinerator temperature field optimization method and system based on CFD numerical simulation, and the method comprises the steps: firstly recognizing a dynamic interference event in the operation of an incinerator, extracting the feature information of the event, constructing a hearth space association graph, learning the association relation between the interference feature and the temperature field abnormality through a graph neural network, and outputting a temperature abnormality association link; inputting the data into a CFD simulation system to generate and execute a target simulation task; acquiring dynamic response data of a temperature field; analyzing the data to determine a simulation correction direction; generating a parameter adjustment instruction to update simulation parameters; and finally, extracting a final simulation parameter as an optimization scheme, so that the temperature field of the incinerator can be accurately optimized.

Description

Technical Field

[0001] The present invention relates to the field of computer technology, and in particular to a method and system for optimizing the temperature field of an incinerator based on CFD numerical simulation. Background Art

[0002] In incinerator operation, maintaining and optimizing the temperature field plays a crucial role in incineration efficiency, pollutant emission control, and equipment lifespan. During actual incinerator operation, incinerators can be affected by a variety of dynamic disturbances, such as sudden changes in fuel composition, fluctuations in feed rate, and abnormal adjustments in the combustion air volume. These dynamic disturbances, characterized by randomness and uncertainty, can disrupt the incinerator's thermal equilibrium and cause abnormal temperature field fluctuations.

[0003] Traditional incinerator temperature field control methods mainly rely on empirical formulas and simple feedback adjustment mechanisms. Empirical formulas are often summarized based on specific operating conditions and are difficult to accurately adapt to the complex and changeable actual operating environment. Simple feedback adjustment mechanisms usually make adjustments only after the temperature has significantly deviated. There is a problem of adjustment lag, and it is impossible to respond to the impact of dynamic interference events on the temperature field in a timely and effective manner. In addition, traditional methods lack in-depth analysis and utilization of the heat conduction relationship in the internal space of the incinerator furnace, making it difficult to accurately grasp the root cause and propagation path of temperature field anomalies, resulting in poor temperature field optimization effects. It is easy for local temperatures to be too high or too low, affecting the overall operating performance of the incinerator. Summary of the Invention

[0004] In view of the above-mentioned problems, in combination with the first aspect of the present invention, an embodiment of the present invention provides an incinerator temperature field optimization method based on CFD numerical simulation, the method comprising:

[0005] Identify dynamic interference events during the operation of the incinerator and extract event feature information of the dynamic interference events, wherein the event feature information includes the interference type, the interference occurrence time period and the interference impact range;

[0006] Construct a spatial association graph for the incinerator furnace, using each spatial unit of the furnace as a graph node and the heat conduction relationship between each spatial unit of the furnace as the graph edge weight. Embed the event feature information into the corresponding graph node, learn the association between interference features and temperature field anomalies through a graph neural network, and output the temperature anomaly association link corresponding to the interference event.

[0007] Input the temperature anomaly associated link into the CFD simulation system to generate a target simulation task for the associated link, wherein the target simulation task includes a target simulation area, simulation duration, and key monitoring temperature indicators;

[0008] Execute the target simulation task, obtain the temperature field dynamic response data of the associated links, analyze the embedding changes of the graph nodes corresponding to the temperature field dynamic response data through the graph neural network, determine the correlation strength between the temperature field response and the interference event, and then determine the simulation correction direction that needs to be adjusted;

[0009] generating a CFD simulation parameter adjustment instruction according to the simulation correction direction, inputting the CFD simulation parameter adjustment instruction into the CFD simulation system to update simulation parameters, re-executing the target simulation task, and obtaining corrected temperature field response data;

[0010] Compare the embedding differences of the graph nodes corresponding to the temperature field response data before and after correction to determine whether the abnormal temperature field pattern has been alleviated. If not, re-identify the dynamic interference event characteristics and update the furnace space association graph until the abnormal temperature field pattern is eliminated. Extract the final simulation parameters as the incinerator temperature field optimization plan.

[0011] On the other hand, an embodiment of the present invention also provides an incinerator temperature field optimization system based on CFD numerical simulation, including a processor and a machine-readable storage medium, the machine-readable storage medium is connected to the processor, the machine-readable storage medium is used to store programs, instructions or codes, and the processor is used to execute the programs, instructions or codes in the machine-readable storage medium to implement the above method.

[0012] Based on the above aspects, the embodiment of the present invention identifies dynamic interference events in the operation of the incinerator and extracts event feature information, constructs an incinerator furnace space association graph and embeds event feature information into the graph nodes, uses a graph neural network to learn the correlation between interference features and temperature field anomalies, reveals the internal mechanism and propagation path of temperature field anomalies, can accurately locate the temperature anomaly correlation link corresponding to the interference event, inputs the temperature anomaly correlation link into the CFD simulation system to generate a target simulation task, performs simulation analysis on key links in a targeted manner, improves simulation efficiency and accuracy, executes the target simulation task to obtain temperature field dynamic response data, and determines the correlation strength between the temperature field response and the interference event and the simulation correction direction through graph neural network analysis, generates CFD simulation parameter adjustment instructions according to the simulation correction direction and updates the parameters, re-executes the simulation task to obtain the corrected temperature field response data, compares the data differences before and after correction to determine whether the temperature field anomaly pattern is alleviated, and if not, re-identifies and updates until the anomaly is eliminated. The finally extracted simulation parameters are used as an optimization scheme, which can effectively eliminate temperature field anomalies, improve the stability and uniformity of the incinerator temperature field, improve incineration efficiency, reduce pollutant emissions, and enhance the adaptability of the incinerator under complex working conditions. BRIEF DESCRIPTION OF THE DRAWINGS

[0013] Figure 1It is a schematic diagram of the execution flow of the incinerator temperature field optimization method based on CFD numerical simulation provided by an embodiment of the present invention.

[0014] Figure 2 It is a schematic diagram of exemplary hardware and software components of an incinerator temperature field optimization system based on CFD numerical simulation provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0015] The present invention will be described in detail below with reference to the accompanying drawings. Figure 1 This is a flow chart of an incinerator temperature field optimization method based on CFD numerical simulation provided by an embodiment of the present invention. The incinerator temperature field optimization method based on CFD numerical simulation is introduced in detail below.

[0016] Step S110: identifying dynamic interference events during the operation of the incinerator, and extracting event feature information of the dynamic interference events, wherein the event feature information includes interference type, interference occurrence time period, and interference impact range.

[0017] This embodiment uses a special incinerator for medical waste in domestic waste as an application scenario. The incinerator is used to treat special waste such as infectious waste and pathological waste, and its operational stability is directly related to the treatment effect and environmental safety. During operation, dynamic interference events can come from many aspects, such as the rupture of the container for loading medical waste, resulting in uneven feeding, a slight leakage in the combustion-supporting agent supply pipeline affecting combustion efficiency, and the shedding of local refractory materials in the furnace, which changes the heat reflection characteristics. These events will disrupt the original thermal balance in the furnace and cause temperature field anomalies. In order to effectively identify dynamic interference events, it is necessary to rely on a monitoring system covering the key areas of the incinerator, continuously collect and analyze various types of operating data, capture abnormal signals, and then extract the core feature information of the event.

[0018] Step S111: collecting operation data at a preset frequency through each monitoring point in the incinerator operation status monitoring network, wherein the operation data includes feed data, burner data, flue gas data and furnace wall data.

[0019] The waste incinerator's operating status monitoring network includes multiple monitoring points, which are distributed according to functional areas. Weight sensors, infrared spectrometers, and conveyor belt speed monitors are installed near the feed inlet to collect feed data, including the feed weight per unit time, the spectral characteristics of the waste composition (reflecting the ratio of organic and inorganic matter), and the conveyor belt's operating speed. The burner area is equipped with thermocouples, pressure sensors, and flow metering devices to collect burner data, specifically covering the temperature distribution of the combustion flame, fuel supply pressure, combustion air flow rate, and oxygen content. Gas analyzers, temperature sensors, and pressure transmitters are arranged at the flue gas outlet to collect flue gas data, including the concentration of various pollutants in the flue gas, flue gas temperature, and flow pressure. Infrared thermometers are installed at different heights and angles on the inner wall of the furnace to obtain furnace wall data, namely the real-time temperature values ​​of various areas of the wall.

[0020] The collection frequency for each monitoring point is set based on the dynamic nature of the data. Weight and velocity information from the feed data is collected at relatively short intervals, while spectral signature analysis is performed at slightly longer intervals. Temperature and pressure data from the burner are collected more frequently to capture transient changes in the flame. Pollutant concentrations from the flue gas data are monitored at set intervals, while temperature and pressure are collected continuously. Furnace wall temperature data is collected at moderate intervals, balancing data volume and timeliness. All collected data is aggregated via wired transmission to a central data storage unit, forming a continuous operational data sequence.

[0021] Step S112: Taking each monitoring point as the initial node and the standardized spatial distance and data correlation between monitoring points as the graph edge weight, a monitoring point spatial correlation graph is constructed, and node anomaly detection is performed on the monitoring point spatial correlation graph through a graph neural network to capture abnormal nodes.

[0022] Each monitoring point is defined as the initial node of the monitoring point spatial association graph. Each node carries a unique identifier that includes its installation location and the type of monitoring data. For example, the node corresponding to the weight sensor at the feed inlet is labeled "Feed Weight Monitoring Point," and the node corresponding to the thermocouple of Burner 1 is labeled "Burner 1 Temperature Monitoring Point."

[0023] Step S1121: Walk through the distribution of monitoring points in the incinerator operation status monitoring network, use each monitoring point as the initial node of the monitoring point spatial association graph, and record the spatial coordinates and monitoring data type of each initial node.

[0024] Using pre-stored drawings of monitoring point installations, the spatial coordinates of each monitoring point in the incinerator's three-dimensional coordinate system are obtained, with the coordinate origin set at the center of the incinerator's feed port. At the same time, the monitoring data types are clearly marked in the node attributes, such as "feed weight," "combustion temperature," and "flue gas pollutant concentration." For example, the spatial coordinates of the infrared spectrometer at the feed port are (x1, y1, z1), and the monitoring data type is "feed component spectral characteristics"; the spatial coordinates of the infrared thermometer in the middle layer of the east wall of the furnace are (x2, y2, z2), and the monitoring data type is "wall temperature." This information is stored in the node attribute table as the basic data for constructing the association graph.

