Flow metering method, system and device based on Internet of Things

Abstract

The invention belongs to the technical field of flow metering, and discloses a flow metering method, system and device based on the Internet of Things, and the method comprises the steps of monitoring data collection and transmission, monitoring data preprocessing, initial volume flow analysis, mass flow analysis and flow monitoring data display. The system comprises a monitoring data collection and transmission module, a monitoring data preprocessing module, an initial volume flow analysis module, a mass flow analysis module and a flow monitoring data display module. Multi-dimensional data acquisition is carried out by comprehensively using ultrasonic waves and a differential pressure type flow sensor, working condition identification is carried out by combining data acquisition nodes of the internet of things and a convolutional neural network, the volume flow is calculated according to a fusion weight matrix, and the mass flow is converted according to a mass flow calculation model. The mass flow rate of the natural gas can be measured at high precision under different working conditions, and the problem that the measurement precision is limited in a full working condition range in a traditional mode is solved.

Description

Technical Field

[0001] The present invention belongs to the technical field of flow measurement, and in particular relates to a flow measurement method, system and device based on the Internet of Things. Background Art

[0002] With the continuous adjustment of energy structure, natural gas, as a clean and efficient energy source, has been applied more and more widely, covering many fields such as industrial production, chemical preparation and residents' life. In the process of natural gas transportation from the production end to the use end, accurate measurement of its flow is not only related to the rational allocation and efficient utilization of energy, but also an important guarantee for the fairness and justice of natural gas trade settlement. However, traditional natural gas flow measurement technology faces many challenges in practical applications, especially the limited measurement accuracy under complex working conditions, which has become a key bottleneck restricting the development of natural gas flow measurement technology.

[0003] In the prior art, natural gas flow measurement mainly relies on a single sensor or a fixed weight fusion method. For example, ultrasonic flow sensors and differential pressure flow sensors are commonly used measurement tools, but there are differences in the measurement accuracy of the two under different working conditions. Ultrasonic flow sensors perform well at low flow rates, while differential pressure flow sensors are more accurate at high flow rates. However, traditional methods usually use fixed weights to fuse data from different sensors. This method cannot flexibly adjust the weights according to changes in working conditions, resulting in limited measurement accuracy within the full range of working conditions. In addition, traditional methods fail to fully utilize the impact of environmental parameters (such as temperature, pressure, and natural gas composition) on flow measurement, further limiting the improvement of measurement accuracy.

[0004] In recent years, the rapid development of Internet of Things technology and machine learning algorithms has provided new ideas for solving the above problems. Internet of Things technology can realize the real-time collection and transmission of multi-source data, providing a richer data foundation for flow measurement; while machine learning algorithms can recognize and model complex working conditions by learning and analyzing historical data. Therefore, it is possible to consider combining these two technologies to develop an adaptive flow measurement method. Summary of the invention

[0005] In view of the above-mentioned problems, the purpose of the present invention is to provide a flow metering method, system and device based on the Internet of Things, which can measure the natural gas mass flow rate with high precision under different operating conditions, and solve the problem of limited measurement accuracy of traditional methods within the full range of operating conditions.

[0006] To achieve the above object, the present invention provides the following technical solution: a flow metering method based on the Internet of Things, comprising the following steps: S1. Monitoring data collection and transmission: Acquire monitoring data during the normal natural gas transportation process and transmit the monitoring data to the IoT data collection node. The monitoring data includes the initial volume flow velocity vector collected by the ultrasonic flow sensor, the volume flow velocity vector collected by the differential pressure flow sensor, and the environmental monitoring data collected in real time by the environmental multi-parameter sensor; S2. Monitoring data preprocessing: After receiving the monitoring data, the IoT data collection node preprocesses it; S3. Analyze the initial volume flow rate: identify the working condition based on the preprocessed monitoring data, generate a working condition feature vector, and analyze the volume flow rate based on the working condition feature vector; S4. Analyze mass flow: Analyze mass flow based on volume flow and environmental monitoring data; S5. Flow monitoring data display: The volume flow, mass flow and monitoring data are transmitted to the remote monitoring center through the Internet of Things communication module, and visual monitoring is carried out in the remote monitoring center.

