Gas pipeline leakage detection system and method
Through cloud data centers and advanced algorithm models, the high cost and low efficiency problems of traditional gas pipeline leakage detection methods are solved, high-precision and real-time leakage detection and positioning are achieved, and the intelligence and practicality of the detection system are improved.
Patent Information
- Application Number
- CN202410998487.0
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-07-24
- Publication Date
- 2024-11-12
- Estimated Expiration
- Not applicable · inactive patent
AI Technical Summary
Traditional gas pipeline leak detection methods have high maintenance costs, low efficiency, poor real-time and accuracy, and it is impossible to detect potential leakage risks and accurately locate the leakage points in the first time.
The leak detection device of the gas pipeline is communicated with the cloud data center, and the leakage detection and positioning is carried out using digital twin models and deep learning algorithms. Combined with the reinforcement learning algorithm optimization startup strategy, the leakage detection model and DQN algorithm of the CNN-GNN-BiLSTM-ICPO algorithm are constructed. Start the policy optimization model to realize real-time data processing and leakage point positioning.
It improves the management efficiency and real-time nature of leakage detection, reduces labor and material costs, improves the accuracy and accuracy of detection data, enhances intelligence and functionality, and ensures the accurate positioning and practicality of leakage points.
Smart Images

Figure CN118935266A_ABST
Abstract
Description
Technical Field
[0001] The present invention belongs to the technical field of gas leakage detection, and in particular relates to a gas pipeline leakage detection system and method. Background Art
[0002] With the widespread use of gas pipelines, pipeline leakage is becoming increasingly serious. Leakage not only leads to energy waste, but may also cause safety accidents. Traditional leak detection methods rely on manual inspections and simple sensor monitoring, which are costly to maintain, inefficient, and have the risk of missed detection. In addition, the real-time and accuracy of sensor-based monitoring are poor, and it is impossible to detect potential leakage risks in the first place and accurately locate the leakage point. Summary of the invention
[0003] In order to solve the problems of high maintenance cost, low efficiency, poor real-time performance and poor accuracy in the prior art, the present invention aims to provide a gas pipeline leakage detection system and method.
[0004] The technical solution adopted by the present invention is:
[0005] A gas pipeline leakage detection system comprises a cloud data center and a plurality of gas pipeline leakage detection devices, wherein the cloud data center is respectively communicatively connected with the plurality of gas pipeline leakage detection devices, and the plurality of gas pipeline leakage detection devices are arranged at a plurality of gas pipelines of a gas network in a one-to-one correspondence.
[0006] Furthermore, the gas pipeline leakage detection device includes a main control module, a negative pressure pump, a sensor, a detection air chamber, a detection pipeline, a plurality of three-way pipelines, a plurality of air filters and a plurality of one-way valves. The detection pipeline is arranged above the gas pipeline. The end of the detection pipeline adopts a sealing structure, and the head end of the detection pipeline is fixedly connected to the input end of the detection air chamber. The output end of the detection air chamber is provided with a negative pressure pump, and the interior of the detection air chamber is provided with a sensor. The output end of the sensor is electrically connected to the input end of the main control module, the output end of the main control module is electrically connected to the control end of the negative pressure pump, and the communication end of the main control module is communicatively connected to the cloud data center. A plurality of three-way pipelines are evenly connected in series to the detection pipeline, and the input end of each three-way pipeline is provided with a one-way valve and an air filter.
[0007] Furthermore, a filter membrane is arranged inside the air filter, and the filter membrane is made of PEFT hydrophobic small molecule membrane material.
[0008] Furthermore, the sensor is a laser methane sensor.
[0009] Furthermore, the cloud data center includes a data preprocessing unit, a model building unit, a leakage detection and positioning unit, and a startup strategy optimization unit which are connected in sequence, and the data preprocessing unit and the startup strategy optimization unit are respectively communicatively connected to a plurality of gas pipeline leakage detection devices.
[0010] Furthermore, the leakage detection and positioning unit is provided with a digital twin model of the gas pipeline network and a leakage detection and positioning model constructed based on a deep learning algorithm, and the startup strategy optimization unit is provided with a startup strategy optimization model constructed based on a reinforcement learning algorithm.
