Large-range multi-point synchronous detection system for leakage of sulfur-containing natural gas pipeline

Abstract

The invention relates to the field of natural gas pipeline detection, in particular to a large-range multi-point synchronous detection system for sulfur-containing natural gas pipeline leakage. A large-range multi-point synchronous detection system for sulfur-containing natural gas pipeline leakage comprises a main control computer and a plurality of hydrogen sulfide synchronous detectors connected with the main control computer, and each hydrogen sulfide synchronous detector comprises an acquisition detection module, a positioning transmission module, a control module, a display module and a power supply module which are connected in sequence. The method has the beneficial effects that multi-point synchronous detection in a large-range space around sulfur-containing natural gas pipeline leakage can be realized, the influence of external weather factors on a detection result is reduced, and a spatial distribution rule of hydrogen sulfide highly toxic gas concentration is accurately obtained; and secondly, the hydrogen sulfide concentration in a large range can be monitored through deployment of the unmanned aerial vehicle, the influence of external weather factors on a detection result is reduced, personnel can be prevented from determining a leakage area of the hydrogen sulfide concentration, and the safety of on-site emergency disposal personnel is ensured.

Description

Technical Field

[0001] The invention relates to the field of natural gas pipeline detection, and in particular to a large-scale multi-point synchronous detection system for leakage of sulfur-containing natural gas pipelines. Background Art

[0002] my country is rich in high-sulfur natural gas resources and has great potential for exploitation. With the continuous development of the economy and society, accelerating the development and utilization of domestic high-sulfur gas reservoirs is of great significance to promoting the development of the national economy, reducing environmental pollution and ensuring national energy security. Sour natural gas pipelines contain a large amount of hydrogen sulfide, which is highly toxic and dangerous. Once a leak occurs, it will pose a serious threat to the lives and property of surrounding personnel. In order to reduce the potential risks brought by sour natural gas leaks and provide reliable information for emergency disposal such as pipeline repairs, it is necessary to monitor the hydrogen sulfide concentration in a large area around the pipeline leak location in real time. At present, when a sour natural gas pipeline leaks, the staff can use a portable gas concentration detection device to monitor the hydrogen sulfide concentration around the leak point and record the changes in the hydrogen sulfide concentration. Hydrogen sulfide concentration detection sensors can also be arranged at some key locations to measure the hydrogen sulfide concentration within the monitoring range. However, since the sensor is arranged at a single point, only the hydrogen sulfide concentration at a certain location can be obtained, and the regional concentration distribution of hydrogen sulfide cannot be accurately obtained.

[0003] When a sour natural gas pipeline leaks, existing technologies require workers to enter the leakage area to detect hydrogen sulfide concentration, which poses a risk of human poisoning. In addition, portable and fixed sensors can only detect hydrogen sulfide concentrations within a limited range, and are unable to perceive the diffusion law of hydrogen sulfide within the regional range, and are unable to provide a scientific basis for a large-scale multi-point synchronous detection system for sour natural gas pipeline leaks. Summary of the invention

[0004] In order to solve the above problems, the present invention provides a large-scale multi-point synchronous detection system for sour natural gas pipeline leakage. The specific scheme is as follows: The detection system comprises a main control machine and a plurality of hydrogen sulfide synchronous detectors connected to the main control machine. The hydrogen sulfide synchronous detectors comprise a collection and detection module, a positioning and transmission module, a control module, a display module and a power supply module which are connected in sequence.

[0005] The main control machine communicates with the hydrogen sulfide concentration synchronous detector through wireless signals. The hydrogen sulfide concentration synchronous detectors at different deployment locations are also wirelessly connected through wireless transmission modules to establish a wireless communication network. The hydrogen sulfide concentration synchronous detector at any location can transmit data with the host through the wireless communication network to prevent data loss caused by the distance between a hydrogen sulfide concentration synchronous detector and the main control machine; deploying multiple hydrogen sulfide synchronous detectors at different points can detect the hydrogen sulfide concentration at multiple points at the same time and accurately obtain the regional concentration distribution of hydrogen sulfide.

[0006] Furthermore, the collection and detection module is composed of a sampling port, a sampling tube, and a vacuum sampler connected in sequence from top to bottom; the positioning and transmission module includes a satellite positioning module and a wireless transmission module; and the control module is a single-chip microcomputer controller.

