Kalman filtering method and system for air tightness detection equipment

Abstract

The invention relates to the technical field of air tightness detection, and discloses a Kalman filtering method and system for air tightness detection equipment, and the Kalman filtering method for the air tightness detection equipment comprises the steps: carrying out the modeling of a multi-detection-point air tightness detection system through a graph structure model, and generating system topological representation; automatically identifying and detecting a network dynamic topological structure by applying an edge functional verification algorithm, and outputting a real-time topological relation matrix; representing system state evolution by adopting a continuous time neural differential equation model, and generating continuous time domain state representation; analyzing a historical topology change mode by using a time sequence prediction model, and predicting a future topology evolution trajectory; estimating and updating the system state in real time by using a continuous depth Kalman filtering algorithm, and outputting a high-precision air tightness detection result; the method can adapt to the dynamic change of a detection network topology structure in an industrial environment, and improves the accuracy of air tightness detection.

Description

Technical Field

[0001] The present invention relates to the technical field of airtightness detection, and more specifically, it relates to a Kalman filtering method and system for airtightness detection equipment. Background Art

[0002] In large industrial systems, airtightness detection is an important link to ensure production safety. Especially in high-speed fluid systems such as chemical and petroleum industries, any minor leakage may lead to serious safety accidents and economic losses. Currently, airtightness detection systems generally adopt a multi-detection-point distributed deployment method to capture possible leakage points in the system.

[0003] Existing airtightness detection technologies mainly include independent detection methods based on traditional Kalman filtering, which independently process and analyze the data of each detection point. However, this method has the following main technical problems: it cannot effectively utilize the spatio-temporal correlation between detection points, resulting in insufficient detection accuracy, especially in the case of minor leakage, it cannot accurately locate the leakage point; it cannot adapt to the dynamic changes of the detection network topology in the industrial environment, and operations such as valve switching and pipeline switching often lead to false alarms; it cannot accurately model and capture the rapid topology changes in high-speed gas flow systems in the continuous time domain, resulting in system response lag.

[0004] Therefore, there is a need for an airtightness detection method that can make full use of the spatio-temporal correlation of multiple detection points, adapt to the dynamic changes of the network topology, and achieve accurate modeling in the continuous time domain, so as to improve the detection accuracy and system reliability. Summary of the Invention

[0005] The present invention provides a Kalman filtering method and system for airtightness detection equipment, which solves the technical problems in the related art of being unable to effectively utilize the spatio-temporal correlation of multiple detection points, being unable to adapt to the dynamic changes of the network topology, and being unable to achieve accurate modeling in the continuous time domain.

[0006] The present invention provides a Kalman filtering method for airtightness detection equipment, including: Modeling a multi-detection-point airtightness detection system using a graph structure model to generate a system topology representation; Based on the system topology representation, applying an edge functional verification algorithm to automatically identify the dynamic topology structure of the detection network and output a real-time topology relationship matrix; Combining the real-time topology relationship matrix, using a continuous-time neural differential equation model to represent the system state evolution and generate a continuous-time domain state representation; Based on the real-time topology relationship matrix and the time-domain state representation, using a time series prediction model to analyze the historical topology change pattern and predict the future topology evolution trajectory; Based on the comprehensive continuous-time domain state representation and predicted future topological evolution trajectory, the continuous-depth Kalman filtering algorithm is applied to estimate and update the system state in real time, and high-precision airtightness detection results are output.

[0007] In a preferred embodiment, the steps of modeling the multi-detection-point airtightness detection system using the graph structure model include: Model the airtightness detection system as a graph structure, where nodes represent detection points and edges represent physical connection relationships; Use a graph convolutional network to extract the spatial correlation features between detection points; Combine a gated recurrent unit to capture the time dependence of the detection point states; Output the spatio-temporal feature representation of the detection system.

[0008] In a preferred embodiment, the steps of applying the edge functional verification algorithm to automatically identify the dynamic topological structure of the detection network include: Automatically discover the true physical connection relationships based on the pressure wave propagation characteristics; Introduce a graph attention algorithm to adaptively adjust the influence weights between nodes; Construct a weighted adjacency matrix; Output the topologically related matrix updated in real time.