[0011] A gas pipeline leakage detection method, based on a gas pipeline leakage detection system, comprises the following steps:
[0012] Initialize the gas pipeline leakage detection system, build a gas pipeline leakage detection device at each gas pipeline in the gas pipeline network, and based on the preset startup strategy, use the gas pipeline leakage detection device to collect some historical leakage detection data and send it to the cloud data center;
[0013] Based on the cloud data center, a digital twin model is constructed according to the basic information data of the gas pipeline network. A leak detection positioning model and a startup strategy optimization model are constructed based on the digital twin model, several historical leak detection data, and preset startup strategies.
[0014] Based on the gas pipeline leakage detection device, based on the preset startup strategy, real-time leakage detection data of the corresponding position is collected, the number and real-time detection cycle of the gas pipeline leakage detection device are used as additional information, and the real-time leakage detection data and additional information are uploaded to the cloud data center;
[0015] Based on the cloud data center, the real-time leakage detection data is input into the digital twin model, and according to the additional information, several real-time leakage detection data are preprocessed and matrix converted to obtain the real-time leakage detection data matrix of the same real-time detection period;
[0016] According to the digital twin model and the real-time leak detection data matrix, the leak detection and positioning model is used to detect and locate the gas pipeline leak, and the real-time gas pipeline leak point positioning result is obtained. According to the real-time gas pipeline leak point positioning result, early warning visualization is performed on the digital twin model;
[0017] The startup strategy optimization model is used to optimize the startup strategy of the target gas pipeline leakage detection device near the real-time gas pipeline leakage point positioning result, and the real-time optimized startup strategy is obtained, and the real-time optimized startup strategy is sent to the corresponding target gas pipeline leakage detection device.
[0018] Furthermore, based on the cloud data center, a digital twin model is constructed according to the basic information data of the gas pipeline network, and a leak detection positioning model and a startup strategy optimization model are constructed according to the digital twin model, a number of historical leak detection data and a preset startup strategy, including the following steps:
[0019] Based on the cloud data center, a 3D simulation model and digital twin database of the gas pipeline network are constructed according to the basic information data of the gas pipeline network. The digital twin model is constructed by combining the 3D simulation model and the digital twin database, and some historical leakage detection data are stored in the digital twin database.
[0020] Based on the gas pipeline network topology information of the digital twin model and several historical leakage detection data stored in the digital twin database, a leakage detection positioning model is constructed using a deep learning algorithm, and several historical gas pipeline leakage point positioning results are generated;
[0021] Based on several historical gas pipeline leakage point location results and preset startup strategies, a startup strategy optimization model was constructed using the reinforcement learning algorithm.
[0022] Furthermore, the leakage detection and positioning model is constructed based on the CNN-GNN-BiLSTM-ICPO algorithm, and the leakage detection and positioning model includes a data feature extraction module constructed based on the CNN algorithm, a topological feature extraction module constructed based on the GNN algorithm, a leakage detection and positioning prediction module constructed based on the BiLSTM algorithm, and a positioning prediction optimization module constructed based on the ICPO algorithm.
[0023] Furthermore, the startup strategy optimization model is built based on the DQN algorithm.
[0024] The beneficial effects of the present invention are:
[0025] The present invention discloses a gas pipeline leakage detection system and method, which adopts a cloud data center to uniformly manage and analyze detection data, thereby improving management efficiency and real-time performance and reducing human and material cost investment; compared with traditional dispersed sensors, the gas pipeline leakage detection device is convenient to maintain, has a high degree of automation, and has high precision in collected data, and avoids the problem of poor detection accuracy caused by the volatilization of gas in the air; batch leakage detection is performed based on real-time leakage detection data, which has higher comprehensiveness and accuracy than simple threshold analysis, and improves the intelligence level and analysis efficiency of leakage detection, and locates the leakage point in combination with the topological information of the gas pipeline network, thereby improving practicality and functionality.
[0026] Other beneficial effects of the present invention will be further described in the specific embodiments. BRIEF DESCRIPTION OF THE DRAWINGS
[0027] Figure 1 It is a structural block diagram of the gas pipeline leakage detection system in the present invention.
[0028] Figure 2 It is a structural schematic diagram of the gas pipeline leakage detection device in the present invention.
[0029] Figure 3 It is a flow chart of the gas pipeline leakage detection method in the present invention.
[0030] In the figure, 1. main control module; 2. negative pressure pump; 3. sensor; 4. detection air chamber; 5. detection pipeline; 6. three-way pipeline; 7. air filter; 8. one-way valve. DETAILED DESCRIPTION
[0031] The present invention will be further explained below in conjunction with the accompanying drawings and specific embodiments.