[0007] The hydrogen sulfide concentration synchronous detector is deployed on site manually. The sampling port is installed on the sampling tube to ensure that the sampling port has a certain installation height, which is conducive to improving the reliability of the hydrogen sulfide concentration detection results. The single-chip microcomputer controller controls the functions of the system. When multiple systems are deployed at multiple points, the wireless transmission module transmits data between different deployment locations. The positioning transmission module obtains the precise deployment location coordinates of the hydrogen sulfide concentration synchronous detector. The power module supplies power to the system.

[0008] Furthermore, the vacuum sampler is composed of a sampling air inlet pipe, a vacuum sampling tank, and a drain pipe connected in sequence; the sampling air inlet pipe is provided with a sampling air inlet automatic valve; one end of the sampling air inlet pipe is connected to the sampling port, and the other end is connected to the vacuum sampling tank; the drain pipe is provided with an automatic drain valve; the device also includes a nitrogen tank connected to the vacuum sampling tank through a nitrogen air inlet pipe and a nitrogen air inlet automatic valve arranged on the nitrogen air inlet pipe; the device also includes a miniature vacuum pump, the miniature vacuum pump is connected to the vacuum sampling tank through an exhaust pipe, and the exhaust pipe is provided with an exhaust automatic valve; the device also includes a hydrogen sulfide sensor, a temperature sensor and a pressure sensor arranged inside the vacuum sampling tank.

[0009] When the vacuum sampler is sampling, the nitrogen inlet automatic valve, the drain automatic valve and the sampling inlet automatic valve are closed, the exhaust automatic valve is opened, and the micro vacuum pump works to evacuate the vacuum sampling tank. When the vacuum degree measured by the pressure sensor reaches the preset value, the exhaust valve is closed, the micro vacuum pump stops working, and the sampling inlet automatic valve opens quickly. When the pressure measured by the pressure sensor returns to atmospheric pressure and the temperature measured by the temperature sensor stabilizes, the measurement results of the hydrogen sulfide sensor are recorded to complete the hydrogen sulfide concentration measurement.

[0010] After every 10 measurements, open the sampling air inlet automatic valve and the drainage automatic valve for 60 seconds to drain the accumulated liquid in the vacuum sampling tank. After the accumulated liquid is drained, close the sampling air inlet automatic valve, open the exhaust automatic valve, start the micro vacuum pump, and when the vacuum degree measured by the pressure sensor reaches the preset value, close the exhaust automatic valve, open the nitrogen air inlet automatic valve, and let nitrogen enter the vacuum sampling tank for purging.

[0011] Furthermore, it also includes a bracket located at the bottom of the vacuum sampling tank, and the hydrogen sulfide sensor is placed on the bracket.

[0012] A bracket is provided to support and fix the hydrogen sulfide sensor so that the hydrogen sulfide sensor has a certain height in the vacuum sampler, thereby ensuring the accuracy and stability of detection and preventing the accumulation of liquid from affecting the hydrogen sulfide sensor.

[0013] Furthermore, it also includes a control box, in which the positioning and transmission module, the control module, and the power supply module are located; the acquisition and detection module is fixedly connected to the outside of the control box; and the positioning and transmission module also includes an antenna.

[0014] The positioning and transmission module and the control module are located in the control box, namely the satellite positioning module and the wireless transmission module; the single-chip microcomputer controller is located in the control box, and the collection and detection module, namely the sampling port, the sampling tube, and the vacuum sampler are located outside the control box; the vacuum collector is fixed on the outer surface of the control box; the positioning and transmission module also includes an antenna, which is fixed on the outer surface of the control box and is wirelessly connected to the positioning and transmission module, and the antenna enhances the data transmission capability of the wireless module.

[0015] Furthermore, it also includes a balancing module, which is arranged in a control box and consists of an acceleration sensor, a buffer base and a buffer spring fixed on the buffer base; the acceleration sensor is connected to the positioning transmission module; the control box is semi-closed, and the buffer base is buckled on the bottom surface of the control box.

[0016] When the hydrogen sulfide concentration synchronous detector lands on the ground, the buffer base contacts the ground first, and the buffer spring is compressed to absorb part of the impact force for buffering, thereby reducing the damage to the structural stability of the hydrogen sulfide concentration synchronous detector caused by the impact of landing.

[0017] The acceleration sensor is used to sense the movement state and posture of the hydrogen sulfide concentration synchronous detector. If the hydrogen sulfide concentration synchronous detector falls over, an alarm will be issued to assist in identifying abnormal hydrogen sulfide concentration detection results caused by the device tipping over and the hydrogen sulfide sampling port being close to the ground.