[0009] In a preferred embodiment, the steps of representing the system state evolution using a continuous-time neural differential equation model include: Construct a continuous-time differential equation for the system state evolution; Define the neural network structure, using a combination structure of graph convolution and multi-layer perceptron; Solve the differential equation through an adaptive numerical integration method to obtain the state representation in the continuous-time domain.

[0010] In a preferred embodiment, the steps of analyzing the historical topological change patterns using a time series prediction model include: Construct a spatio-temporal sequence prediction model; Design the prediction function structure, using a spatio-temporal attention network architecture; Adjust the model parameters in advance based on the predicted topological changes; Output the predicted topological structure and pre-adjusted model parameters.

[0011] In a preferred embodiment, the steps of applying the continuous-depth Kalman filtering algorithm to estimate and update the system state in real time include: Use a stochastic differential equation to represent the system dynamics model; Design the observation equation; Dynamically generate the Kalman filter state transition matrix through the graph attention algorithm; Execute the continuous discrete extended Kalman filtering algorithm, including the prediction step and the update step; Output the filtered system state estimation value and the anomaly detection result.

[0012] In a preferred embodiment, the graph convolutional network includes an input layer, an intermediate layer, and an output layer, wherein: The input layer receives the original node features; The intermediate layer contains multiple hidden units and uses an activation function to enhance the non-linear expression ability; The output layer maps the features to the hidden state space.

[0013] In a preferred embodiment, the step of automatically discovering the true physical connection relationship based on the pressure wave propagation characteristics includes: Analyze the time-delay correlation of the pressure signals between the detection points; Judge whether there is a physical connection between the detection points according to the correlation exceeding the preset threshold.

[0014] In a preferred embodiment, the spatio-temporal attention network includes: An encoder for encoding the graph structure into a hidden state vector; An attention module for calculating the importance weights of the graph structures at different time points; A recurrent neural network for processing sequence information; A decoder for decoding the hidden state into a predicted graph structure.

[0015] In a preferred embodiment, an airtightness detection system applicable to a dynamic topology network is used to execute a Kalman filtering method for an airtightness detection device, including: A multi-detection point graph structure modeling module for modeling the airtightness detection system and generating a system topology representation; A dynamic topology identification module for automatically identifying the dynamic topology structure of the detection network by applying an edge functionality verification algorithm; A continuous time state representation module for representing the system state evolution using a continuous time neural differential equation model; A topology prediction module for analyzing the historical topology change pattern using a time series prediction model and predicting the future topology evolution trajectory; A continuous depth Kalman filtering module for performing real-time estimation and update of the system state and outputting a high-precision airtightness detection result.

[0016] The beneficial effects of the present invention are as follows: By modeling the structural relationship between the detection points, the collaborative filtering of the entire system is realized, the detection accuracy is improved, and the leakage point positioning accuracy is improved; It can adapt to the dynamic changes of the detection network topology in the industrial environment in real time, shortening the response time; representing the system state evolution by continuous-time differential equations, accurately capturing the rapid topology changes in the high-speed gas flow system, and reducing the model prediction error; It can predict the future topology evolution trajectory, adjust the model parameters in advance, and improve the system stability. Description of the Drawings

[0017] Figure 1 It is a flowchart of a Kalman filtering method for an airtightness detection device according to the present invention. Detailed Embodiments

[0018] Now, the subject matter described herein will be discussed with reference to example embodiments. It should be understood that discussing these embodiments is only to enable those skilled in the art to better understand and thus implement the subject matter described herein. Without departing from the scope of protection of the content of this specification, changes can be made to the functions and arrangements of the elements discussed. Each example can omit, substitute, or add various processes or components as needed. Additionally, the features described in some examples can also be combined in other examples.

[0019] In at least one embodiment of the present invention, an airtightness detection method applicable to a dynamic topology network is disclosed, as Figure 1 shown, including the following steps: Step 1, use a graph structure model to model the multi-detection point airtightness detection system and generate a system topology representation; The specific steps are as follows: 1.1, model the airtightness detection system as a graph structure , where the node set represents the set of detection points, and the edge set represents the physical connection relationship. For any detection points , if there is a physical connection between them, then the edge . Each detection point has a feature vector , which contains sensor measurement values such as pressure, temperature, and flow rate.