[0025] Step S1122: Calculate the normalized spatial distance between any two initial nodes, combine the historical correlation coefficient of the monitoring data of the two initial nodes, and use the product of the inverse of the normalized spatial distance and the historical correlation coefficient as the graph edge weight between the two initial nodes.

[0026] For any two initial nodes, the straight-line distance is calculated based on their spatial coordinates, and then divided by the maximum spatial diagonal length of the incinerator furnace to obtain the standardized spatial distance, which ranges from 0 to 1. For example, if the actual spatial distance between nodes A and B is D, and the maximum spatial diagonal length of the furnace is L, then the standardized spatial distance is the ratio of D to L.

[0027] At the same time, the historical monitoring data series of the two nodes over the past period are retrieved and the correlation coefficient between them is calculated. The calculation process is as follows: After normalizing the two sets of data series to eliminate the influence of dimension, the ratio of their covariance and the product of their respective standard deviations is calculated to obtain the historical correlation coefficient, which ranges from -1 to 1.

[0028] The inverse of the normalized spatial distance is multiplied by the historical correlation coefficient to obtain the edge weight between the two initial nodes. The closer the distance between the two nodes and the more consistent the data trends, the larger the edge weight, indicating a stronger spatial and functional connection between them.

[0029] Step S1123: Based on the initial nodes and graph edge weights, a monitoring point spatial association graph is constructed, and the graph autoencoder in the graph neural network is used to pre-train the monitoring point spatial association graph to learn the embedding vectors of each initial node under normal operating conditions.

[0030] Based on the attribute information of the initial nodes and the calculated graph edge weights, a monitoring point spatial association graph is constructed. The monitoring point spatial association graph uses nodes to represent monitoring points and weighted edges to represent the association strength between monitoring points.

[0031] A graph autoencoder is used to pre-train this association graph. The graph autoencoder consists of an encoder and a decoder. The encoder includes multiple graph convolutional layers, each of which aggregates node features with those of its adjacent nodes. Specifically, the features of each node are weighted and combined with those of its adjacent nodes according to the graph edge weights. This is then transformed nonlinearly using an activation function to obtain a more abstract feature representation. After processing through multiple graph convolutional layers, the encoder outputs a low-dimensional embedding vector for each node.

[0032] The decoder receives these embeddings and attempts to reconstruct the original node features and graph edge weights through operations such as transposed convolution. During pre-training, historical data from the incinerator during normal operation is used as training samples. Backpropagation is used to adjust network parameters to minimize the error between the reconstructed results and the original input. After sufficient iterative training, the graph autoencoder learns stable embeddings for each initial node under normal operation. These stable embeddings contain both the node's own characteristics and its relationships with other nodes.

[0033] Step S1124: Collect the operating data of each monitoring point in real time, convert the operating data into a real-time node feature vector, input the pre-trained graph autoencoder, and calculate the reconstruction error between the real-time node embedding vector and the initial node embedding vector under normal operating conditions.

[0034] Real-time operational data is first preprocessed, including outlier removal (removing data that clearly exceeds the physically meaningful range), missing value filling (using interpolation of data from adjacent moments), and data normalization (converting data of varying magnitudes to the same numerical range). After processing, the data from each monitoring point is organized into a vector, the real-time node feature vector. The dimensionality of the vector is determined by the number of indicators in the monitoring data. For example, a feature vector for a burner monitoring point might contain normalized values ​​for multiple indicators, such as flame temperature, fuel pressure, and air flow.

[0035] The real-time node feature vector is fed into the encoder portion of a pre-trained graph autoencoder to generate a real-time node embedding vector. This embedding vector is then fed into the decoder to obtain reconstructed node features and graph edge weights. The difference between the reconstructed result and the real-time input data is calculated, which is the reconstruction error. The reconstruction error is calculated by combining the node feature reconstruction error and the graph edge weight reconstruction error, and the two are weighted and summed according to a predetermined ratio to obtain the total reconstruction error.

[0036] Step S1125: Set a reconstruction error threshold, mark nodes whose reconstruction errors exceed the reconstruction error threshold as candidate abnormal nodes, perform neighborhood node verification on the candidate abnormal nodes, and check whether the reconstruction errors of the neighboring nodes of the candidate abnormal nodes also exceed the reconstruction error threshold.

[0037] The reconstruction error threshold is determined by analyzing the historical reconstruction error data during normal operation of the incinerator. The statistical distribution of the historical reconstruction error is calculated, and an appropriate quantile is selected as the threshold so that only a very small number of nodes will have reconstruction errors exceeding the threshold during normal operation.

[0038] The real-time reconstruction error of each node is compared with the threshold. Nodes with reconstruction errors greater than the threshold are marked as candidate outlier nodes. For each candidate outlier node, all adjacent nodes with direct edges to it are found in the monitoring point spatial association graph and checked to see if their real-time reconstruction errors also exceed the threshold.

[0039] Step S1126: If the reconstruction errors of the nodes exceeding a preset ratio among the neighboring nodes of the candidate abnormal node exceed the reconstruction error threshold, the candidate abnormal node is determined to be a final abnormal node, and all final abnormal nodes are extracted to form an abnormal node set.

[0040] The preset ratio is determined by the density of connections in the spatial correlation graph of the monitoring points. The denser the connections, the lower the preset ratio. For example, if a candidate anomaly node has multiple adjacent nodes, and the reconstruction error of more than one-third of these adjacent nodes exceeds the threshold, this indicates that the anomaly at that node is not isolated but is influenced by anomalies in surrounding nodes, or is itself the source of anomaly diffusion. In this case, the candidate anomaly node is determined to be the final anomaly node.

[0041] All final abnormal nodes that meet the above conditions are aggregated to form an abnormal node set.

[0042] Step S113: Count the duration of data fluctuations of abnormal nodes, and determine the start time and end time of interference corresponding to the abnormal nodes in combination with the incinerator operation mechanism to form the interference occurrence period.

[0043] For each node in the abnormal node set, retrieve the time series of its monitoring data and locate the time point in the series when the data begins to deviate from the normal range. This is recorded as the preliminary start time. Then, find the time point when the data returns to the normal range and remains stable without fluctuations. This is recorded as the preliminary end time.

[0044] These two time points are verified in combination with the operating mechanism of the incinerator. For example, if the data of a burner monitoring point is abnormal, according to the response characteristics of the combustion system, the temperature fluctuation caused by the change in fuel supply has a certain lag. If the initial start time is earlier than the time when the fuel supply system parameters change, the start time needs to be adjusted to the time when the fuel supply parameters change. In the above way, the initial start time and end time are corrected to obtain the accurate start time and end time of the interference. The time period between the two is the interference occurrence period corresponding to the abnormal node. For multiple abnormal nodes, if their interference occurrence periods overlap, the earliest start time and the latest end time are taken as the interference occurrence period of the entire dynamic interference event.

[0045] Step S114: Determine the spatial boundary of the interference impact range based on the distribution position of the abnormal node and the data fluctuation amplitude, and simultaneously determine the interference type based on the monitoring data type corresponding to the abnormal node.

[0046] Mark the spatial coordinates of the abnormal nodes in the 3D incinerator model and observe their distribution. If the abnormal nodes are concentrated in a certain area, such as near the feed inlet, expand the interference impact range outward from that area, taking into account the direction of airflow and heat transfer within the furnace. The spatial boundaries should be determined by referring to the incinerator's structural drawings to ensure that the boundaries do not cross obvious structural divisions (such as refractory brick partitions within the furnace).

[0047] Data fluctuations are measured by calculating the degree of deviation between the abnormal data and the mean of the normal data range. For each abnormal node, the extent to which its monitored data exceeds the upper limit of normal or falls below the lower limit of normal is calculated. The average fluctuation of multiple abnormal nodes is used as a reference for the overall impact of the interference.

[0048] Determine the interference type based on the monitoring data type corresponding to the abnormal node. If the abnormal nodes are mainly feed-related monitoring points, the interference type may be "feed anomaly", including feed quantity fluctuations, sudden changes in feed composition, etc.; if the abnormal nodes are concentrated in the burner area, the interference type may be "burner operation abnormality", such as unstable flame, insufficient fuel supply, etc.; if the flue gas monitoring point is abnormal, it may be "flue gas emission system abnormality".

[0049] Step S115: establishing dynamic interference event classification rules, and classifying dynamic interference events into different priority levels based on the number of nodes covered by the interference impact range, the duration of the interference period, and the data fluctuation amplitude of the abnormal nodes.

[0050] The dynamic interference event classification rules include three evaluation indicators: the number of covered nodes, duration, and data fluctuation range. Each indicator is divided into multiple levels. For example, the number of covered nodes is divided into "small," "medium," and "large," the duration is divided into "short," "medium," and "long," and the data fluctuation range is divided into "minor," "moderate," and "significant."

[0051] Each indicator level is assigned a corresponding score. The more nodes covered, the longer the duration, and the greater the fluctuation, the higher the score. The scores of the three indicators are added together to obtain a total score. Dynamic interference events are divided into three levels: "low priority," "medium priority," and "high priority" based on the range of the total score. For example, a total score in the lower range is low priority, indicating that the interference impact is minor and can be handled according to the normal process; a total score in the higher range is high priority, indicating that the interference may have a serious impact on the operation of the incinerator and requires immediate attention.

[0052] Step S116: Integrate the interference type, interference occurrence period, interference impact range, and interference priority level to form an event feature information set including priority information.

[0053] The interference type (e.g., "feed anomaly"), interference occurrence period (e.g., start time T1 to end time T2), spatial boundary coordinates of the interference impact range, and interference priority level (e.g., "medium priority") determined in the previous step are aggregated and organized into a structured data set, known as the event feature information set. This event feature information set is stored in a standardized data format and contains multiple fields, each corresponding to a piece of feature information, facilitating subsequent access and processing. For example, the value of the "interference type" field in the set is "feed anomaly - composition mutation," the value of the "interference occurrence period" field is [T1, T2], the value of the "spatial boundary" field is a polygonal boundary formed by a set of three-dimensional coordinate points, and the value of the "priority" field is "medium priority."

[0054] Step S120: Construct a spatial association graph of the incinerator furnace, with each spatial unit of the furnace as a graph node and the heat conduction relationship between the units as the graph edge weight. The event feature information is embedded in the corresponding graph node, and the association relationship between the interference feature and the temperature field anomaly is learned through the graph neural network, and the temperature anomaly association link corresponding to the interference event is output.