[0032] Embodiment 1:
[0033] like Figure 1 As shown, this embodiment provides a gas pipeline leakage detection system, including a cloud data center and a plurality of gas pipeline leakage detection devices. The cloud data center is communicatively connected to the plurality of gas pipeline leakage detection devices, respectively, and the plurality of gas pipeline leakage detection devices are arranged one by one at a plurality of gas pipelines 9 of the gas network.
[0034] A gas pipeline leakage detection device is used to collect real-time leakage detection data of a corresponding position based on a preset startup strategy / real-time optimized startup strategy, use the number and real-time detection period of the gas pipeline leakage detection device as additional information, and upload the real-time leakage detection data and additional information to a cloud data center;
[0035] The cloud data center is used to build a digital twin model based on the basic information data of the gas pipeline network, and to build a leak detection positioning model and a startup strategy optimization model based on the digital twin model, several historical leak detection data, and preset startup strategies;
[0036] It is also used to input the real-time leakage detection data uploaded by the gas pipeline leakage detection device into the digital twin model, and pre-process and matrix-convert a number of real-time leakage detection data according to the additional information to obtain the real-time leakage detection data matrix of the same real-time detection period;
[0037] It is also used to detect and locate gas pipeline leaks using a leak detection and positioning model based on the digital twin model and the real-time leak detection data matrix, obtain real-time gas pipeline leak point positioning results, and perform early warning visualization on the digital twin model based on the real-time gas pipeline leak point positioning results;
[0038] It is also used to use the startup strategy optimization model to optimize the startup strategy of the target gas pipeline leakage detection device near the real-time gas pipeline leakage point positioning result, obtain the real-time optimized startup strategy, and send the real-time optimized startup strategy to the corresponding target gas pipeline leakage detection device.
[0039] As a preference, Figure 2 As shown, the gas pipeline leakage detection device includes a main control module 1, a negative pressure pump 2, a sensor 3, a detection air chamber 4, a detection pipeline 5, a plurality of three-way pipelines 6, a plurality of air filters 7 and a plurality of one-way valves 8. The detection pipeline 5 is arranged above the gas pipeline 9, the end of the detection pipeline 5 adopts a sealing structure, and the head end of the detection pipeline 5 is fixedly connected to the input end of the detection air chamber 4, the output end of the detection air chamber 4 is provided with a negative pressure pump 2, and the interior of the detection air chamber 4 is provided with a sensor 3, the output end of the sensor 3 is electrically connected to the input end of the main control module 1, the output end of the main control module 1 is electrically connected to the control end of the negative pressure pump 2, and the communication end of the main control module 1 is communicatively connected to the cloud data center, a plurality of three-way pipelines 6 are evenly connected in series to the detection pipeline 5, and a one-way valve 8 and an air filter 7 are provided at the input end of each three-way pipeline 6.
[0040] Based on the preset startup strategy / real-time optimized startup strategy, the main control module 1 controls the negative pressure pump 2 to evacuate air when the real-time detection cycle arrives, so that negative pressure is formed in the detection pipe 5, and the air at the input end of the three-way pipe 6 is sucked into the detection pipe 5, and enters the detection air chamber 4 through the detection pipe 5. The air filter 7 can separate methane gas, water and other gases, and increase the concentration of methane gas entering the detection air chamber 4. The one-way valve 8 prevents air leakage in the detection pipe 5. The sensor 3 is used to check the concentration of methane gas and generate digital real-time leakage detection data, which is uploaded to the cloud data center through the main control module 1. The main control module 1 is configured with wireless transmission channels such as WIFI modules or Bluetooth modules, which realizes remote transmission of real-time leakage detection data, preset startup strategies and real-time optimized startup strategies, avoids the laying of communication cables, and saves cost investment.
[0041] Preferably, a filter membrane is disposed inside the air filter 7, and the filter membrane is made of a hydrophobic small molecule membrane material of polytetrafluoroethylene (PTFE).
[0042] Preferably, the sensor 3 is a laser methane sensor.
[0043] Preferably, the cloud data center includes a data preprocessing unit, a model building unit, a leakage detection and positioning unit, and a startup strategy optimization unit which are connected in sequence, and the data preprocessing unit and the startup strategy optimization unit are respectively communicatively connected to a plurality of gas pipeline leakage detection devices.