[0020] 1.2, adopt a graph convolutional network to extract the spatial correlation features between detection points. For the node feature matrix at time (where is the number of detection points, is the feature dimension), the graph convolutional operation is defined as: ; where, represents the The hidden representation of the nodes in the layer, i.e., the node features after the graph convolution operation, is the adjacency matrix with self-connections added, is the original adjacency matrix, is the identity matrix of order is the pair-wise angle matrix, represents the node degree, , indicating that the input of the initial layer is equal to the original node feature matrix, is the learnable weight matrix of the layer, represents the sigmod activation function.

[0021] The graph convolutional network used in this embodiment consists of three layers: an input layer, an intermediate layer, and an output layer. Among them, the input layer receives the original node features, with a dimension of ; the intermediate layer contains 64 hidden units and uses the ReLU activation function to enhance the non-linear expression ability; the output layer maps the features to the -dimensional hidden state space.

[0022] The specific applications of the graph convolutional network in the airtightness detection scenario include: effectively extracting features such as the pressure difference and temperature correlation between adjacent detection points; automatically learning the attenuation law of the pressure wave transmitted between detection points according to the physical connection relationship; in the complex pipe network structure of a chemical plant, being able to automatically identify the topological relationship differences caused by valve state changes.

[0023] 1.3, combined with the gated recurrent unit (GRU) to capture the time dependence of the detection point state. For the th node, the GRU update formula is: ; ; ; ; Among them, is the reset gate, is the update gate, is the candidate hidden state, is the hidden state of node at time , is the hidden state of node at time , is the hidden state of node at time The input features represents element-wise multiplication , , are respectively used for the calculations of the reset gate, update gate, and candidate hidden state , , are respectively the bias vectors of the reset gate, update gate, and candidate hidden state represents the sigmod activation function is the hyperbolic tangent activation function

[0024] The GRU network in this embodiment is configured as a single-layer structure, with the hidden state dimension set to 32, and the input is the node features at the current moment concatenated with the hidden state at the previous moment . In the airtightness detection system, the specific applications of the GRU network include: Capturing the time-series change laws of parameters such as the pressure and temperature at the detection points, such as identifying the propagation delay of pressure waves in the oil pipeline system; Learning the short-term and long-term memory characteristics of system parameters under different operating conditions (such as flow rate changes, temperature fluctuations); In a high-speed gas flow system, being able to distinguish the pressure fluctuations caused by normal operations from the abnormal pressure drop patterns caused by leaks

[0025] 1.4, outputting the spatio-temporal feature representation of the detection system , where is the hidden state dimension. This representation simultaneously includes the system topology structure information and the time evolution characteristics of the node states

[0026] Step 2, based on the system topology representation, applying the edge functional verification algorithm to automatically identify the dynamic topology structure of the detection network and output the real-time topology relationship matrix; The specific steps are as follows: 2.1, Automatically discovering the real physical connection relationships based on the pressure wave propagation characteristics. For any two detection points and , by analyzing the time-delay correlation of the pressure signals to determine whether there is a physical connection between them: ; where, represents the time-delay correlation of the pressure signals, and are respectively the pressure values of the detection points and at time , and They are detection points and the mean values at, is the observation window size, is the time delay parameter. If ( is the preset threshold), then it is considered that there is a physical connection between the detection points and .

[0027] 2.2, Introduce the graph attention algorithm to adaptively adjust the influence weights between nodes. For each pair of connected nodes , calculate the attention coefficient: ; ; where, is the unnormalized attention score, is the normalized attention coefficient, , are the attention weight matrix and the feature transformation matrix respectively, represents the vector concatenation operation, represents the neighbor set of node , LeakyReLU is an activation function with a small negative slope part, , are the hidden state representations of points and respectively, represents the exponential function.

[0028] 2.3, Construct the weighted adjacency matrix , where the element represents the influence intensity of node on node at time .

[0029] 2.4, Output the topologically related matrix updated in real time , where, represents the topologically related matrix at time , represents the weighted adjacency matrix at time , represents the node feature matrix at time . This matrix synthesizes the physical connection relationship and node state characteristics of the system and can reflect the dynamic changes of the system topology structure.