[0018] Furthermore, it also includes a partition arranged in the control box, which divides the area composed of the control box buffer base into two layers, the positioning and transmission module, the control module, the acceleration sensor, and the power module are located in the upper area of ​​the partition; the power module is a battery; and the buffer spring is located in the lower area.

[0019] The partition is used to distinguish each electric module from the buffer spring and the buffer base.

[0020] Furthermore, the display module is an indication alarm device, which is fixed on the outer surface of the control box.

[0021] The indicating alarm device displays the concentration level of hydrogen sulfide through different colors and sounds.

[0022] Furthermore, the control box is provided with a lifting ring and also includes a drone, and the drone is connected to the lifting ring through a sling to lift the hydrogen sulfide synchronous detector.

[0023] The hydrogen sulfide concentration synchronous detector can be deployed manually on site or remotely by drone via a lifting ring. When a sour natural gas pipeline leaks, the existing technology requires workers to enter the leak area to detect hydrogen sulfide concentration, which poses a risk of poisoning. Therefore, the hydrogen sulfide concentration synchronous detector can be deployed in the leak area via drone to prevent poisoning.

[0024] Furthermore, a method for predicting the diffusion law of hydrogen sulfide is provided, and the prediction method is as follows: (1) Data collection and preprocessing: Collect hydrogen sulfide gas concentration data, including time, location, and concentration information, clean the data, remove invalid or abnormal data, and normalize the data so that it falls within an appropriate numerical range; (2) Feature selection: According to the changing rules and influencing factors of gas concentration, select the time features, spatial features and environmental features related to the prediction, and further process and transform the features; (3) Neural network structure design: Construct an LSTM neural network model, including an input layer, an LSTM layer, a fully connected layer, and an output layer; the input layer receives the preprocessed data features; the LSTM layer is responsible for capturing the long-term dependencies in the time series data; the fully connected layer maps the output of the LSTM layer to the predicted value; the output layer outputs the predicted hydrogen sulfide concentration; (4) Data partitioning: Divide the data set into training set, validation set and test set. The training set is used to train the model, the validation set is used to adjust hyperparameters and monitor model performance, and the test set is used to evaluate the final performance of the model; (5) Model training: Use the training set data to train the neural network and update the model parameters through the back-propagation algorithm; monitor the performance on the validation set and select the optimal model parameters and structure; (6) Hyperparameter tuning: Hyperparameters can be optimized by grid search, random search or Bayesian optimization. Perform tuning to evaluate the model performance under different hyperparameter combinations and select the optimal hyperparameter combination.

[0025] The beneficial effects of the present invention are: The test results are displayed on the main control machine in real time and uploaded to the central control through the network to guide the implementation of emergency response measures. It can realize multi-point synchronous detection in a large range of space around the leakage of sour natural gas pipelines, reduce the impact of external weather factors on the test results, and accurately obtain the spatial distribution law of the concentration of highly toxic hydrogen sulfide gas; secondly, through the deployment of drones, the hydrogen sulfide concentration in a large area can be monitored to reduce the impact of external weather factors on the test results, and personnel can be avoided from leaking areas where the hydrogen sulfide concentration is not yet clear, ensuring the safety of on-site emergency response personnel; secondly, the impact force during the deployment of drones is prevented from damaging the structural stability of the hydrogen sulfide concentration synchronous detector, and the posture of the deployed hydrogen sulfide concentration synchronous detector is sensed by the acceleration sensor to ensure the reliability of the test results. BRIEF DESCRIPTION OF THE DRAWINGS

[0026] Figure 1 The multi-point synchronous detection system of embodiment 1; Figure 2 This is a structural diagram of a synchronous detector for hydrogen sulfide concentration in Example 1; Figure 3 This is a diagram of the internal structure of the vacuum sampler of Example 1.

[0027] Figure 4 This is a diagram of the device for deploying a synchronous hydrogen sulfide concentration detector on a drone in Example 2.