[0007] Preferably, in S1, collecting the initial volume flow velocity vector by an ultrasonic flow sensor comprises the following steps: S111, transmitting an ultrasonic pulse into the natural gas pipeline through a transmitting transducer of the ultrasonic flow sensor, and receiving the ultrasonic signal after being transmitted through the natural gas medium through a receiving transducer; S112, obtaining the inner diameter of the pipeline, and measuring the propagation time of the ultrasonic wave in the natural gas medium along the flow and against the flow; S113, calculating the initial volume flow velocity vector of the natural gas based on the downstream propagation time, the upstream propagation time, and the inner diameter of the pipeline by using the initial volume flow velocity vector calculation formula; The initial volume flow velocity vector calculation formula is: ; in, is the initial volume flow rate vector, C is a constant, is the time for downstream propagation, is the time of countercurrent propagation, and D is the inner diameter of the pipe.

[0008] Preferably, in S1, collecting the volume flow velocity vector by a differential pressure flow sensor comprises the following steps: S121. When natural gas flows through the throttling device in the differential pressure flow sensor, the pressure difference generated before and after the throttling device is measured by the pressure sensor. The differential pressure flow sensor is installed at a preset position of the natural gas pipeline. S122. According to the natural gas differential pressure-flow rate characteristic relationship curve set in the database, the measured pressure difference is converted into a volume flow rate vector of the natural gas.

[0009] Preferably, in S3, generating the operating condition characteristic vector comprises the following steps: S311. Obtaining a natural gas operating condition identification model on an IoT data collection node; S312, analyzing the monitoring data through a natural gas operating condition identification model to generate an operating condition feature vector; The natural gas operating condition identification model is obtained as follows: Obtain historical natural gas flow monitoring data sets under various working conditions and divide them into training sets and test sets; Mark the working condition category of the natural gas flow monitoring data group corresponding to each working condition information in the training set; The labeled natural gas flow monitoring data set is input into the convolutional neural network for training, and a primary natural gas operating condition recognition model is output; The primary natural gas operating condition identification model is tested through the test set, and the natural gas operating condition identification model is output.

[0010] Preferably, in S3, analyzing the volume flow rate based on the operating condition characteristic vector includes the following steps: S321, according to the generated working condition feature vector, retrieve the fusion weight matrix in the database, the fusion weight matrix includes the weights of the ultrasonic flow sensor and the differential pressure flow sensor for the measurement results under different working conditions; S322, weighted fusion of the initial volume flow velocity vector and the volume flow velocity vector is performed according to the natural gas volume flow velocity formula to obtain the final natural gas volume flow velocity , S323, combined with the cross-sectional area S of the pipe, through the formula Calculate the natural gas volume flow .

[0011] Preferably, the natural gas volume flow rate formula is: ; in, For the jth working condition, the ultrasonic flow sensor The weight of For the jth working condition, the differential pressure flow sensor is The weight of j is the working condition type number. is the initial volume flow velocity vector, is the volume flow velocity vector.

[0012] Preferably, in S4, analyzing the mass flow rate based on the volume flow rate and environmental monitoring data comprises the following steps: S401, obtaining natural gas volume flow, and extracting temperature, pressure and natural gas composition data from environmental monitoring data; S402. Calculate the mass flow rate of natural gas using a mass flow rate calculation model.

[0013] Preferably, the mass flow calculation model is: ; in, is the mass flow rate, is the volume flow rate, is the molar mass of the i-th component, i is the number of the natural gas component type, n is the total number of natural gas component types, , is the mole fraction of the ith component, Z is the compressibility factor of natural gas, R is the universal gas constant, T is the temperature, and P is the pressure.

[0014] A flow metering system based on the Internet of Things is used to implement the above method, including a monitoring data collection and transmission module, a monitoring data preprocessing module, an initial volume flow analysis module, a mass flow analysis module, and a flow monitoring data display module, wherein: The monitoring data collection and transmission module is used to obtain the monitoring data during the normal natural gas transportation process and transmit the monitoring data to the IoT data collection node. The monitoring data includes the initial volume flow velocity vector collected by the ultrasonic flow sensor, the volume flow velocity vector collected by the differential pressure flow sensor, and the environmental monitoring data collected in real time by the environmental multi-parameter sensor; The monitoring data preprocessing module is used to preprocess the monitoring data after it is received by the IoT data collection node; An initial volume flow analysis module is used to identify the working condition based on the pre-processed monitoring data, generate a working condition feature vector, and analyze the volume flow based on the working condition feature vector; Analytical mass flow module, used to analyze mass flow based on volume flow and environmental monitoring data; The flow monitoring data display module is used to transmit the volume flow, mass flow and monitoring data to the remote monitoring center through the Internet of Things communication module, and perform visual monitoring in the remote monitoring center.