[0030] Step 3, Combine the real-time topologically related matrix and use the continuous-time neural differential equation model to represent the system state evolution and generate the continuous-time domain state representation; The specific steps are as follows: 3.1, Construct a continuous-time differential equation for the evolution of the system state. For the system state at time , its evolution law is expressed as: ; where represents how the system state changes continuously over time, is the node hidden state matrix, is the time-varying graph structure, is a neural network with parameter , describing the state change rate.

[0031] 3.2, Define the structure of the neural network . Adopt a combined structure of graph convolution and multi-layer perceptron: ; where is a parameterized neural network function, represents the set of learnable parameters of this neural network, represents the system state vector at time , represents the changing graph structure at time , is the time variable, GCN represents the graph convolutional network, MLP represents the multi-layer perceptron, is the time-dependent coefficient.

[0032] The neural differential equation network in this embodiment is specifically implemented as: The GCN part adopts a 2-layer structure with 32 hidden units in each layer, and uses the Tanh activation function to ensure smooth and differentiable output; The MLP part contains a 3-layer fully connected network, and the hidden layer dimensions are 64, 32, and 32 respectively, and also uses the Tanh activation function.

[0033] The time-dependent coefficient is set to 0.01 to balance the time sensitivity of the system.

[0034] In the airtightness detection system, the specific applications of this network include: Real-time tracking of pressure wave propagation in the complex pipe network of a petrochemical plant, and accurately modeling the continuous change of the system state; Adapting to topological mutations caused by operations such as rapid valve opening and closing, and continuously capturing the transition state of the system; In the high-speed gas flow scenario, being able to distinguish normal operating condition fluctuations from slow pressure drops caused by minor leaks.

[0035] 3.3, Solve the differential equation by an adaptive numerical integration method to obtain the state representation in the continuous time domain. For the time interval , use the 5th-order Runge-Kutta method to solve: ; where, represents the system state vector at the termination time , represents the system state vector at the initial time , denotes the definite integral operation over the time interval , is a neural network function with parameter , representing the instantaneous change rate of the system state, represents the system state vector at time , represents the change graph structure at time , represents an infinitesimal time increment.

[0036] 3.4, Output the system state representation in the continuous time domain , which can accurately capture the continuous state changes in the high-speed gas flow system and overcome the information loss problem caused by discrete time step updates.

[0037] Step 4, Based on the real-time topological relationship matrix and the time domain state representation, use a time series prediction model to analyze the historical topological change patterns and predict the future topological evolution trajectory; The specific steps are as follows: 4.1, Construct a spatio-temporal sequence prediction model. This model takes the topological structure sequence at the past time points as input and predicts the topological structure at the future time , where , , , and the first respectively represent the historical topological structure sequences at times , , , , represents the earliest historical time point, represents the historical time window size, represents the time step, represents the current time point.

[0038] 4.2, Design the structure of the prediction function and adopt a spatio-temporal attention network architecture: ; Among them, represents the topological structure matrix at the future moment predicted by the model , , , respectively represent the topological structure matrices at the current moment , the previous moment , and the earliest historical moment considered ; represents the time step, represents the size of the historical time window, represents the current time point; It is implemented as a cyclic attention network: ; ; ; Among them, represents the initial hidden state vector, represents the hidden state vector after processing to the -th historical time point, represents the topological structure matrix at the future moment predicted by the model, represents the encoder function, represents the long short-term memory network unit, represents the attention mechanism function, represents the system topological structure at the historical moment , represents the time step, represents the index of the historical time point, represents the decoder function, represents the hidden state vector after processing to the -th historical time point.

[0039] The spatio-temporal attention network in this embodiment is specifically implemented as follows: The Encoder uses a 3-layer graph convolutional network to encode the graph structure into a hidden state vector; The Attention module uses scaled dot-product attention to calculate the importance weights of the graph structures at different time points; The LSTM is configured with a 2-layer structure, and each layer contains 128 hidden units; The Decoder uses a 3-layer transposed graph convolutional network to decode the hidden state into a predicted graph structure.