[0028] Reference numerals: 1—main control machine, 2—hydrogen sulfide synchronous detector, 3—sling, 4—drone, 201—sampling port, 202—sampling tube, 203—vacuum sampler, 204—satellite positioning module, 205—wireless transmission module, 206—acceleration sensor, 207—buffer base, 208—buffer spring, 209—partition, 210—single chip microcomputer controller, 211—battery, 212—indicator alarm device, 213—antenna, 214—control box, 215—hanging ring, 203-1—sampling air intake automatic valve, 203-2—sampling air intake pipe, 203-3—nitrogen air intake pipe, 203-4—nitrogen air intake automatic valve, 203-5—nitrogen tank, 203-6—hydrogen sulfide sensor, 203-7—bracket, 203-8—drain pipe, 203-9—drain automatic valve, 203-10—exhaust pipe, 203-11—exhaust automatic valve, 203-12—micro vacuum pump, 203-13—vacuum sampling tank, 203-14—temperature sensor, 203-15—pressure sensor. DETAILED DESCRIPTION

[0029] Example 1 A large-scale multi-point synchronous detection system for sulfur-containing natural gas pipeline leakage, the detection system comprises a main control machine 1 and five identical hydrogen sulfide synchronous detectors 2 connected to the main control machine 1, the hydrogen sulfide synchronous detectors 2 comprise a collection and detection module, a positioning and transmission module, a control module, a display module, and a power supply module connected in sequence.

[0030] The acquisition and detection module is composed of a sampling port 201, a sampling tube 202, and a vacuum sampler 203 connected in sequence from top to bottom; the positioning and transmission module includes a satellite positioning module 204 and a wireless transmission module 205; the control module is a single-chip microcomputer controller 210; the display module is an indication alarm device 212; and the power module is a battery 211.

[0031] The vacuum sampler 203 is composed of a sampling air inlet pipe 203-2, a vacuum sampling tank 203-13, and a drain pipe 203-8 connected in sequence. The sampling air inlet pipe 203-2 is provided with a sampling air inlet automatic valve 203-1. One end of the sampling air inlet pipe 203-2 is connected to the sampling port 201, and the other end is connected to the vacuum sampling tank 203-13. The drain pipe 203-8 is provided with an automatic drain valve 203-9, and the other end of the drain pipe 203-8 is communicated with the outside of the vacuum sampler 203; it also includes a nitrogen tank 203-5 connected to the vacuum sampling tank 203-13 through a nitrogen inlet pipe 203-3 and a nitrogen inlet automatic valve 203-4 arranged on the nitrogen inlet pipe 203-3; it also includes a miniature vacuum pump 203-12, and the miniature vacuum pump 203-12 is connected to the vacuum sampling tank 203-13 through an exhaust pipe 203-10, and the exhaust pipe 203-10 is provided with an automatic exhaust valve 203-11; it also includes a hydrogen sulfide sensor 203-6, a temperature sensor 203-14 and a pressure sensor 203-15 arranged inside the vacuum sampling tank 203-13.

[0032] It also includes a bracket 203-7 located at the bottom of the vacuum sampling tank 203-13, and the hydrogen sulfide sensor 203-6 is placed and fixed on the bracket 203-7.

[0033] It also includes a semi-enclosed control box 214, and the satellite positioning module 204 and the wireless transmission module 205 are located in the control box 214; the collection and detection module, namely the sampling port 201, the sampling tube 202, and the vacuum sampler 203 are located outside the control box 214, and the vacuum sampler 203 is fixed on the upper surface of the control box 214; the positioning and transmission module also includes an antenna 213, and the antenna 213 is vertically fixed on the outer surface of the control box 214 and is infinitely connected to the positioning and transmission module, and the indication alarm device 212 is located on the outer upper surface of the control box 214.

[0034] It also includes a balancing module, which is arranged in the control box 214 and consists of an acceleration sensor 206, a buffer base 207, and a buffer spring 208 fixedly arranged on the buffer base 207; the acceleration sensor 206 is connected to the positioning transmission module; The buffer base 207 is buckled on the bottom surface of the control box 214, serving as the bottom surface of the entire hydrogen sulfide synchronous detector 2, and also includes a partition 209 arranged in the control box 214. The partition 209 divides the area formed by the control box buffer base into two layers, the upper and lower layers. The positioning and transmission module, the control module, the acceleration sensor 206, and the power module are located in the upper area of ​​the partition 209; the power module is a battery 211; and the buffer spring 208 is located in the lower area.