[0044] Preferably, the leakage detection and positioning unit is provided with a digital twin model of the gas pipeline network and a leakage detection and positioning model constructed based on a deep learning algorithm, and the startup strategy optimization unit is provided with a startup strategy optimization model constructed based on a reinforcement learning algorithm.
[0045] Embodiment 2:
[0046] like Figure 3 As shown, this embodiment provides a gas pipeline leakage detection method, based on a gas pipeline leakage detection system, the method includes the following steps:
[0047] S1: Initialize the gas pipeline leakage detection system, build a gas pipeline leakage detection device at each gas pipeline in the gas network, and use the gas pipeline leakage detection device to collect some historical leakage detection data based on the preset startup strategy and send it to the cloud data center;
[0048] S2: Based on the cloud data center, a digital twin model is constructed according to the basic information data of the gas pipeline network, and a leak detection positioning model and a startup strategy optimization model are constructed according to the digital twin model, several historical leak detection data, and preset startup strategies, including the following steps:
[0049] S2-1: Based on the cloud data center and according to the basic information data of the gas pipeline network, a 3D simulation model and a digital twin database of the gas pipeline network are constructed. The digital twin model is constructed by combining the 3D simulation model and the digital twin database, and some historical leakage detection data are stored in the digital twin database;
[0050] The basic information data includes the Geographic Information System (GIS) data, Building Information Modeling (BIM) data, gas pipeline connection information, gas pipeline size information and gas pipeline material information of the gas pipeline network. A 3D geographic model of the gas pipeline network is established based on the GIS data. A 3D building model of the gas pipeline is constructed based on the BIM data, gas pipeline size information and gas pipeline material information. The 3D building model of the gas pipeline is converted into a 3D topological model of the gas pipeline network based on the gas pipeline connection information. The 3D topological model and the 3D geographic model are combined to obtain a 3D simulation model of the gas pipeline network. The 3D simulation model displays the topological information of the gas pipeline network, providing data support for subsequent leakage location. The data of the gas pipeline leakage detection device is digitally mapped and converted into a data stream, and a digital twin database is constructed to store the data stream.
[0051] S2-2: Based on the gas pipeline network topology information of the digital twin model and several historical leakage detection data stored in the digital twin database, a deep learning algorithm is used to build a leakage detection positioning model and generate several historical gas pipeline leakage point positioning results, including the following steps:
[0052] S2-2-1: performing data preprocessing and matrix conversion on a number of historical leakage detection data to obtain a number of historical leakage detection data matrices;
[0053] Data preprocessing includes data cleaning, data removal, and format unification, which improves data quality and provides data support for subsequent model training;
[0054] S2-2-2: Use the Convolutional Neural Network (CNN)-Graph Neural Network (GNN)-Bidirectional Long Short-Term Memory (BiLSTM)-Improved Crested Porcupine Optimizer (ICPO) algorithm to build an initial leak detection and positioning model. The leak detection and positioning model includes a data feature extraction module based on the CNN algorithm, a topological feature extraction module based on the GNN algorithm, a leak detection and positioning prediction module based on the BiLSTM algorithm, and a positioning prediction optimization module based on the ICPO algorithm.
[0055] The data feature extraction module is used to extract the data features of the leakage detection data matrix; through the deep neural network structure, it can mine the deep features in the data and improve the accuracy of the model prediction;
[0056] A topological feature extraction module is used to extract the topological features of the gas pipeline network topology information of the digital twin model; extract the edge features of the gas pipelines in the gas network, the point features of the pipeline connections, and the graph features of the gas pipeline network;
[0057] The leakage detection and positioning prediction module is used to perform leakage detection and positioning prediction based on data features and topological features to obtain leakage detection results and initial leakage point positioning results;
[0058] The positioning prediction optimization module is used to optimize the leakage point positioning result of the leakage detection result indicating that there is a leakage point, so as to obtain the optimal leakage point positioning result;
[0059] S2-2-3: Based on several historical leak detection data matrices and the gas pipeline network topology information of the digital twin model, the initial leak detection and positioning model is optimized and trained to obtain the optimal leak detection and positioning model, and several historical gas pipeline leak point positioning results are generated;
[0060] S2-3: Based on several historical gas pipeline leakage point location results and preset startup strategies, a startup strategy optimization model is constructed using a reinforcement learning algorithm, including the following steps:
[0061] S2-3-1: According to the startup strategy optimization problem, define the simulation environment of the Deep Q Network (DQN) algorithm, and build the DQN algorithm's deep Q network, agency, and experience storage pool. S2-3-2: According to the preset startup strategy, define the DQN algorithm's action space A = [a1,...,a j ,...,a I' ], based on the historical gas pipeline leakage location results, define the state space S = [s1,...,s i ,...,s I ], where s i is the i-th state value, i is the state indicator, I is the size of the state space, a j is the jth action value, j is the action indicator, and I' is the size of the action space;
[0062] S2-3-3: According to the action space A = [a1, ..., a j ,...,a I' ] and the state space S = [s1,...,s i ,...,s I ], define the reward function of the DQN algorithm;
[0063] S2-3-4: Build an initial startup strategy optimization model based on the simulation environment, state space, action space, reward function, deep Q network, agency, and experience storage pool;
[0064] The neural network structure is a deep Q network, which includes two neural networks with the same structure but different parameters: one is used to predict Q values (online Q network), and the other is used to generate target Q values (target Q network). The parameters of the target Q network are regularly copied from the online Q network, but are not updated frequently, which helps to improve the stability of the learning process.