[0040] In the airtightness detection system, the specific applications of this network include: Predicting topological changes in the oil pipeline system caused by temperature changes or flow adjustments; Identifying and predicting the periodic change patterns of the system state caused by periodic operations (such as timed valve opening and closing); Early sensing of impending topological changes in complex industrial systems, such as pipeline reconstruction caused by production process switching.

[0041] 4.3, Adjusting the model parameters in advance based on the predicted topological changes. For the predicted topological structure , calculate the difference from the current topology : ; If ( is a preset threshold, represents the Frobenius norm), then trigger the pre-adjustment process of the model parameters, where, represents the difference matrix between the predicted topology and the current topology, represents the topology structure matrix at the future moment predicted by the model, and represents the topology structure matrix at the current moment .

[0042] 4.4, Output the predicted topological structure and the pre-adjusted model parameters, realizing the early prediction and adaptation to topological changes, and improving the stability and accuracy of the system during topological changes.

[0043] Step 5, Integrating the continuous-time domain state representation and the predicted future topological evolution trajectory, applying the continuous deep Kalman filter algorithm to perform real-time estimation and update of the system state, and outputting high-precision airtightness detection results; The specific steps are as follows: 5.1, Represent the system dynamics model using a stochastic differential equation. The evolution law of the system state is expressed as: ; Where, represents the differential change of the system state, is the drift function, representing the deterministic part; is the diffusion function, representing the random part; is the Wiener process, representing the system noise.

[0044] 5.2, Design the observation equation. For the system state at time , the relationship between the observed value is: ; Among them, represents the observation value vector at time , is the observation matrix, is the observation noise, which follows a Gaussian distribution .

[0045] 5.3, Dynamically generate the Kalman filter state transition matrix through the graph attention algorithm. At discrete time points , the state transition matrix is generated by the graph attention convolutional network: ; Among them, represents the state transition matrix from time to , is the graph structure at time , is the node hidden state matrix at the previous time, and GATConv represents the graph attention convolution operation.

[0046] 5.4, Execute the continuous-discrete extended Kalman filter algorithm. It includes a prediction step and an update step: Prediction step: ; ; Update step: ; ; ; Among them, is the state prediction value at time, is the prediction error covariance matrix, is the Kalman gain, is the state update value, is the updated error covariance matrix, is the process noise covariance matrix, represents the observation matrix, represents the transpose matrix of, represents the actual observation value at time, represents the expected observation value calculated based on the predicted state, represents the identity matrix, represents the observation noise covariance matrix.

[0047] The continuous depth Kalman filter algorithm in this embodiment is specifically implemented as: The integration process uses the Dormand-Prince method with adaptive step size control, and the maximum relative error is set to 1e-6; Process noise covariance matrix and the observation noise covariance matrix are respectively configured as diagonal matrices, where the diagonal elements of are 0.01, the diagonal elements of are 0.05.

[0048] In the airtightness detection system, the specific applications of this algorithm include: Filtering sensor noise in the multi-detection point system of chemical plants to improve the accuracy of measured values such as pressure, temperature, and flow rate; Continuously estimating the state in a high-speed gas flow system and real-time detecting system parameter anomalies caused by minute leaks; In a complex pipe network structure, accurately locating the leak point through collaborative filtering, especially for multi-point leak situations in large industrial systems; 5.5, outputting the filtered system state estimation value and the anomaly detection result. By analyzing the deviation from the predicted value to determine whether there is an airtightness anomaly in the system, and combining the graph structure information to locate the specific location where the anomaly occurs.

[0049] Real application examples of this embodiment This embodiment has been practically applied in the airtightness detection system of the ethylene production unit in a large petrochemical plant. This unit has a complex pipe network structure, including 86 airtightness detection points, which are interconnected by valves and pipes to form a dynamically changing network topology. Due to frequent process switches during production, the valve switch states change at any time, resulting in continuous dynamic changes in the topology of the detection network.

[0050] Previously, the petrochemical plant used the traditional independent Kalman filtering method to monitor each detection point, but there were the following problems in actual applications: Unable to accurately locate the leak position: When minute leaks occur, multiple interconnected detection points show anomalies simultaneously, and it is difficult for the independent filtering method to determine the true leak point position; High false alarm rate: Topology changes (such as valve switches) often cause sudden changes in sensor data, and the system cannot distinguish fluctuations caused by normal operations from leak anomalies; Response lag: Existing methods cannot effectively handle topology changes across sampling intervals in high-speed gas flow systems, resulting in system response lag.