[0055] After obtaining a set of event feature information, it is necessary to construct a spatial association graph of the incinerator furnace to analyze the intrinsic connection between interference events and temperature field anomalies. First, the furnace is divided into multiple spatial units according to the set spatial division strategy, and each unit serves as a node in the association graph. Then, based on the principle of heat conduction, the heat conduction coefficient between units is calculated as the edge weight between nodes. Next, the event feature information is embedded into the corresponding nodes in the association graph. The relationship between these features and temperature field anomaly patterns is then learned through a graph neural network, ultimately determining the temperature anomaly association link affected by the interference.

[0056] Step S121: Divide the incinerator furnace into multiple space units, use each space unit as a basic node of the furnace space association diagram, and record the spatial coordinates and geometric properties of each basic node.

[0057] Based on the waste incinerator's furnace structure, a structured meshing method was used to divide the furnace's interior into multiple spatial units. In areas with drastic heat flux variations, such as near the burner and in the feed inlet, the spatial unit size was kept small to improve analysis accuracy. In areas with gentler heat flux variations, such as the furnace top and corners, the spatial unit size was increased to reduce computational complexity.

[0058] Each spatial unit serves as a foundational node in the furnace spatial association diagram, recording its spatial coordinates (such as the 3D coordinates of the unit's center point) and geometric attributes (such as the unit's length, width, and height, volume, and surface area). For example, the spatial unit located directly in front of burner 1 has its center point coordinates (x3, y3, z3), geometric attributes including length a, width b, and height c, a volume of a×b×c, and a surface area calculated as the sum of the areas of all its faces. This information is stored in the basic node's attribute database and serves as the basis for subsequent calculations.

[0059] Step S122: Based on the heat conduction mechanism of the incinerator, the heat conduction coefficient between adjacent space units is calculated, and the heat conduction coefficient is used as the edge weight between the corresponding basic nodes in the furnace space association graph to construct the initial furnace space association graph.

[0060] Adjacent spatial units are those that share a common surface or edge. According to the heat conduction mechanism, the heat conduction coefficient between two adjacent spatial units is related to the contact area between the units, the thermal conductivity of the material, and the distance between the units. The calculation first determines the shared contact area between the two adjacent units. Then, based on the material properties of this area within the furnace (such as the thermal conductivity of the refractory material) and the distance between the unit centers, the heat conduction coefficient is calculated using the textual description logic of the heat conduction formula.

[0061] The calculated heat transfer coefficient is used as the graph edge weight between two adjacent basic nodes. A larger weight indicates a stronger heat transfer capability between the two spatial units. Following this approach, weighted edges are established for all adjacent basic nodes, forming an initial furnace spatial association graph. This initial furnace spatial association graph reflects the path and intensity of heat transfer between different spatial units within the furnace.

[0062] Step S123: Determine the node features corresponding to each abnormal temperature field mode to form an abnormal mode node feature library.

[0063] By analyzing cases of temperature field anomalies that occurred during the historical operation of waste incinerators, we identified common temperature field anomaly patterns, such as localized high-temperature zones, abnormal temperature gradients, and frequent temperature fluctuations. For each anomaly pattern, we extracted the characteristics of the corresponding nodes in the furnace spatial association graph. For example, the node characteristics corresponding to a localized high-temperature zone include the degree to which the node's temperature exceeds the normal range, the duration of the high temperature, and the temperature difference with surrounding nodes. Node characteristics corresponding to an abnormal temperature gradient include the temperature change rate between adjacent nodes and the distribution range of the abnormal gradient.

[0064] These features are categorized and stored by anomaly pattern, forming an anomaly pattern node feature library. Each anomaly pattern in the library corresponds to a feature set containing multiple feature items. Each feature item has a clear definition and description, facilitating comparison with subsequent node features extracted in real time.

[0065] Step S124: The event feature information is used as an interference feature vector and embedded into the corresponding basic node in the initial furnace space association graph. The furnace space association graph embedded with the interference feature vector is trained using a graph convolutional neural network to learn the mapping relationship between the interference feature vector and the features in the abnormal pattern node feature library, thereby forming an interference temperature association learning model.

[0066] The event feature information set is converted into a numerical interference feature vector, where the dimension of the vector is consistent with the number of feature items. For example, the interference type is converted into a numerical value through coding, the interference occurrence period is converted into a numerical value of duration, the interference impact range is converted into a numerical value of the number of spatial units covered, and the interference priority level is also converted into a corresponding numerical value.

[0067] Based on the spatial boundaries of the interference impact range, the basic nodes related to the interference in the initial furnace spatial association graph are determined, and the interference feature vectors are embedded into the attributes of these nodes. If the interference impact range involves multiple nodes, the same interference feature vector is embedded in each related node.

[0068] A graph convolutional neural network is used to train the furnace space association graph embedded with interference feature vectors to learn the mapping relationship between interference features and temperature field abnormal pattern features, forming an interference-temperature association learning model.

[0069] Step S1241: constructing a training data set, wherein the training data set includes interference feature vectors of historical interference events, corresponding furnace space association graphs, and a set of manually labeled temperature anomaly nodes.

[0070] Cases of various dynamic interference events that occurred during the past operation of the incinerator were collected. Each case included the interference feature vector at that time (obtained through historical data extraction), the corresponding furnace space association diagram (constructed according to the furnace structure and heat conduction characteristics at that time), and a set of manually labeled temperature anomaly nodes (determined by technical personnel based on the temperature monitoring data and operation records at that time).

[0071] These cases were organized into a training dataset, with each sample consisting of three components: an interference feature vector, a furnace spatial correlation map, and a set of annotations for temperature anomaly nodes. The training dataset was then divided into training and validation subsets based on a set ratio for model training and performance verification.

[0072] Step S1242: Input the furnace space association graph in the training data set into the graph convolutional neural network, aggregate the node features through the graph convolution layer, and generate a node embedding vector. The process of aggregating the node features through the graph convolution layer combines the weighted sum of the node's own features and the features of adjacent nodes.

[0073] The graph convolutional neural network consists of multiple graph convolutional layers. The furnace spatial association graph from the training dataset is fed into the network's first graph convolutional layer. The features of each node (including basic attribute features and embedded interference feature vectors) are aggregated with the features of neighboring nodes. During aggregation, the features of neighboring nodes are weighted by the graph edge weight (heat transfer coefficient). The features of neighboring nodes with larger weights contribute more to the aggregated data.

[0074] The aggregated features are transformed nonlinearly using an activation function to produce the node features output by this layer. These features are then fed into the next graph convolutional layer for deeper aggregation and transformation. After processing through multiple graph convolutional layers, the features of each node are compressed into a low-dimensional node embedding vector that contains both the node's own features and its associations with surrounding nodes.

[0075] Step S1243: In the fully connected layer of the graph convolutional neural network, the node embedding vector is compared with the features in the abnormal pattern node feature library, the similarity loss between the node embedding vector and the features in the abnormal pattern node feature library is calculated, and the network parameters of the graph convolutional neural network are optimized using the gradient descent algorithm.

[0076] The fully connected layer of the graph convolutional neural network receives the node embedding vector output by the graph convolutional layer and compares it with various features in the abnormal pattern node feature library. This comparison is achieved by calculating the similarity between the node embedding vector and each feature in the feature library. The higher the similarity, the more likely the node belongs to the corresponding temperature anomaly pattern.

[0077] The loss between the similarity of the actual output and the manually labeled set of temperature anomaly nodes, known as the similarity loss, is calculated. Using a gradient descent algorithm, the parameters of each network layer (such as the weight matrix and bias terms) are adjusted based on the loss value, gradually reducing the loss. Through multiple iterative training cycles, the network parameters are continuously optimized, improving the model's accuracy in identifying temperature anomaly nodes.

[0078] Step S1244: Introduce the attention mechanism layer to assign attention weights to the node embedding vectors output by the graph convolution layer, so that the graph convolutional neural network pays more attention to node features related to temperature anomalies.

[0079] An attention mechanism layer is introduced between the graph convolutional layer and the fully connected layer. This attention mechanism layer assigns an attention weight to each node by calculating the correlation between each node's embedding vector and the temperature anomaly pattern features. Nodes with higher correlation receive larger attention weights, and are given greater importance in subsequent feature processing.

[0080] Specifically, the attention mechanism layer processes the node embedding vectors output by the graph convolutional layer, calculates the attention score for each node through a small neural network, and then converts the score into an attention weight using the softmax function. After multiplying the node embedding vector by the corresponding attention weight, the vector is input into the fully connected layer for processing. This approach automatically focuses on node features that are more critical for identifying temperature anomalies, improving model performance.

[0081] Step S1245: Divide the training data set into a training subset and a validation subset, use the training subset to train the network parameters of the graph convolutional neural network, and use the validation subset to verify the matching degree between the temperature anomaly node set output by the graph convolutional neural network and the manually labeled temperature anomaly node set.

[0082] The training dataset is divided into a training subset and a validation subset in proportion, for example, the majority of the data is used as the training subset and a smaller portion as the validation subset. The graph convolutional neural network is trained using the training subset, and the validation subset is used to validate the trained model after each training iteration.

[0083] During the validation process, the furnace spatial correlation graph from the validation subset was input into the model. The resulting set of temperature anomaly nodes was compared with the manually labeled set, and the matching degree was calculated. The matching degree was calculated as the number of intersections of the model-identified anomaly nodes and the manually labeled anomaly nodes, divided by the number of unions of the two. A higher matching degree indicates better generalization of the model.

[0084] Step S1246: If the matching degree is lower than the preset matching degree standard, adjust the number of layers, convolution kernel size and attention weight calculation method of the graph convolutional neural network, and retrain the graph convolutional neural network.

[0085] The preset matching standard is determined based on actual application requirements, for example, it can be set to a higher ratio. If the matching degree of the validation subset is lower than this standard, it means that the model performance does not meet the requirements and the network structure parameters need to be adjusted.

[0086] Adjustments include increasing or decreasing the number of graph convolutional layers, changing the size of the convolution kernel (i.e., the range of adjacent nodes considered during each aggregation), and modifying the calculation function of the attention weights in the attention mechanism layer (e.g., using a different relevance metric). After these adjustments, the model is retrained using the training subset and revalidated using the validation subset until the match reaches the preset standard.

[0087] Step S1247: If the matching degree reaches the preset matching degree standard, stop training to form a stable interference temperature association learning model.

[0088] When the matching degree of the validation subset reaches the preset standard, the model demonstrates good recognition and generalization capabilities. Training is then stopped, the current network parameters are saved, and a stable interference temperature association learning model is formed. This interference temperature association learning model receives a furnace spatial association graph embedded with interference feature vectors and outputs prediction results for the corresponding set of temperature anomaly nodes.