[0065] S2-3-5: According to several historical gas pipeline leakage point location results and preset startup strategies, the initial startup strategy optimization model is optimized and trained, and the generated startup strategy optimization experience is stored in the experience storage pool to obtain the final startup strategy optimization model;
[0066] S3: Based on the gas pipeline leakage detection device and the preset startup strategy, real-time leakage detection data of the corresponding position is collected, the number and real-time detection cycle of the gas pipeline leakage detection device are used as additional information, and the real-time leakage detection data and additional information are uploaded to the cloud data center;
[0067] The serial number is used to construct the position of the real-time leakage detection data in the real-time leakage detection data matrix, and the real-time detection period is used to ensure the real-time nature of the data when constructing the real-time leakage detection data matrix;
[0068] S4: Based on the cloud data center, the real-time leakage detection data is input into the digital twin model, and according to the additional information, a number of real-time leakage detection data are preprocessed and matrix converted to obtain the real-time leakage detection data matrix of the same real-time detection period;
[0069] The real-time leak detection data matrix is used to uniformly analyze several detection points of the gas pipeline in the gas pipeline network. Compared with the traditional threshold analysis method, the analysis based on the real-time leak detection data matrix can obtain the location information and diffusion information of the gas pipeline network leakage point;
[0070] S5: Based on the digital twin model and the real-time leak detection data matrix, the leak detection and positioning model is used to perform gas pipeline leak detection and positioning, and the real-time gas pipeline leak point positioning result is obtained. Based on the real-time gas pipeline leak point positioning result, early warning visualization is performed on the digital twin model, including the following steps:
[0071] S5-1: using the data feature extraction module of the leakage detection positioning model to extract real-time data features of the real-time leakage detection data matrix;
[0072] S5-2: Use the topological feature extraction module to extract the topological features of the gas pipeline network topology information of the digital twin model;
[0073] S5-3: Use the leakage detection and positioning prediction module to perform leakage detection and positioning prediction according to the real-time data characteristics and topological characteristics, and obtain the real-time leakage detection results and the initial real-time leakage point positioning results;
[0074] The real-time leakage detection results include the detection point detection data information represented by the real-time leakage detection data matrix. When the preset threshold is exceeded, the abnormal state is represented. If there is also an abnormal state near the detection point, the leakage point prediction result can be obtained based on the data characteristics. At the same time, combined with the topological characteristics, the real-time leakage point positioning result can be obtained, and the probability distribution of the leakage point in the gas pipeline in the gas pipeline network can be located;
[0075] S5-4: Using the positioning prediction optimization module, the initial real-time leakage point positioning result whose real-time leakage detection result is that there is a leakage point is optimized to obtain the optimal leakage point positioning result, including the following steps:
[0076] S5-4-1: splitting the real-time probability distribution of the initial real-time leakage point location result into a number of candidate real-time leakage point location positions;
[0077] S5-4-2: Encode the candidate real-time leak point location as an ICPO individual of the ICPO optimization algorithm;
[0078] S5-4-3: Set ICPO population parameters, maximum number of iterations and fitness function of ICPO optimization algorithm;
[0079] S5-4-4: According to the number of individuals in the ICPO population parameters, the Circle chaotic mapping sequence is used for initialization to obtain the initial ICPO population;
[0080] The formula is:
[0081]
[0082] In the formula, It is the initial ICPO individual of Circle chaos mapping; is the randomly generated initial ICPO individual; i' is the ICPO individual indicator;
[0083] S5-4-5: Introduce a cyclic population reduction mechanism to limit the number of individuals in the ICPO population parameters and obtain updated ICPO population parameters for the next iteration;
[0084] The formula is:
[0085]