[0051] The specific implementation of this embodiment in this ethylene unit includes the following examples: Example of multi - detection - point system modeling: In practical applications, we collected the basic information of 86 detection points, including sensor type, position coordinates, and initial connection relationships, and constructed an initial graph - structure model. The characteristic information collected for each detection point includes data such as pressure (Bar), temperature (°C), flow rate (m³ / h), and gas - component concentration, with a sampling frequency of 10 Hz. Table 1 shows an example of the characteristic information of some core detection points: Table 1: Example of characteristic information of some detection points;

[0052] The input of the graph convolutional network is an 86×12 - dimensional feature matrix (86 detection points, 12 feature parameters for each detection point), and the output is an 86×32 - dimensional hidden - state matrix, effectively extracting the spatial - correlation features between detection points.

[0053] Example of edge - functionality verification algorithm: In practical applications, this algorithm automatically identifies the true physical connection relationships of the detection network by analyzing the propagation characteristics of pressure waves in the system. We injected a small - amplitude pressure - pulse signal (peak value 0.1 Bar, duration 0.5 seconds) at a key position (P01) in the system, and then analyzed the time - delay response of the pressure signals at other detection points. The results are shown in Table 2: Table 2: Results of time - delay correlation analysis of pressure - wave propagation;

[0054] Example of topology prediction and dynamic adaptation: During the process - switching process, the system can predict future topology - structure changes based on historical topology - change patterns. Table 3 shows the prediction results of the system for the state changes of key valves during a typical process - switching operation (from normal production mode to low - load mode): Table 3: Results of topology - prediction and dynamic - adaptation example;

[0055] Verification of technical effects: We selected "improvement in detection accuracy" and "improvement in dynamic - adaptation ability" as the two most important technical effects of this implementation method for verification and analysis: Verification of the improvement effect of detection accuracy: To verify the improvement effect of this implementation method in air - tightness detection accuracy, we simulated different degrees of leakage in different areas of the ethylene plant and compared the detection results of the traditional independent Kalman - filtering method with those of this implementation method. Table 4 shows the comparison results of 10 simulated leakage tests: Table 4: Results of the comparative experiment on the accuracy of airtightness detection;

[0056] As can be seen from Table 4, compared with the traditional method, the detection time of this embodiment is shortened by 74.0% on average, and the positioning accuracy of the leakage point is improved by 71.1%, which fully proves the significant advantage of this embodiment in terms of detection accuracy. Especially for micro-leakage (<0.5 L / min), the detection ability of this embodiment is improved more significantly.

[0057] Verification of the improvement effect of dynamic adaptation ability: To verify the improvement effect of this embodiment in terms of the ability to adapt to dynamic topology changes, we recorded the response of the system to topology changes during the production process switch. Table 5 shows the comparison results between the traditional method and this embodiment in 3 typical process switch scenarios: Table 5: Results of the comparative experiment on the ability to adapt to dynamic topology changes;

[0058] As can be seen from Table 5, in the scenario of dynamic topology changes, the response time of this embodiment is shortened by 85.2% on average, and the number of false alarms is reduced from an average of 11.33 times to 0.33 times, which fully verifies the significant improvement of this embodiment in terms of the ability to adapt to dynamic topology structures.

[0059] The embodiments of the present invention have been described above, but these embodiments are not limited to the above specific embodiments. The above specific embodiments are merely illustrative and not restrictive. Under the inspiration of this embodiment, those of ordinary skill in the art can also make more equivalent embodiments in various forms, all of which fall within the protection scope of this embodiment.

Claims

1. A Kalman filtering method for airtightness detection equipment, characterized in that: The following steps are involved: Use the graph structure model to model the multi-detection point airtightness detection system and generate the system topology representation; Based on the system topology representation, the edge functional verification algorithm is applied to automatically identify and detect the dynamic topology structure of the network and output the real-time topology relationship matrix; Combined with the real-time topological relationship matrix, the continuous-time neural differential equation model is used to represent the system state evolution and generate a continuous-time domain state representation; Based on the real-time topological relationship matrix and time domain state representation, the time series prediction model is used to analyze the historical topological change pattern and predict the future topological evolution trajectory; The continuous time domain state representation and the predicted future topological evolution trajectory are integrated, and the continuous deep Kalman filter algorithm is applied to estimate and update the system state in real time to output high-precision airtightness detection results.