[0089] Step S125: Iteratively optimize the interference temperature association learning model through historical interference temperature data, so that the interference temperature association learning model can output a set of temperature anomaly nodes corresponding to the interference event, and determine the furnace function link corresponding to the temperature anomaly node set as the temperature anomaly association link; wherein, when the interference temperature association learning model outputs the temperature anomaly association link, the association confidence is output synchronously, and the association confidence is obtained by calculating the similarity of the graph node embedding. If the association confidence is lower than the preset association confidence threshold, the feature embedding and model training are re-executed.

[0090] Collect more historical interference events and their corresponding temperature field data to form new training samples, and regularly iterate and optimize the interference-temperature correlation learning model. The optimization process is similar to the initial training process, inputting new samples into the model and adjusting network parameters to further improve the model's accuracy.

[0091] The model calculates the association confidence level while outputting a set of temperature anomaly nodes. This confidence level is calculated by calculating the average similarity between the node embedding vectors output by the model and the corresponding features in the anomaly pattern node feature library. The higher the similarity, the greater the association confidence level, indicating a higher reliability of the model's prediction results.

[0092] Set a preset association confidence threshold. If the output association confidence is lower than the threshold, it means that the model's recognition reliability of the current interference event is insufficient. It is necessary to recheck the embedding process of the event feature information. It may be that the embedded node range is inaccurate or there is a deviation in the feature vector conversion. After adjustment, retrain the model and predict until the association confidence reaches above the threshold.

[0093] The furnace functional link corresponding to the set of temperature anomaly nodes output by the model is identified as the temperature anomaly association link. For example, if the temperature anomaly nodes are concentrated in the spatial unit near burner 2, the corresponding temperature anomaly association link is "burner 2 working area."

[0094] Step S130: inputting the temperature anomaly associated link into the CFD simulation system to generate a target simulation task for the associated link, wherein the target simulation task includes a target simulation area, simulation duration, and key monitoring temperature indicators.

[0095] The identified temperature anomaly-related link is input into a CFD (computational fluid dynamics) simulation system. The system generates a targeted simulation task based on the link's characteristics and relevant parameters. The target simulation task requires a clear spatial scope, time span, and key temperature indicators to be monitored to ensure that the simulation results accurately reflect the temperature field changes in the associated link under the influence of the interference event.

[0096] Step S131: calling the CFD simulation system to receive the temperature anomaly association link and the corresponding association confidence. If the association confidence is higher than a preset association confidence threshold, starting the target simulation task generation process.

[0097] The CFD simulation system receives information about temperature anomaly associations and their corresponding association confidence levels through an interface. The CFD simulation system internally sets a confidence threshold, determined based on the required accuracy of the simulation results. If the received confidence level exceeds the threshold, the identification of the temperature anomaly association is reliable, and the target simulation task generation process is initiated. If the confidence level is below the threshold, a prompt message is returned, requesting the re-determination of the temperature anomaly associations.

[0098] Step S132: Use the graph neural network to sort the importance of the nodes corresponding to the temperature anomaly association links in the furnace space association graph, extract the top K nodes and their adjacent nodes, and determine the spatial range of the target simulation area. The spatial range of the target simulation area includes the spatial units corresponding to the potential nodes where interference may spread.

[0099] The nodes in the furnace spatial correlation graph corresponding to temperature anomalies are fed into the graph neural network. The neural network ranks these nodes by importance by analyzing factors such as the connection strength and feature importance of the nodes in the correlation graph. The ranking is based on factors such as the node's correlation with the interference feature vector and its position in the heat conduction path. Nodes with higher importance require more sophisticated processing in the simulation.

[0100] Extract the top-ranked nodes and consider their neighboring nodes to capture potential areas where interference may spread. Combine the spatial cells corresponding to these nodes to form the spatial extent of the target simulation region. For example, if the top-ranked nodes are concentrated around burner 3, the target simulation region will include burner 3 and the spatial cells within the specified range around it, ensuring that the interference diffusion process in this area is captured.

[0101] Step S133: according to the interference occurrence period and interference duration characteristics in the event characteristic information, the time span of the simulation duration is set, and the time span includes the reference stable period before the interference occurs, the interference occurrence period and the recovery period after the interference is eliminated.

[0102] The simulation duration is determined by referring to the interference occurrence period (start time T1 and end time T2) in the event characteristic information and combining it with the duration characteristics of the interference (such as whether the interference is quickly eliminated after it occurs or persists for a long time). The time span consists of three parts: the baseline stabilization period before the interference occurs (a period extending from T1 onwards), which is used to obtain baseline temperature field data under normal operating conditions; the interference occurrence period (from T1 to T2), which simulates the temperature field changes during the interference event; and the recovery period after the interference is eliminated (a period extending from T2 onwards), which is used to observe whether the temperature field can return to normal.

[0103] For example, if the interference occurs for 10 minutes, the baseline stabilization period can be set to 5 minutes before the interference occurs, and the recovery period can be set to 5 minutes after the interference ends. The total simulation time is 20 minutes to fully capture the impact of the interference on the temperature field.

[0104] Step S134: Based on the node characteristics of the temperature field abnormality pattern, determine the key monitoring temperature indicators, which include the core temperature parameters corresponding to the temperature field abnormality pattern and the associated derived temperature parameters. At the same time, set the monitoring frequency of different simulation time periods, and the monitoring frequency during the interference period is higher than that in other simulation time periods.

[0105] Key temperature monitoring indicators are determined based on the node characteristics of the abnormal temperature field pattern. Core temperature parameters directly reflect temperature anomalies, such as the temperature value of the abnormal node and the temperature difference between adjacent nodes. Derived temperature parameters indirectly reflect temperature field stability, such as the temperature change rate and the frequency of temperature fluctuation. For example, for a localized high temperature abnormality pattern, the core temperature parameter is the temperature value of the high temperature zone, and the derived temperature parameter is the rate of temperature increase.

[0106] The monitoring frequency is set for different simulation periods. During the interference period, the temperature field changes dramatically, so the monitoring frequency is set to a higher level to capture subtle temperature changes. During the baseline stabilization period and recovery period, the temperature field changes relatively slowly, and the monitoring frequency can be appropriately reduced to reduce the amount of data.

[0107] Step S135: Allocate CFD simulation resources based on the interference priority level to generate a target simulation task for resource optimization.

[0108] The CFD simulation system allocates computing resources based on the priority level of dynamic interference events. High-priority interference events are allocated more computing cores and memory resources, and a finer computational grid is used to improve simulation accuracy and speed. Low-priority interference events can be allocated fewer resources and a simpler computational setup can be used.

[0109] The target simulation area, simulation duration, key monitoring temperature indicators and resource allocation information are integrated to generate a complete target simulation task. The target simulation task is stored in the form of a structured task description file, which contains all necessary parameter settings for execution by the CFD simulation system.

[0110] Step S140: Execute the target simulation task, obtain the temperature field dynamic response data of the associated link, analyze the graph node embedding changes corresponding to the temperature field dynamic response data through the graph neural network, determine the correlation strength between the temperature field response and the interference event, and then determine the simulation correction direction that needs to be adjusted.

[0111] The CFD simulation system performs simulations according to the target simulation task settings, acquiring dynamic temperature field response data for temperature anomaly-related links at different times. A graph neural network analyzes the changes in graph node embedding caused by this data, assessing the correlation between the temperature field response and the disturbance event. Based on the evaluation results, the simulation corrections required are determined to optimize the temperature field simulation results.

[0112] Step S141: calling the CFD simulation system to load the target simulation area geometry model corresponding to the target simulation task, importing the initial operation parameters, and starting the simulation operation according to the set time step.

[0113] The CFD simulation system reads the target simulation area geometry model file specified in the target simulation task. This target simulation area geometry model file contains geometric information such as the furnace structure and equipment layout within the area. It also imports the incinerator's initial operating parameters, including the initial temperature distribution, the initial flow rate and pressure of the fluids (air and flue gas), and the initial fuel supply rate. These parameters are set based on historical data from the incinerator's normal operation.

[0114] Set the simulation time step. This size is determined by the required simulation accuracy and available computing resources. During periods of interference, the time step can be set to a smaller value to improve the ability to capture rapidly changing processes. After starting the simulation, the system numerically calculates the temperature and flow fields within each time step according to the basic equations of fluid mechanics and heat conduction, gradually advancing the simulation.

[0115] Step S142: During the simulation operation, the temperature data corresponding to each node in the target simulation area is collected according to the monitoring frequency to form a temperature field dynamic response data sequence, and the interference simulation state of each time node is synchronously recorded.

[0116] During simulations, the CFD system collects temperature data for all spatial nodes within the target simulation area at each monitoring moment, based on a preset monitoring frequency. This data includes the instantaneous temperature value and temperature gradient at each node. This data is arranged in chronological order to form a dynamic temperature response data sequence, reflecting how the temperature field changes over time.

[0117] At the same time, the system records the interference simulation status at each time point, such as whether interference is occurring and the intensity level of the interference, so as to facilitate subsequent analysis of the corresponding relationship between the temperature field response and the interference status. For example, at the time point of the interference period, the specific manifestation of the interference (such as a sudden increase in feed volume) is recorded, while during the baseline stabilization period and recovery period, it is recorded as no interference or the interference is eliminated.

[0118] Step S143: Embed the temperature field dynamic response data sequence into the corresponding node of the furnace space association graph to generate a dynamic node embedding vector. Calculate the change in the dynamic node embedding vector before and after the interference occurs through the node embedding update algorithm of the graph neural network.

[0119] The temperature data for each time point in the temperature field dynamic response data sequence is embedded into the corresponding basic nodes of the furnace spatial association graph according to the node correspondence relationship, and the node feature information is updated. The node embedding update algorithm of the graph neural network is used to process the association graph after embedding the new temperature data, generating dynamic node embedding vectors for each node at different time points.

[0120] This update algorithm, similar to the encoding process of a graph autoencoder, aggregates the node's own temperature features with those of neighboring nodes through graph convolution operations to generate a new embedding vector. The difference in the dynamic node embedding vector for the same node before the disturbance (baseline stability period) and after the disturbance (disturbance period) is calculated to obtain the change. A larger change indicates that the node's temperature features are more significantly affected by the disturbance.