[0086] In the formula, S t+1 is the number of individuals in the ICPO population parameter of the t+1th iteration; S t is the number of individuals in the ICPO population parameter of the tth iteration; S min is the minimum number of individuals in the ICPO population parameter; a' is the function evaluation parameter; V is the function evaluation cycle parameter; V max Evaluate loop parameters for the maximum function;
[0087] S5-4-6: Calculate the initial fitness value of the initial ICPO individuals in the initial ICPO population according to the fitness function;
[0088] The formula is:
[0089]
[0090] In the formula, f(x i' ) is the ICPO individual x i' The fitness value of For ICPO individual x i' The distance deviation function of For ICPO individual x i' The operating cost function of For ICPO individual x i' The time deviation function of are weight coefficients; ν is the compensation coefficient;
[0091] S5-4-7: According to the initial fitness value and the updated ICPO population parameters, the first defense strategy, the second defense strategy, the third defense strategy and the fourth defense strategy are used to update the initial ICPO population to obtain an updated ICPO population;
[0092] The formula for the first defense strategy is:
[0093]
[0094] In the formula, For the updated ICPO individuals within the first defense range; is the initial ICPO individual within the first defense range; τ1 is a random number based on normal distribution; τ2 is a random value in the interval [0,1]; It is the optimal solution within the first defense range; is the vector generated between the true optimal solution within the first defense range and the optimal solution randomly selected from the ICPO population; i' is the ICPO individual indicator; t is the iteration indicator;
[0095] The formula for the second defense strategy is:
[0096]
[0097] In the formula, For the updated ICPO individuals within the second defense range; The initial ICPO individual within the second defense range; is the search upper limit vector of the second defense range; τ3 is a random value in the interval [0,1]; are the r1th and r2th initial ICPO individuals respectively; r1 and r2 are two random integers between [1, S]; is the vector generated between the true optimal solution within the second defense range and the optimal solution randomly selected from the ICPO population;
[0098] The formula for the third defense strategy is:
[0099]
[0100] In the formula, For the updated ICPO individuals within the third defense range; The initial ICPO individual within the third defense range; is the search upper limit vector of the third defense range; are the r2th and r3th initial ICPO individuals respectively; r3 is a random integer between [1, S]; The odor diffusion factor defined for the fitness function; λ t It is a defense factor; Control parameters for search direction;
[0101] The formula for the fourth defense strategy is:
[0102]
[0103] In the formula, For the updated ICPO individuals within the fourth defense range; The initial ICPO individual within the fourth defense range; is the optimal solution within the fourth defense range; τ4 and τ5 are both random values in the interval [0,1]; λ t It is a defense factor; Control parameters for search direction; is the average force affecting the search direction; a' is the convergence speed factor;
[0104] S5-4-8: Use the dynamic reverse learning algorithm to perform dynamic reverse learning on the updated ICPO population to generate a dynamic reverse ICPO population;
[0105] The formula is:
[0106]
[0107] In the formula, is the ICPO individual in dynamic reverse direction; γ is the decreasing inertia coefficient; L max , L min are the maximum and minimum values of the vector space respectively; For the updated ICPO individual;
[0108] S5-4-9: According to the fitness function, calculate the fitness values of all ICPO individuals in the updated ICPO population and the dynamically reversed ICPO population, and take the ICPO individual with the minimum fitness value as the optimal individual;
[0109] S5-4-10: If the number of iterations reaches the maximum number of iterations or the fitness value of the optimal individual meets the requirements, the optimal solution corresponding to the current optimal individual is output to obtain the optimal leakage point location result;
[0110] S6: Using the startup strategy optimization model, optimizing the startup strategy of the target gas pipeline leakage detection device near the real-time gas pipeline leakage point positioning result, obtaining the real-time optimized startup strategy, and sending the real-time optimized startup strategy to the corresponding target gas pipeline leakage detection device, including the following steps:
[0111] S6-1: According to the real-time gas pipeline leakage point location results, the state space of the startup strategy optimization model is updated to obtain the updated state space S'=[s'1,...,s' i" ,...,s' I ], where s' i" is the updated i'th state value, i" is the state indicator;
[0112] S6-2: According to the preset startup strategy of the nearby target gas pipeline leakage detection device, update the action space A'=[a'1,...,a' j' ,...,a' I ], where a' j' is the updated j'th action value, j' is the action indicator;