2. A Kalman filtering method for airtightness detection equipment according to claim 1, characterized in that: The step of modeling the multi-detection point airtightness detection system using the graph structure model includes: The airtightness detection system is modeled as a graph structure, where nodes represent detection points and edges represent physical connection relationships; A graph convolutional network is used to extract spatial correlation features between detection points; Incorporating gated recurrent units to capture the temporal dependency of the checkpoint states; Output the spatiotemporal feature representation of the detection system.

3. The Kalman filtering method for airtightness detection equipment according to claim 1, characterized in that: The step of automatically identifying and detecting the dynamic topology structure of the network by applying the edge functional verification algorithm comprises: Automatically discover the real physical connection relationship based on the pressure wave propagation characteristics; Introduce graph attention algorithm to adaptively adjust the influence weights between nodes; Construct a weighted adjacency matrix; Output a topological relationship matrix that is updated in real time.

4. A Kalman filtering method for airtightness detection equipment according to claim 1, characterized in that: The step of using the continuous-time neural differential equation model to represent the system state evolution comprises: Construct continuous-time differential equations for the evolution of the system state; Define the neural network structure, using a combination of graph convolution and multi-layer perceptron; The differential equations are solved by an adaptive numerical integration method to obtain the state representation in the continuous time domain.

5. The Kalman filtering method for airtightness detection equipment according to claim 1, characterized in that: The step of analyzing the historical topological change pattern using the time series prediction model includes: Build a spatiotemporal series prediction model; Design the prediction function structure and adopt the spatiotemporal attention network architecture; Adjust model parameters in advance based on predicted topological changes; Output the predicted topology and pre-tuned model parameters.

6. A Kalman filtering method for airtightness detection equipment according to claim 1, characterized in that: The step of applying the continuous deep Kalman filter algorithm to estimate and update the system state in real time includes: Use stochastic differential equations to represent system dynamics models; Design observation equations; Dynamically generate the Kalman filter state transfer matrix through the graph attention algorithm; Execute a continuous discrete extended Kalman filter algorithm, including a prediction step and an update step; Output the filtered system state estimation and anomaly detection results.

7. A Kalman filtering method for airtightness detection equipment according to claim 2, characterized in that: The graph convolutional network includes an input layer, an intermediate layer and an output layer, wherein: The input layer receives raw node features; The middle layer contains multiple hidden units and uses activation functions to enhance nonlinear expression capabilities; The output layer maps the features to the hidden state space.

8. A Kalman filtering method for airtightness detection equipment according to claim 3, characterized in that: The step of automatically discovering the real physical connection relationship based on the pressure wave propagation characteristics includes: Analyze the time-lag correlation of pressure signals between detection points; It is determined whether there is a physical connection between the detection points based on the correlation exceeding the preset threshold.

9. A Kalman filtering method for airtightness detection equipment according to claim 5, characterized in that: The spatiotemporal attention network includes: An encoder for encoding the graph structure into a hidden state vector; An attention module for calculating the importance weights of graph structures at different time points; Recurrent neural networks for processing sequence information; A decoder for decoding hidden states into a predicted graph structure.

10. An airtightness detection system suitable for a dynamic topology network, used to execute a Kalman filtering method for an airtightness detection device as described in any one of claims 1 to 9, characterized in that: include: Multi-detection point graph structure modeling module, used to model the airtightness detection system and generate the system topology representation; Dynamic topology recognition module, used to automatically identify and detect the dynamic topology structure of the network by applying edge functional verification algorithm; A continuous-time state representation module, which is used to represent the system state evolution using a continuous-time neural differential equation model; The topology prediction module is used to analyze the historical topology change pattern and predict the future topology evolution trajectory using the time series prediction model; The continuous deep Kalman filter module is used to estimate and update the system status in real time and output high-precision airtightness detection results.

Images