[0035] When it is necessary to detect the pipeline leakage area, 5 hydrogen sulfide synchronous detectors 2 are placed in the leakage area, and the sampling detection is started through the control of the main control machine 1. When the vacuum sampler 203 is sampling, the nitrogen air intake automatic valve 203-4, the drainage automatic valve 203-9 and the sampling air intake automatic valve 203-1 are closed, the exhaust automatic valve 203-11 is opened, and the micro vacuum pump 203-12 works to evacuate the vacuum sampling tank 203-13. When the vacuum degree measured by the pressure sensor 203-15 reaches the preset value, the exhaust automatic valve 203-11 is closed, the micro vacuum pump 203-12 stops working, and the sampling air intake automatic valve 203-1 is quickly opened. When the pressure measured by the pressure sensor 203-15 returns to atmospheric pressure and the temperature measured by the temperature sensor 203-14 is stable, the measurement result of the hydrogen sulfide sensor 203-6 is recorded to complete the hydrogen sulfide concentration measurement.

[0036] After every 10 measurements, open the sampling air inlet automatic valve 203-1 and the drainage automatic valve 203-9 for 60 seconds to discharge the accumulated liquid in the vacuum sampling tank 203-13; after the accumulated liquid is discharged, close the sampling air inlet automatic valve 203-1, open the exhaust automatic valve 203-11, start the micro vacuum pump 203-12, and when the vacuum degree measured by the pressure sensor 203-15 reaches the preset value, the exhaust automatic valve 203-11 is closed, and the nitrogen air inlet automatic valve 203-4 is opened, and nitrogen enters the vacuum sampling tank 203-13 for purging.

[0037] After the detection technology is collected, the detection results are transmitted to the main control computer 1.

[0038] Example 2 Different from the first embodiment, the control box 214 is further provided with a lifting ring 215 , and also includes a drone 4 , and the drone 4 is connected to the lifting ring 215 via a sling 3 to lift the hydrogen sulfide synchronous detector 2 .

[0039] Because the detection area is a sulfide leak, in order to avoid poisoning of personnel, remote deployment is carried out using drones through hanging rings.

[0040] Example 3 The detection results obtained in Example 2 are displayed by the main control computer. The diffusion law of hydrogen sulfide can be predicted based on the collected data. The prediction method is as follows: (1) Data collection and preprocessing: Collect hydrogen sulfide gas concentration data, including time, location, and concentration information, clean the data, remove invalid or abnormal data, and normalize the data so that it falls within an appropriate numerical range; (2) Feature selection: According to the changing rules and influencing factors of gas concentration, select the time features, spatial features and environmental features related to the prediction, and further process and transform the features; (3) Neural network structure design: Construct an LSTM neural network model, including an input layer, an LSTM layer, a fully connected layer, and an output layer; the input layer receives the preprocessed data features; the LSTM layer is responsible for capturing the long-term dependencies in the time series data; the fully connected layer maps the output of the LSTM layer to the predicted value; the output layer outputs the predicted hydrogen sulfide concentration; (4) Data partitioning: Divide the data set into training set, validation set and test set. The training set is used to train the model, the validation set is used to adjust hyperparameters and monitor model performance, and the test set is used to evaluate the final performance of the model; (5) Model training: Use the training set data to train the neural network and update the model parameters through the back-propagation algorithm; monitor the performance on the validation set and select the optimal model parameters and structure; (6) Hyperparameter tuning: Tune the hyperparameters through grid search, random search, or Bayesian optimization methods, evaluate the model performance under different hyperparameter combinations, and select the optimal hyperparameter combination.

Claims

1. A large-scale multi-point synchronous detection system for sour natural gas pipeline leakage, characterized by: The detection system comprises a main control machine (1) and a plurality of hydrogen sulfide synchronous detectors (2) connected to the main control machine (1), wherein the hydrogen sulfide synchronous detectors (2) comprise a collection and detection module, a positioning and transmission module, a control module, a display module and a power supply module which are connected in sequence.

2. A large-scale multi-point synchronous detection system for sour natural gas pipeline leakage according to claim 1, characterized in that: The collection and detection module is composed of a sampling port (201), a sampling tube (202), and a vacuum sampler (203) which are connected in sequence from top to bottom; The positioning transmission module comprises a satellite positioning module (204) and a wireless transmission module (205); The control module is a single chip microcomputer controller (210).