[0113] The preset startup strategy includes the startup time, startup cycle, startup frequency of the negative pressure pump and sensor, and the transmission time, transmission cycle and transmission frequency of the main control module upload time;
[0114] S6-3: Update the state space S' = [s'1, ..., s' i' ,...,s' I ] as the input of the starting policy optimization model, based on the deep Q network, using the agency to obtain the updated action space A'=[a'1,...,a' j' ,...,a' I ];
[0115] S6-4: Use the reward function to obtain the reward value of the possible action, and update the predicted Q value of the possible action according to the reward value to obtain the updated Q value of the possible action;
[0116] The formula is:
[0117] Q(s' p' ,a' p' )=(1-α)·Q(s p' ,a p' )+α·(R(sp' ,a p' ,s' p' )+γ·Q max (s p' ,a p' ))
[0118] In the formula, Q(s' p' ,a' p' ) is the updated state value s' p' and the updated action value a' p' The corresponding updated Q value; Q(s p' ,a p' ) is the state value s p' and action value a p' The corresponding predicted Q value; α is the learning rate; Q max (s p' ,a p' ) is the highest predicted Q value;
[0119] S6-5: Repeat the above steps until the iteration number threshold is reached, and use the greedy strategy to take the possible action corresponding to the highest updated Q value as the execution action;
[0120] The execution action includes an action to optimize the preset startup strategy, for example, shortening the startup time, increasing the startup frequency, etc.;
[0121] S6-6: Optimize the real-time startup strategy according to the execution action to obtain the real-time optimized startup strategy, and send the real-time optimized startup strategy to the corresponding target gas pipeline leakage detection device.
[0122] The present invention discloses a gas pipeline leakage detection system and method, which adopts a cloud data center to uniformly manage and analyze detection data, thereby improving management efficiency and real-time performance and reducing human and material cost investment; compared with traditional dispersed sensors, the gas pipeline leakage detection device is convenient to maintain, has a high degree of automation, and has high precision in collected data, and avoids the problem of poor detection accuracy caused by the volatilization of gas in the air; batch leakage detection is performed based on real-time leakage detection data, which has higher comprehensiveness and accuracy than simple threshold analysis, and improves the intelligence level and analysis efficiency of leakage detection, and locates the leakage point in combination with the topological information of the gas pipeline network, thereby improving practicality and functionality.
[0123] The present invention is not limited to the above optional implementations, and anyone can derive other various forms of products under the enlightenment of the present invention. The above specific implementations should not be understood as limiting the scope of protection of the present invention. The scope of protection of the present invention should be based on the definition in the claims, and the description can be used to interpret the claims.
Claims
1. A gas pipeline leak detection system, Features: It comprises a cloud data center and a plurality of gas pipeline leakage detection devices, wherein the cloud data center is respectively connected to the plurality of gas pipeline leakage detection devices for communication, and the plurality of gas pipeline leakage detection devices are arranged one by one at a plurality of gas pipelines (9) of a gas network.
2. A gas pipeline leakage detection system according to claim 1, Features: The gas pipeline leakage detection device comprises a main control module (1), a negative pressure pump (2), a sensor (3), a detection air chamber (4), a detection pipeline (5), a plurality of three-way pipelines (6), a plurality of air filters (7) and a plurality of one-way valves (8). The detection pipeline (5) is arranged above the gas pipeline (9). The end of the detection pipeline (5) adopts a sealing structure, and the head end of the detection pipeline (5) is fixedly connected to the input end of the detection air chamber (4). The output end of the detection air chamber (4) is provided with a negative pressure pump (2), and the interior of the detection air chamber (4) is provided with a sensor (3). The output end of the sensor (3) is electrically connected to the input end of the main control module (1), the output end of the main control module (1) is electrically connected to the control end of the negative pressure pump (2), and the communication end of the main control module (1) is communicatively connected to a cloud data center. The plurality of three-way pipelines (6) are evenly connected in series to the detection pipeline (5), and the input end of each three-way pipeline (6) is provided with a one-way valve (8) and an air filter (7).