[0121] Step S144: Construct an association strength analysis dimension, which includes a time dimension, a space dimension, and a feature dimension. The time dimension analyzes the synchronization between the occurrence of interference and the change of the dynamic node embedding vector. The space dimension analyzes the overlap between abnormal nodes and interference-affected nodes. The feature dimension analyzes the matching degree between the change of the dynamic node embedding vector and the interference feature.

[0122] Correlation strength analysis involves three dimensions: time, space, and features. In the time dimension, we analyze the synchronization between the time of interference and the time when the dynamic node embedding vector begins to change significantly. If the time of vector change coincides with or is close to the time of interference, the correlation in the time dimension is high.

[0123] In the spatial dimension, the number of overlaps between abnormal nodes in the temperature anomaly association link and nodes whose embedded vectors have changed due to interference is counted. The greater the number of overlaps, the higher the degree of association in the spatial dimension.

[0124] In the feature dimension, the matching degree between the change characteristics (e.g., magnitude and trend) of the dynamic node embedding vector and the interference feature vector (e.g., interference type and intensity) is analyzed. For example, if the change in the corresponding node embedding vector for the interference feature of an increase in feed rate shows an increasing temperature trend, then the matching degree in the feature dimension is high.

[0125] Step S145: Calculate the association scores of each association strength analysis dimension through the graph neural network, standardize the association scores of each association strength analysis dimension, and sum them up according to the preset weights to obtain the comprehensive association strength. If the comprehensive association strength is lower than the preset comprehensive association strength threshold, recheck the parameter settings of the target simulation task.

[0126] The graph neural network calculates a correlation score for each correlation strength analysis dimension. The score for the time dimension is determined by the degree of synchronization, with better synchronization resulting in a higher score. The score for the spatial dimension is calculated based on the percentage of overlap, with a higher percentage resulting in a higher score. The score for the feature dimension is based on the degree of matching, with a higher degree of matching resulting in a higher score.

[0127] Normalize the correlation scores of the three dimensions to the same numerical range (e.g., 0 to 1). Set preset weights based on the importance of each dimension, such as giving higher weights to the time and space dimensions and slightly lower weights to the feature dimension. The standardized scores are weighted and summed to obtain the overall correlation strength.

[0128] The preset comprehensive correlation strength threshold is determined based on historical data statistics. If the calculated comprehensive correlation strength is lower than the threshold, it means that the correlation between the temperature field response and the interference event is not significant enough. It may be that the parameter settings of the target simulation task are unreasonable (such as the simulation area is too small, the time step is too large, etc.). It is necessary to recheck and adjust the parameter settings and re-execute the simulation task.

[0129] Step S146: If the comprehensive correlation strength is higher than the preset comprehensive correlation strength threshold, based on the temperature data characteristics corresponding to the dynamic node embedding vector change, combined with the incinerator temperature field optimization target, determine the simulation correction direction, which includes the heat source adjustment direction, the fluid flow adjustment direction and the heat exchange adjustment direction.

[0130] When the combined correlation strength exceeds a preset threshold, it indicates a significant correlation between the temperature field response and the disturbance event, which can be used to determine the direction of simulation correction. The temperature data characteristics corresponding to the dynamic node embedding vector change include the distribution of temperature anomaly nodes, the magnitude of temperature change, and the temperature gradient. Incinerator temperature field optimization objectives include temperature field uniformity (temperature differences between regions are within a reasonable range), temperature stability (minimal temperature fluctuations), and temperature range conformity (temperatures are within a set reasonable range).

[0131] Determine the specific direction of simulation correction based on the difference between the temperature data characteristics and the optimization target. If the temperature data shows that the temperature in a local area exceeds the reasonable range, adjust the heat source; if the temperature gradient is too large, optimize the fluid flow; if the temperature changes too quickly, adjust the heat exchange components.

[0132] Step S1461: extracting temperature data features corresponding to the dynamic node embedding vector variation, wherein the temperature data features include the location of the temperature abnormality node, the temperature variation amplitude, the temperature fluctuation frequency, and the temperature gradient distribution between nodes.

[0133] Corresponding temperature data features are extracted from the dynamic node embedding vector changes. The location of the temperature anomaly node is determined by the node's spatial coordinates; the temperature variation amplitude is the difference between the abnormal node temperature and the normal range; the temperature fluctuation frequency is the number of times the temperature exceeds the normal range per unit time; and the inter-node temperature gradient distribution is the spatial distribution of the temperature difference between adjacent nodes. For example, if the temperature variation amplitude of nodes in a certain area is large, the fluctuation frequency is high, and the temperature gradient with surrounding nodes is obvious, these are all temperature data features that require attention.

[0134] Step S1462: Obtain the incinerator temperature field optimization target, which includes a temperature field uniformity target, a temperature stability target, and a temperature range compliance target.

[0135] The incinerator's temperature field optimization targets are set by technical personnel based on the process requirements for medical waste treatment. The temperature field uniformity target requires that temperature differences between various areas within the furnace be controlled within a set range to avoid localized overcooling or overheating. The temperature stability target requires that the amplitude and frequency of temperature fluctuations be within allowable limits to ensure a stable combustion process. The temperature range compliance target requires that the furnace temperature be maintained within a range suitable for the complete combustion of medical waste without generating excessive amounts of harmful pollutants.

[0136] Step S1463: If the temperature data characteristics show that the local node temperature is higher than the temperature range corresponding to the temperature interval compliance target, and the correlation strength analysis shows that the embedding change of the local node is significantly correlated with the embedding change of the heat source input node, then the simulation correction direction is determined to be the heat source adjustment direction, specifically, reducing the heat source input intensity corresponding to the heat source input node.

[0137] When temperature data characteristics indicate that the temperature of a local node consistently exceeds the upper limit of the temperature range compliance target, and the correlation strength analysis shows significant synchronization and high matching between the dynamic embedding vector changes of this local node and the embedding vector changes of the heat source input node (such as the node corresponding to the burner), it indicates that the local high temperature is caused by excessive heat input. In this case, the simulation correction direction is determined to be heat source regulation. The specific measure is to reduce the fuel supply or combustion intensity corresponding to the heat source input node to reduce heat input and bring the local temperature down to a reasonable range.

[0138] Step S1464: If the temperature data characteristics show that the temperature gradient between nodes exceeds the temperature gradient range corresponding to the temperature field uniformity target, and the correlation strength analysis shows that the temperature gradient change is strongly correlated with the node embedding change corresponding to the fluid flow path, then the simulation correction direction is determined to be the fluid flow regulation direction, specifically, the flow velocity parameters of the node area corresponding to the fluid flow path are optimized.

[0139] When the temperature gradient between nodes exceeds the temperature field uniformity target range, and correlation strength analysis shows a strong correlation between the temperature gradient change and the node embedding vector changes along the fluid (e.g., combustion air, flue gas) flow path (e.g., the temperature gradient changes synchronously with changes in flow velocity), this indicates that the uneven temperature distribution is caused by poor fluid flow. In this case, the simulation correction direction is determined to be fluid flow regulation. Specific measures include adjusting the flow velocity parameters of the relevant node areas along the fluid flow path, such as increasing or decreasing the air volume in certain areas, to improve fluid mixing and heat transfer, and reduce the temperature gradient.

[0140] Step S1465: If the temperature data characteristics show that the node temperature change rate exceeds the temperature change rate range corresponding to the temperature stability target, and the correlation strength analysis shows that the temperature change rate is strongly correlated with the embedded change of the heat exchange node, then the simulation correction direction is determined to be the heat exchange adjustment direction, specifically, adjusting the heat exchange component parameters corresponding to the heat exchange node.

[0141] When the rate of change of node temperature exceeds the limit set for the temperature stability objective, and correlation strength analysis shows that this rapid change is strongly correlated with changes in the embedding vector of heat exchange nodes (such as the nodes corresponding to the cooling water pipes on the furnace wall), it indicates that the temperature instability is caused by abnormal heat exchange efficiency. In this case, the simulation correction direction is determined to be heat exchange adjustment. Specific measures include adjusting the parameters of the heat exchange components corresponding to the heat exchange nodes, such as changing the cooling water flow rate or flow rate, adjusting the heat exchange efficiency, reducing the temperature change rate, and stabilizing the temperature field.

[0142] Step S1466: Verify the feasibility of the determined simulation correction direction, and combine the parameter adjustment range of the CFD simulation system and the actual operation limitations of the incinerator to ensure that the simulation correction direction is within the feasible range. If it is not feasible, re-analyze the temperature data characteristics corresponding to the dynamic node embedded vector change and adjust the simulation correction direction.

[0143] Verify the feasibility of the initially determined simulation correction direction. Check the adjustment range of the corresponding parameters in the CFD simulation system to ensure that the parameter changes required by the correction direction are within the system's allowable range. At the same time, consider the actual operating limitations of the incinerator, such as the maximum and minimum output power of the burner and the speed range of the fan, to ensure that the correction measures can be implemented in actual operation.

[0144] If the simulated correction direction exceeds the system parameter adjustment range or actual operating limits, it is determined to be infeasible and the temperature data characteristics need to be reanalyzed to find other possible correction directions. For example, if reducing the heat source input intensity causes the combustion temperature to fall below the minimum requirement for medical waste treatment, this direction needs to be abandoned and the local high temperature needs to be reanalyzed to determine if other factors are causing the local high temperature and adjust the correction direction accordingly.

[0145] Step S150: Generate a CFD simulation parameter adjustment instruction according to the simulation correction direction, input the CFD simulation parameter adjustment instruction into the CFD simulation system to update simulation parameters, re-execute the target simulation task, and obtain the corrected temperature field response data.

[0146] Based on the determined simulation correction direction, specific CFD simulation parameter adjustment instructions are generated, clearly defining the parameters to be adjusted, the direction of adjustment, and the magnitude of the adjustment. These instructions are input into the CFD simulation system, which updates the relevant simulation parameters. The simulation is then re-executed according to the original target simulation task settings to obtain the corrected temperature field response data to observe the improvement effect of the correction measures on the temperature field.

[0147] Step S151: Establish a mapping relationship library between simulation correction directions and CFD parameters. The mapping relationship library records the core parameter types and parameter adjustment constraints corresponding to different simulation correction directions. The constraints are classified into physical constraints and numerical constraints. Physical constraints include the equipment safe operation parameter range, and numerical constraints include simulation stability requirements.

[0148] A mapping library between simulation correction directions and CFD parameters was established by analyzing the incinerator's physical model and the numerical model used in CFD simulations. For heat source adjustment, core parameters include the burner's fuel flow rate and combustion temperature setpoint. Fluid flow adjustment corresponds to parameters such as fan air volume, wind speed, and fluid inlet angle. Heat exchange adjustment corresponds to parameters such as cooling medium flow rate and temperature.