3. A large-scale multi-point synchronous detection system for sour natural gas pipeline leakage according to claim 2, characterized in that: The vacuum sampler (203) is composed of a sampling air intake pipe (203-2), a vacuum sampling tank (203-13), and a drainage pipe (203-8) which are connected in sequence; the sampling air intake pipe (203-2) is provided with a sampling air intake automatic valve (203-1); one end of the sampling air intake pipe (203-2) is connected to the sampling port (201), and the other end is connected to the vacuum sampling tank (203-13); The drainage pipe (203-8) is provided with an automatic drainage valve (203-9), one end of which is connected to the vacuum sampling tank (203) and the other end of which is in communication with the outside of the vacuum sampler (203); It also includes a nitrogen tank (203-5) connected to the vacuum sampling tank (203-13) via a nitrogen intake pipe (203-3) and a nitrogen intake automatic valve (203-4) arranged on the nitrogen intake pipe (203-3); it also includes a micro vacuum pump (203-12), the micro vacuum pump (203-12) is connected to the vacuum sampling tank (203-13) via an exhaust pipe (203-10), and the exhaust pipe (203-10) is provided with an exhaust automatic valve (203-11); It also includes a hydrogen sulfide sensor (203-6), a temperature sensor (203-14) and a pressure sensor (203-15) which are arranged inside the vacuum sampling tank (203-13).

4. A large-scale multi-point synchronous detection system for sour natural gas pipeline leakage according to claim 3, characterized in that: It also includes a bracket (203-7) located at the bottom of the vacuum sampling tank (203-13), and the hydrogen sulfide sensor (203-6) is placed on the bracket (203-7).

5. A large-scale multi-point synchronous detection system for sour natural gas pipeline leakage according to claim 1, characterized in that: It also includes a control box (214), wherein the positioning transmission module, the control module and the power supply module are located inside the control box (214); the acquisition and detection module is fixedly connected outside the control box; and the positioning transmission module also includes an antenna (213).

6. A large-scale multi-point synchronous detection system for sour natural gas pipeline leakage according to claim 5, characterized in that: It also includes a balancing module, which is arranged in the control box (214) and consists of an acceleration sensor (206), a buffer base (207), and a buffer spring (208) fixedly arranged on the buffer base (207); the acceleration sensor (206) is connected to the positioning transmission module; The control box (214) is semi-enclosed, and the buffer base (207) is buckled onto the bottom surface of the control box (214).

7. A large-scale multi-point synchronous detection system for sour natural gas pipeline leakage according to claim 6, characterized in that: It also includes a partition (209) disposed in the control box (214), wherein the partition (209) divides an area formed by the control box (214) and the buffer base (207) into upper and lower layers, and the positioning transmission module, the control module, the acceleration sensor (206), and the power supply module are located in the upper area of ​​the partition (209); The power module is a battery (211); The buffer spring (208) is located in the lower region.

8. The large-scale multi-point synchronous detection system for sour natural gas pipeline leakage according to claim 1 is characterized by: The display module is an indication alarm device (212) and is fixed on the outer surface of the control box (204).

9. A large-scale multi-point synchronous detection system for sour natural gas pipeline leakage according to any one of claims 5 to 8, characterized in that: The control box (204) is also provided with a lifting ring (215), and also includes a drone (4), wherein the drone (4) is connected to the lifting ring (215) via a sling (3) to lift the hydrogen sulfide synchronous detector (2).

10. A method for predicting the diffusion law of hydrogen sulfide, using a large-scale multi-point synchronous detection system for sour natural gas pipeline leakage as claimed in claim 9, characterized in that: The prediction method is as follows: (1) Data collection and preprocessing: Collect hydrogen sulfide gas concentration data, including time, location, and concentration information, clean the data, remove invalid or abnormal data, and normalize the data so that it falls within an appropriate numerical range; (2) Feature selection: According to the changing rules and influencing factors of gas concentration, select the time features, spatial features and environmental features related to the prediction, and further process and transform the features; (3) Neural network structure design: Construct an LSTM neural network model, including an input layer, an LSTM layer, a fully connected layer, and an output layer; the input layer receives the preprocessed data features; the LSTM layer is responsible for capturing the long-term dependencies in the time series data; the fully connected layer maps the output of the LSTM layer to the predicted value; the output layer outputs the predicted hydrogen sulfide concentration; (4) Data partitioning: Divide the data set into training set, validation set, and test set; The training set is used to train the model, the validation set is used to adjust hyperparameters and monitor model performance, and the test set is used to evaluate the final performance of the model; (5) Model training: Use the training set data to train the neural network and update the model parameters through the back-propagation algorithm; monitor the performance on the validation set and select the optimal model parameters and structure; (6) Hyperparameter tuning: Tune the hyperparameters through grid search, random search, or Bayesian optimization methods, evaluate the model performance under different hyperparameter combinations, and select the optimal hyperparameter combination.

Images