3. A gas pipeline leakage detection system according to claim 2, Features: The air filter (7) is provided with a filter membrane inside, and the filter membrane is made of PEFT hydrophobic small molecule membrane material.
4. A gas pipeline leakage detection system according to claim 2, Features: The sensor (3) is a laser methane sensor.
5. A gas pipeline leakage detection system according to claim 1, Features: The cloud data center includes a data preprocessing unit, a model building unit, a leakage detection and positioning unit, and a startup strategy optimization unit which are connected in sequence. The data preprocessing unit and the startup strategy optimization unit are respectively connected to a plurality of gas pipeline leakage detection devices for communication.
6. A gas pipeline leakage detection system according to claim 5, Features: The leakage detection and positioning unit is provided with a digital twin model of the gas pipeline network and a leakage detection and positioning model constructed based on a deep learning algorithm, and the startup strategy optimization unit is provided with a startup strategy optimization model constructed based on a reinforcement learning algorithm.
7. A gas pipeline leakage detection method, based on the gas pipeline leakage detection system according to any one of claims 1 to 6, Features: The method comprises the following steps: Initialize the gas pipeline leakage detection system, build a gas pipeline leakage detection device at each gas pipeline in the gas pipeline network, and based on the preset startup strategy, use the gas pipeline leakage detection device to collect some historical leakage detection data and send it to the cloud data center; Based on the cloud data center, a digital twin model is constructed according to the basic information data of the gas pipeline network. A leak detection positioning model and a startup strategy optimization model are constructed based on the digital twin model, several historical leak detection data, and preset startup strategies. Based on the gas pipeline leakage detection device, based on the preset startup strategy, real-time leakage detection data of the corresponding position is collected, the number and real-time detection cycle of the gas pipeline leakage detection device are used as additional information, and the real-time leakage detection data and additional information are uploaded to the cloud data center; Based on the cloud data center, the real-time leakage detection data is input into the digital twin model, and according to the additional information, several real-time leakage detection data are preprocessed and matrix converted to obtain the real-time leakage detection data matrix of the same real-time detection period; According to the digital twin model and the real-time leak detection data matrix, the leak detection and positioning model is used to detect and locate the gas pipeline leak, and the real-time gas pipeline leak point positioning result is obtained. According to the real-time gas pipeline leak point positioning result, early warning visualization is performed on the digital twin model; The startup strategy optimization model is used to optimize the startup strategy of the target gas pipeline leakage detection device near the real-time gas pipeline leakage point positioning result, and the real-time optimized startup strategy is obtained, and the real-time optimized startup strategy is sent to the corresponding target gas pipeline leakage detection device.
8. A gas pipeline leakage detection method according to claim 7, Features: Based on the cloud data center, a digital twin model is constructed according to the basic information data of the gas pipeline network, and a leak detection positioning model and a startup strategy optimization model are constructed according to the digital twin model, several historical leak detection data and preset startup strategies, including the following steps: Based on the cloud data center, a 3D simulation model and digital twin database of the gas pipeline network are constructed according to the basic information data of the gas pipeline network. The digital twin model is constructed by combining the 3D simulation model and the digital twin database, and some historical leakage detection data are stored in the digital twin database. Based on the gas pipeline network topology information of the digital twin model and several historical leakage detection data stored in the digital twin database, a leakage detection positioning model is constructed using a deep learning algorithm, and several historical gas pipeline leakage point positioning results are generated; Based on several historical gas pipeline leakage point location results and preset startup strategies, a startup strategy optimization model was constructed using the reinforcement learning algorithm.
9. A gas pipeline leakage detection method according to claim 8, Features: The leakage detection and positioning model is constructed based on the CNN-GNN-BiLSTM-ICPO algorithm, and the leakage detection and positioning model includes a data feature extraction module constructed based on the CNN algorithm, a topological feature extraction module constructed based on the GNN algorithm, a leakage detection and positioning prediction module constructed based on the BiLSTM algorithm, and a positioning prediction optimization module constructed based on the ICPO algorithm.
10. A gas pipeline leakage detection method according to claim 8, Features: The startup strategy optimization model is constructed based on the DQN algorithm.
Citation Information
Cited By
Sweeping robot control system based on reinforcement learning
CN120233780A
Gas pipeline safety monitoring and management method, device, equipment, medium and product
CN120430897A