[0149] The mapping library also records the constraints for parameter adjustments. Physical constraints are based on the safe operation requirements of the incinerator equipment, such as the burner fuel flow rate cannot exceed its maximum design value and the cooling medium pressure cannot fall below the safety threshold. Numerical constraints are based on the stability requirements of CFD simulations, such as flow rate changes cannot be too drastic to cause numerical calculation divergence, and temperature gradient adjustments must ensure the numerical stability of the energy conservation equation.

[0150] Step S152: Based on the determined simulation correction direction, the corresponding core parameter type is extracted from the mapping relationship library, the current value of the core parameter type in the CFD simulation system is obtained, and the reasonable range of the core parameter type adjustment is determined in combination with the dynamic node embedding vector change corresponding to the temperature field dynamic response data.

[0151] Based on the determined simulation correction direction (e.g., heat source adjustment direction), the corresponding core parameter type (e.g., burner fuel flow) is searched from the mapping relationship library. The current value of the core parameter is obtained through the parameter query interface of the CFD simulation system.

[0152] Analyze the changes in the dynamic node embedding vectors in the temperature field's dynamic response data. Larger changes indicate potentially larger adjustments are needed. Consider the parameter constraints to determine a reasonable range for adjusting core parameters. For example, if the current fuel flow rate is a certain value and the vector change indicates a significant temperature reduction is needed, determine the range within which the fuel flow rate should be adjusted downward, within the limits permitted by the physical constraints. This ensures that the adjusted parameters effectively improve the temperature field without violating the constraints.

[0153] Step S153: Generate a CFD simulation parameter adjustment instruction, which includes a target parameter identifier, an adjustment direction, an adjustment range, and a verification rule for the adjusted parameter value, so that the adjusted parameter value meets the constraint condition.

[0154] CFD simulation parameter adjustment instructions are generated using a standardized format consisting of multiple fields. The target parameter identifies the specific parameter to be adjusted (e.g., "burner 1 fuel flow"). The adjustment direction indicates whether the parameter is to be increased or decreased. The adjustment magnitude is determined within a reasonable range and can be expressed as a percentage or absolute change relative to the current value. Validation rules for the adjusted parameter values ​​are used to check whether the parameter adjustment complies with the constraints, such as verifying that the adjusted fuel flow is within the range of physical constraints.

[0155] For example, an adjustment instruction may be: target parameter identifier "burner 1 fuel flow", adjustment direction "decrease", adjustment range "10%", and verification rule "the adjusted value must be greater than the minimum fuel flow of the burner and less than the current value".

[0156] Step S154: importing CFD simulation parameter adjustment instructions through the parameter update interface of the CFD simulation system, automatically replacing the original parameter values, and backing up the parameter configuration before adjustment.

[0157] The CFD simulation system provides a dedicated parameter update interface to receive generated parameter adjustment commands. The system parses the command content, locates the corresponding parameter based on the target parameter identifier, and automatically updates the parameter value according to the adjustment direction and magnitude. During the update process, the system automatically executes validation rules to check whether the adjusted parameters meet the constraints. If not, the update is rejected and an error message is displayed.

[0158] After the parameter update is complete, the system automatically backs up the parameter configuration before the adjustment and stores it in a specified backup file so that it can be restored to the state before the adjustment when needed. The backup file contains information such as the original values ​​of all parameters and the adjustment time.

[0159] Step S155: Call the basic configuration of the target simulation task, keep the target simulation area, simulation duration and monitoring frequency unchanged, start the re-simulation operation, collect temperature data at the set frequency, form the corrected temperature field response data, and synchronously record the corrected embedding vector of each node.

[0160] Call the basic configuration of the target simulation task from the task storage of the CFD simulation system to ensure that the spatial range of the target simulation area, the time span of the simulation duration, and the monitoring frequency of each time period are consistent with the previous simulation task to ensure that the simulation results before and after the correction are comparable.

[0161] For example, step S1551: retrieve the basic configuration of the target simulation task from the task management process of the CFD simulation system, confirm the spatial coordinate range of the target simulation area, the time node of the simulation duration, and the monitoring frequency of each simulation period.

[0162] In the task management module of the CFD simulation system, search and retrieve the corresponding target simulation task basic configuration file. This target simulation task basic configuration file records in detail the spatial coordinate range of the target simulation area (such as the boundary defined by multiple three-dimensional coordinate points), the time nodes of the simulation duration (the start and end time of the baseline stability period, the start and end time of the interference occurrence period, and the start and end time of the recovery period), and the monitoring frequency of each simulation period (such as once every longer interval during the baseline stability period and once every shorter interval during the interference occurrence period).

[0163] Confirm the retrieved configuration information to ensure that it is consistent with the original simulation task. If there is any deviation, correct it to ensure that the conditions of the re-simulation are the same as the original simulation conditions.

[0164] Step S1552: Check the computing resource status of the CFD simulation system to determine whether the current computing resources meet the requirements for re-simulation calculations. If the computing resources do not meet the requirements for re-simulation calculations, queue up according to the interference priority level, or adjust the resource allocation of non-critical simulation tasks.

[0165] Use the CFD simulation system's resource monitoring module to view currently available computing resources, including the number of CPU cores, memory capacity, and GPU utilization. Compare these resources with the resources required for the re-simulation to determine whether they are sufficient. The re-simulation resource requirements are determined by factors such as the size of the target simulation area, simulation duration, and mesh fineness.

[0166] If the computing resources meet the requirements, the re-simulation operation is initiated directly. If not, the dynamic interference event is handled according to its priority level. High-priority events can enter the queue to wait for resource release, or the currently executing non-critical simulation task can be suspended to allocate its resources to the re-simulation operation. Low-priority events enter the queue and wait for other tasks to complete before obtaining resources.

[0167] Step S1553: Load the adjusted simulation parameters, confirm that the adjusted simulation parameter values ​​meet the constraints, start re-simulation operation, and monitor the operation status in real time during the re-simulation operation.

[0168] The updated simulation parameters are loaded into the computational module of the CFD simulation system. The system then verifies that the adjusted parameter values ​​meet the physical and numerical constraints. Once verified, the simulation is restarted.

[0169] During the simulation process, monitor the status of the operation in real time, including the calculation progress, the convergence of each time step, and whether there is any numerical instability (such as abnormal jumps in temperature or pressure). If the operation is interrupted or convergence fails, stop the operation immediately, check the parameter settings and model configuration, and restart after troubleshooting the problem.

[0170] Step S1554: Collect temperature data of each node in the target simulation area according to the set monitoring frequency, record the temperature value, temperature change rate and temperature gradient data between nodes at each time node, and form corrected temperature field response data.

[0171] During the resimulation operation, the system collects temperature data for all spatial nodes within the target simulation area at each monitoring time point according to the preset monitoring frequency. The collected data includes the instantaneous temperature value of each node, the temperature change rate per unit time (the ratio of the temperature difference to the time interval), and the temperature gradient between adjacent nodes (the ratio of the temperature difference to the spatial distance).

[0172] These data are sorted in chronological order to form the corrected temperature field response data, which uses the same data format as the temperature field response data before correction to facilitate subsequent comparative analysis.

[0173] Step S1555: Embed the corrected temperature field response data into the corresponding node of the furnace space association graph, generate a corrected node embedding vector through the node embedding algorithm of the graph neural network, and mark and associate it with the node embedding vector before correction.

[0174] The corrected temperature field response data is embedded into the basic nodes of the furnace space association graph according to the node correspondence relationship, and the same graph neural network node embedding algorithm as before (such as the encoding process based on graph convolution) is used to generate the corrected node embedding vector of each node at different time points.

[0175] The corrected node embedding vector is labeled to indicate the corresponding parameter adjustment (e.g., “burner 1 fuel flow reduced by 10 percent”) and associated with the uncorrected node embedding vector to ensure that the two can be compared one-to-one by node and time point.

[0176] Step S1556: After the re-simulation operation is completed, the corrected temperature field response data and the corrected node embedding vector are classified and stored, and a simulation process report is generated to record the parameter adjustment status, operation status and data acquisition status.

[0177] After the re-simulation operation is completed, the corrected temperature field response data and the corrected node embedding vector are classified and stored according to the interference event number and simulation task number, and stored in the system database for subsequent query and analysis.

[0178] Generate a simulation report detailing the parameter adjustments (parameter name, original value, new value, and adjustment range), operation status (start time, end time, any anomalies, convergence), and data collection (number of collected time nodes, total data volume, and data quality assessment). The report is saved as a document to document the temperature field optimization process.

[0179] Step S160: Compare the embedding differences of the graph nodes corresponding to the temperature field response data before and after correction to determine whether the abnormal temperature field pattern has been alleviated. If not, re-identify the dynamic interference event characteristics and update the furnace space association graph until the abnormal temperature field pattern is eliminated, and extract the final simulation parameters as the incinerator temperature field optimization plan.

[0180] The graph node embedding vectors corresponding to the temperature field response data before and after correction are compared, and the differences are analyzed to assess whether the abnormal temperature field pattern has been alleviated. If the abnormal pattern has not been alleviated, it is necessary to re-execute the steps of dynamic interference event identification and furnace space association map update, and repeat the optimization process until the abnormal temperature field pattern is eliminated. Finally, the finalized simulation parameters are extracted as the optimization plan for the incinerator temperature field.

[0181] Step S161: Calculate the difference in embedding vectors of corresponding nodes before and after correction at the same time point. The difference calculation adopts a vector similarity measurement method, such as cosine similarity or Euclidean distance, to obtain a difference value sequence.

[0182] For each node in the furnace spatial association graph, at the same time point (e.g., the same moment after the disturbance), the node embedding vectors before and after correction are extracted. The similarity between the two vectors is calculated using cosine similarity; the closer the cosine similarity is to 1, the more similar the vectors are. Alternatively, the spatial distance between the two vectors is calculated using Euclidean distance; the smaller the distance, the more similar the vectors are.

[0183] The calculated similarity or distance values ​​are sorted by nodes and time points to form a difference value sequence, which reflects the impact of the correction measures on the embedding vector of each node.

[0184] Step S162: Analyze the changes in node embedding features corresponding to the abnormal temperature field pattern according to the difference value sequence. If the difference value of the abnormal feature exceeds the preset improvement threshold, it is determined that the abnormal temperature field pattern is alleviated; otherwise, it is determined that it is not alleviated.

[0185] The preset improvement threshold is set according to the temperature field optimization target. For example, for cosine similarity, a threshold is set. When the similarity between the corrected embedding vector and the normal pattern feature vector is improved by more than the threshold than that before correction, it indicates that the abnormal feature is improved.

[0186] Analyze the changes in the embedded features of nodes related to the abnormal temperature field pattern (such as the temperature anomaly nodes previously determined) in the difference value sequence. If the difference values ​​of these nodes exceed the preset improvement threshold, it means that the corrective measures are effective and the abnormal temperature field pattern has been alleviated; if the difference value does not reach the threshold, it is determined that the abnormal pattern has not been alleviated.

[0187] Step S163: If it is determined that the temperature field abnormal pattern has not been alleviated, return to step S110 to re-identify the dynamic interference event characteristics, and at the same time update the edge weights (based on the latest heat conduction data) and node attributes of the furnace space association graph, and repeat the subsequent steps.

[0188] If the abnormal temperature field pattern persists, it indicates that the previous interference event identification may have been biased, or the furnace space association map may not accurately reflect the current heat conduction state. In this case, return to step S110 to recollect operating data and identify dynamic interference event characteristics. It may be necessary to adjust anomaly detection parameters (such as the reconstruction error threshold) to improve identification accuracy.

[0189] At the same time, based on the latest temperature field data and operating status, the heat transfer coefficients (edge ​​weights) between adjacent nodes in the furnace spatial association graph, as well as the node geometric properties or basic characteristics (such as changes in node properties caused by minor changes in the furnace structure), are updated to make the association graph more consistent with the current actual situation. The process of steps S120 to S150 is then repeated until the abnormal temperature field pattern is alleviated.

[0190] Step S164: If it is determined that the abnormal temperature field mode is eliminated, the simulation parameters in the current CFD simulation system are extracted, including heat source parameters, fluid flow parameters and heat exchange parameters, and the incinerator temperature field optimization plan is formed. The plan includes the specific values ​​of the parameters, applicable interference scenarios and expected temperature field improvement effects.

[0191] When the abnormal temperature field mode is eliminated (that is, all node embedding features related to the abnormal mode are within the normal range), the simulation parameters currently used in the CFD simulation system are extracted, including heat source parameters such as the burner's fuel flow rate and temperature setting value; fluid flow parameters such as the fan's air volume and wind speed; and heat exchange parameters such as the cooling medium's flow rate and temperature.

[0192] These parameters are organized into an incinerator temperature field optimization plan. The incinerator temperature field optimization plan clearly defines the specific values ​​of each parameter, explains the interference scenarios to which the incinerator temperature field optimization plan is applicable (such as temperature anomalies caused by sudden changes in feed composition), and describes the expected temperature field improvement effect based on the simulation results (such as the degree of improvement in temperature field uniformity and the reduction in temperature fluctuation amplitude). The incinerator temperature field optimization plan can serve as the basis for actual adjustment of incinerator operating parameters and guide on-site operations.

[0193] Figure 2 A schematic diagram illustrates exemplary hardware and software components of a CFD numerical simulation-based incinerator temperature field optimization system 100, which can implement the concepts of the present application, according to some embodiments of the present application. For example, a processor 120 can be used in the CFD numerical simulation-based incinerator temperature field optimization system 100 to perform the functions described in the present application.

[0194] The incinerator temperature field optimization system 100 based on CFD numerical simulation can be a general-purpose server or a special-purpose server, both of which can be used to implement the incinerator temperature field optimization method based on CFD numerical simulation of the present application. Although only one server is shown in this application, for convenience, the functions described in this application can be implemented in a distributed manner on multiple similar platforms to balance the processing load.

[0195] For example, the incinerator temperature field optimization system 100 based on CFD numerical simulation may include a network port 110 connected to a network, one or more processors 120 for executing program instructions, a communication bus 130, and storage media 140 in different forms, such as a disk, ROM, or RAM, or any combination thereof. Exemplarily, the incinerator temperature field optimization system 100 based on CFD numerical simulation may also include program instructions stored in ROM, RAM, or other types of non-transitory storage media, or any combination thereof. The method of the present application can be implemented according to these program instructions. The incinerator temperature field optimization system 100 based on CFD numerical simulation also includes an I / O interface 150 between a computer and other input and output devices.

[0196] For ease of explanation, only one processor is described in the incinerator temperature field optimization system 100 based on CFD numerical simulation. However, it should be noted that the incinerator temperature field optimization system 100 based on CFD numerical simulation in this application can also include multiple processors, so the steps performed by one processor described in this application can also be performed jointly or individually by multiple processors. For example, if the processor of the incinerator temperature field optimization system 100 based on CFD numerical simulation executes step A and step B, it should be understood that step A and step B can also be performed jointly by two different processors or individually in one processor. For example, the first processor executes step A, the second processor executes step B, or the first processor and the second processor execute steps A and B together.

[0197] In addition, an embodiment of the present invention further provides a readable storage medium, in which computer executable instructions are preset. When a processor executes the computer executable instructions, the incinerator temperature field optimization method based on CFD numerical simulation as described above is implemented.

[0198] It should be noted that in order to simplify the description of the present invention and thus help understand one or more embodiments of the invention, in the foregoing description of the embodiments of the present invention, multiple features are sometimes combined into one embodiment, figure or description thereof.

Claims

1. A method for optimizing the temperature field of an incinerator based on CFD numerical simulation, characterized in that: The method comprises: Identify dynamic interference events during the operation of the incinerator and extract event feature information of the dynamic interference events, wherein the event feature information includes the interference type, the interference occurrence time period and the interference impact range; Construct a spatial association graph for the incinerator furnace, using each spatial unit of the furnace as a graph node and the heat conduction relationship between each spatial unit of the furnace as the graph edge weight. Embed the event feature information into the corresponding graph node, learn the association between interference features and temperature field anomalies through a graph neural network, and output the temperature anomaly association link corresponding to the interference event. Input the temperature anomaly associated link into the CFD simulation system to generate a target simulation task for the associated link, wherein the target simulation task includes a target simulation area, simulation duration, and key monitoring temperature indicators; Execute the target simulation task, obtain the temperature field dynamic response data of the associated links, analyze the embedding changes of the graph nodes corresponding to the temperature field dynamic response data through the graph neural network, determine the correlation strength between the temperature field response and the interference event, and then determine the simulation correction direction that needs to be adjusted; generating a CFD simulation parameter adjustment instruction according to the simulation correction direction, inputting the CFD simulation parameter adjustment instruction into the CFD simulation system to update simulation parameters, re-executing the target simulation task, and obtaining corrected temperature field response data; Compare the embedding differences of the graph nodes corresponding to the temperature field response data before and after correction to determine whether the abnormal temperature field pattern has been alleviated. If not, re-identify the dynamic interference event characteristics and update the furnace space association graph until the abnormal temperature field pattern is eliminated. Extract the final simulation parameters as the incinerator temperature field optimization plan.

2. The incinerator temperature field optimization method based on CFD numerical simulation according to claim 1 is characterized in that: The identifying of dynamic interference events during the operation of the incinerator and extracting event feature information of the dynamic interference events include: Collecting operation data at a preset frequency through each monitoring point in the incinerator operation status monitoring network, the operation data including feed data, burner data, flue gas data and furnace wall data; Taking each monitoring point as the initial node and the standardized spatial distance and data correlation between monitoring points as the edge weight, a monitoring point spatial correlation graph is constructed. Node anomaly detection is performed on the monitoring point spatial correlation graph through a graph neural network to capture abnormal nodes. Count the duration of data fluctuations of abnormal nodes, and combine with the incinerator operation mechanism to determine the start and end time of interference corresponding to the abnormal nodes to form the interference period; Determine the spatial boundaries of the interference impact range based on the distribution location of abnormal nodes and the data fluctuation amplitude, and simultaneously determine the interference type based on the monitoring data type corresponding to the abnormal nodes; Establish dynamic interference event classification rules, and classify dynamic interference events into different priority levels based on the number of nodes covered by the interference impact range, the duration of the interference period, and the data fluctuation amplitude of the abnormal nodes; The interference type, interference occurrence period, interference impact range and interference priority level are integrated to form an event feature information set containing priority information.

3. The incinerator temperature field optimization method based on CFD numerical simulation according to claim 2 is characterized in that: The method uses each monitoring point as the initial node, the standardized spatial distance and data correlation between monitoring points as the graph edge weight, and constructs a monitoring point spatial correlation graph. Node anomaly detection is performed on the monitoring point spatial correlation graph through a graph neural network to capture abnormal nodes, including: The distribution of monitoring points in the roaming incinerator operation status monitoring network is analyzed. Each monitoring point is used as the initial node of the monitoring point spatial association graph, and the spatial coordinates and monitoring data types of each initial node are recorded. Calculate the normalized spatial distance between any two initial nodes, combine the historical correlation coefficient of the monitoring data of the two initial nodes, and use the product of the inverse of the normalized spatial distance and the historical correlation coefficient as the graph edge weight between the two initial nodes; Based on the initial nodes and graph edge weights, a spatial correlation graph of monitoring points is constructed. The graph autoencoder in the graph neural network is used to pre-train the spatial correlation graph of monitoring points to learn the embedding vectors of each initial node under normal operating conditions. Collect the operating data of each monitoring point in real time, convert the operating data into a real-time node feature vector, input it into the pre-trained graph autoencoder, and calculate the reconstruction error between the real-time node embedding vector and the initial node embedding vector under normal operating conditions; Set a reconstruction error threshold, mark nodes whose reconstruction errors exceed the reconstruction error threshold as candidate abnormal nodes, perform neighborhood node verification on the candidate abnormal nodes, and check whether the reconstruction errors of the neighboring nodes of the candidate abnormal nodes also exceed the reconstruction error threshold; If the reconstruction errors of more than a preset proportion of the neighboring nodes of the candidate abnormal node exceed the reconstruction error threshold, the candidate abnormal node is determined to be the final abnormal node, and all the final abnormal nodes are extracted to form an abnormal node set.

4. The incinerator temperature field optimization method based on CFD numerical simulation according to claim 1 is characterized in that: The incinerator furnace space association graph is constructed, with each furnace space unit as a graph node and the heat conduction relationship between units as the graph edge weight, the event feature information is embedded in the corresponding graph node, the association relationship between interference features and temperature field anomalies is learned through a graph neural network, and the temperature anomaly association link corresponding to the interference event is output, including: The incinerator furnace is divided into multiple spatial units, each spatial unit is used as the basic node of the furnace space association diagram, and the spatial coordinates and geometric properties of each basic node are recorded; Based on the heat conduction mechanism of the incinerator, the heat conduction coefficient between adjacent spatial units is calculated, and the heat conduction coefficient is used as the edge weight between the corresponding basic nodes in the furnace space association graph to construct the initial furnace space association graph; Determine the node features corresponding to each abnormal temperature field mode and form an abnormal mode node feature library; The event feature information is used as an interference feature vector and embedded into the corresponding basic node in the initial furnace space association graph. The furnace space association graph embedded with the interference feature vector is trained using a graph convolutional neural network to learn the mapping relationship between the interference feature vector and the features in the abnormal pattern node feature library, thereby forming an interference temperature association learning model. The interference temperature association learning model is iteratively optimized through historical interference temperature data, so that the interference temperature association learning model can output a set of temperature anomaly nodes corresponding to the interference event, and the furnace function link corresponding to the temperature anomaly node set is determined as the temperature anomaly association link; wherein, when the interference temperature association learning model outputs the temperature anomaly association link, the association confidence is output synchronously. The association confidence is obtained by calculating the similarity of the graph node embedding. If the association confidence is lower than the preset association confidence threshold, the feature embedding and model training are re-executed.

5. The incinerator temperature field optimization method based on CFD numerical simulation according to claim 4 is characterized in that: The graph convolutional neural network is used to train the furnace space association graph embedded with interference features, learn the mapping relationship between the interference feature vector and the features in the abnormal pattern node feature library, and form an interference temperature association learning model, including: Constructing a training data set, wherein the training data set includes interference feature vectors of historical interference events, corresponding furnace space association graphs, and a set of manually annotated temperature anomaly nodes; The furnace spatial association graph in the training dataset is input into the graph convolutional neural network, and the node features are aggregated through the graph convolution layer to generate a node embedding vector. The process of aggregating the node features through the graph convolution layer combines the weighted sum of the node's own features and the features of adjacent nodes; In the fully connected layer of the graph convolutional neural network, the node embedding vector is compared with the features in the abnormal pattern node feature library, the similarity loss between the node embedding vector and the features in the abnormal pattern node feature library is calculated, and the network parameters of the graph convolutional neural network are optimized using the gradient descent algorithm; An attention mechanism layer is introduced to assign attention weights to the node embedding vectors output by the graph convolution layer, allowing the graph convolutional neural network to pay more attention to node features related to temperature anomalies. The training dataset is divided into a training subset and a validation subset. The training subset is used to train the network parameters of the graph convolutional neural network, and the validation subset is used to verify the matching degree between the temperature anomaly node set output by the graph convolutional neural network and the manually labeled temperature anomaly node set. If the matching degree is lower than the preset matching degree standard, adjust the number of layers, convolution kernel size and attention weight calculation method of the graph convolutional neural network, and retrain the graph convolutional neural network; If the matching degree reaches the preset matching degree standard, the training is stopped to form a stable interference temperature association learning model.

6. The incinerator temperature field optimization method based on CFD numerical simulation according to claim 1 is characterized in that: The step of inputting the temperature anomaly associated link into the CFD simulation system and generating a target simulation task for the associated link includes: Calling the CFD simulation system to receive the temperature anomaly correlation link and the corresponding correlation confidence, and if the correlation confidence is higher than a preset correlation confidence threshold, starting the target simulation task generation process; The importance of nodes corresponding to the temperature anomaly association links in the furnace space association graph is ranked by graph neural network, and the top K nodes and their adjacent nodes are extracted to determine the spatial range of the target simulation area. The spatial range of the target simulation area includes the spatial units corresponding to potential nodes where interference may spread. According to the interference occurrence period and interference duration characteristics in the event characteristic information, the time span of the simulation is set, and the time span includes the baseline stability period before the interference occurs, the interference occurrence period, and the recovery period after the interference is eliminated; Based on the node characteristics of the temperature field anomaly pattern, key monitoring temperature indicators are determined. The key monitoring temperature indicators include the core temperature parameters corresponding to the temperature field anomaly pattern and the associated derived temperature parameters. At the same time, the monitoring frequency of different simulation periods is set. The monitoring frequency during the interference period is higher than that in other simulation periods. CFD simulation resources are allocated based on the interference priority level to generate target simulation tasks for resource optimization.

7. The incinerator temperature field optimization method based on CFD numerical simulation according to claim 1 is characterized in that: The target simulation task is executed, the temperature field dynamic response data of the associated link is obtained, and the graph node embedding changes corresponding to the temperature field dynamic response data are analyzed through a graph neural network to determine the correlation strength between the temperature field response and the interference event, thereby determining the simulation correction direction that needs to be adjusted, including: Calling the CFD simulation system to load the target simulation area geometry model corresponding to the target simulation task, importing the initial operating parameters, and starting the simulation operation according to the set time step; During the simulation operation, the temperature data corresponding to each node in the target simulation area is collected according to the monitoring frequency to form a temperature field dynamic response data sequence, and the interference simulation state of each time node is synchronously recorded; The temperature field dynamic response data sequence is embedded into the corresponding nodes of the furnace space association graph to generate a dynamic node embedding vector. The node embedding update algorithm of the graph neural network is used to calculate the change of the dynamic node embedding vector before and after the interference occurs. Constructing correlation strength analysis dimensions, which include time, space, and feature dimensions. The time dimension analyzes the synchronization between interference occurrence and changes in dynamic node embedding vectors. The space dimension analyzes the overlap between abnormal nodes and interference-affected nodes. The feature dimension analyzes the matching degree between changes in dynamic node embedding vectors and interference features. Calculate the correlation scores of each correlation strength analysis dimension through the graph neural network, standardize the correlation scores of each correlation strength analysis dimension, and then sum them up according to the preset weights to obtain the comprehensive correlation strength. If the comprehensive correlation strength is lower than the preset comprehensive correlation strength threshold, recheck the parameter settings of the target simulation task; If the comprehensive correlation strength is higher than the preset comprehensive correlation strength threshold, the simulation correction direction is determined based on the temperature data characteristics corresponding to the dynamic node embedding vector change and combined with the incinerator temperature field optimization goal. The simulation correction direction includes the heat source adjustment direction, the fluid flow adjustment direction and the heat exchange adjustment direction.

8. The incinerator temperature field optimization method based on CFD numerical simulation according to claim 7 is characterized in that: If the comprehensive correlation strength is higher than the preset comprehensive correlation strength threshold, based on the temperature data characteristics corresponding to the dynamic node embedding vector change amount and combined with the incinerator temperature field optimization goal, the simulation correction direction is determined, including: Extracting temperature data features corresponding to the dynamic node embedding vector change, wherein the temperature data features include the location of the temperature abnormal node, the temperature change amplitude, the temperature fluctuation frequency and the temperature gradient distribution between nodes; Obtaining an incinerator temperature field optimization target, wherein the incinerator temperature field optimization target includes a temperature field uniformity target, a temperature stability target, and a temperature range compliance target; If the temperature data characteristics show that the local node temperature is higher than the temperature range corresponding to the temperature interval compliance target, and the correlation strength analysis shows that the embedding change of the local node is significantly correlated with the embedding change of the heat source input node, then the simulation correction direction is determined to be the heat source adjustment direction, specifically, reducing the heat source input intensity corresponding to the heat source input node; If the temperature data characteristics show that the temperature gradient between nodes exceeds the temperature gradient range corresponding to the temperature field uniformity target, and the correlation strength analysis shows that the temperature gradient change is strongly correlated with the node embedding change corresponding to the fluid flow path, then the simulation correction direction is determined to be the fluid flow adjustment direction, specifically, optimizing the flow velocity parameters of the node area corresponding to the fluid flow path; If the temperature data characteristics show that the node temperature change rate exceeds the temperature change rate range corresponding to the temperature stability target, and the correlation strength analysis shows that the temperature change rate is strongly correlated with the embedding change of the heat exchange node, then the simulation correction direction is determined to be the heat exchange adjustment direction, specifically adjusting the heat exchange component parameters corresponding to the heat exchange node; The feasibility of the determined simulation correction direction is verified, and the parameter adjustment range of the CFD simulation system and the actual operation limitations of the incinerator are combined to ensure that the simulation correction direction is within the feasible range. If it is not feasible, the temperature data characteristics corresponding to the dynamic node embedded vector change are re-analyzed to adjust the simulation correction direction.

9. The incinerator temperature field optimization method based on CFD numerical simulation according to claim 1 is characterized in that: Generating a CFD simulation parameter adjustment instruction according to the simulation correction direction, inputting the CFD simulation parameter adjustment instruction into a CFD simulation system to update simulation parameters, re-executing a target simulation task, and obtaining corrected temperature field response data, includes: Establish a mapping relationship library between simulation correction directions and CFD parameters. The mapping relationship library records the core parameter types corresponding to different simulation correction directions and the constraint conditions for parameter adjustment. The constraint conditions are classified into physical constraints and numerical constraints. Physical constraints include the range of equipment safe operation parameters, and numerical constraints include simulation stability requirements. Based on the determined simulation correction direction, the corresponding core parameter type is extracted from the mapping relationship library, and the current value of the core parameter type in the CFD simulation system is obtained. Combined with the dynamic node embedding vector change corresponding to the temperature field dynamic response data, the reasonable range of core parameter type adjustment is determined; Generate a CFD simulation parameter adjustment instruction, wherein the CFD simulation parameter adjustment instruction includes a target parameter identifier, an adjustment direction, an adjustment range, and a verification rule for a parameter value after adjustment, so that the parameter value after adjustment satisfies a constraint condition; Import CFD simulation parameter adjustment instructions through the parameter update interface of the CFD simulation system, automatically replace the original parameter values, and back up the parameter configuration before adjustment; Call the basic configuration of the target simulation task, keep the target simulation area, simulation duration and monitoring frequency unchanged, start the re-simulation operation, collect temperature data at the set frequency, form the corrected temperature field response data, and synchronously record the embedded vector of each node after correction.

10. An incinerator temperature field optimization system based on CFD numerical simulation, characterized in that: It includes a processor and a memory, the memory is connected to the processor, the memory is used to store programs, instructions or codes, and the processor is used to execute the programs, instructions or codes in the memory to implement the incinerator temperature field optimization method based on CFD numerical simulation as described in any one of claims 1 to 9.

Images