[0015] A flow metering device based on the Internet of Things includes a memory, a processor, and a computer program stored in the memory and capable of running on the processor. The above system is implemented by executing the computer program by the processor.

[0016] The beneficial effects of the present invention are: The present invention uses ultrasonic and differential pressure flow sensors for multi-dimensional data collection, combines the Internet of Things data collection nodes with convolutional neural networks to identify operating conditions, calculates the volume flow rate based on the fusion weight matrix, converts the mass flow rate based on the mass flow calculation model, and transmits it to the remote monitoring center through the Internet of Things communication module. It can measure the natural gas mass flow rate with high precision under different operating conditions, solving the problem of limited measurement accuracy of traditional methods within the full range of operating conditions.

[0017] The present invention transmits the collected multi-source data to the IoT data collection node for preliminary processing, and then uses a machine learning algorithm to mine the operating condition characteristics in the data to generate an operating condition characteristic vector, thereby depicting the complex operating conditions of the natural gas. This makes the flow calculation process more in line with the actual operating conditions and avoids the large errors caused by general calculation methods under special conditions.

[0018] The present invention performs weighted fusion on the volume flow rate data collected by different sensors based on the operating condition characteristic vector to obtain an accurate volume flow rate, and then substitutes the environmental monitoring data into the mass flow calculation model to calculate the mass flow rate, thereby realizing a high-precision flow metering method that is adaptive to all operating conditions and takes multiple factors into comprehensive consideration, which is beneficial to the rational planning of natural gas resource allocation, optimization of energy utilization structure, and ensuring the efficient and stable operation of the entire natural gas industry chain. BRIEF DESCRIPTION OF THE DRAWINGS

[0019] Figure 1 It is a schematic diagram of the process of the present invention; Figure 2 Schematic diagram of the modules of the system of the present invention. DETAILED DESCRIPTION

[0020] The specific implementation modes of the present invention are described in detail below in conjunction with the accompanying drawings.

[0021] Embodiment 1: like Figure 1 As shown, a flow metering method based on the Internet of Things includes the following steps: S1. Monitoring data collection and transmission: Acquire monitoring data during the normal natural gas transportation process and transmit the monitoring data to the IoT data collection node. The monitoring data includes the initial volume flow velocity vector collected by the ultrasonic flow sensor, the volume flow velocity vector collected by the differential pressure flow sensor, and the environmental monitoring data collected in real time by the environmental multi-parameter sensor; S2. Monitoring data preprocessing: After receiving the monitoring data, the IoT data collection node preprocesses it; S3. Analyze the initial volume flow rate: identify the working condition based on the preprocessed monitoring data, generate a working condition feature vector, and analyze the volume flow rate based on the working condition feature vector; S4. Analyze mass flow: Analyze mass flow based on volume flow and environmental monitoring data; S5. Flow monitoring data display: The volume flow, mass flow and monitoring data are transmitted to the remote monitoring center through the Internet of Things communication module, and visual monitoring is carried out in the remote monitoring center.

[0022] The IoT data acquisition node can establish connections with a variety of sensors. In the scenario of natural gas flow measurement, it will simultaneously connect to ultrasonic flow sensors, differential pressure flow sensors, and environmental multi-parameter sensors. These sensors continuously generate monitoring data during the natural gas transportation process, and the IoT data acquisition node aggregates the information from different sensors. For example, the ultrasonic flow sensor will generate ultrasonic flow monitoring data, reflecting the natural gas flow information measured by the ultrasonic principle; the differential pressure flow sensor will output differential pressure flow monitoring data, reflecting the flow information obtained based on the differential pressure principle; the environmental multi-parameter sensor will provide environmental monitoring data such as temperature and pressure. The IoT data acquisition node receives these data from different sources and different types, and centrally forms a complete data set.

[0023] The IoT communication module is a hardware module that integrates communication chips, peripheral circuits and communication protocols. Its main function is to connect the IoT data acquisition node to the network and realize data transmission between the device and the remote monitoring center. It can be implemented using publicly known technologies.

[0024] In S1, collecting the initial volume flow velocity vector by the ultrasonic flow sensor includes the following steps: S111, transmitting an ultrasonic pulse into the natural gas pipeline through a transmitting transducer of the ultrasonic flow sensor, and receiving the ultrasonic signal after being transmitted through the natural gas medium through a receiving transducer; S112, obtaining the inner diameter of the pipeline, and measuring the propagation time of the ultrasonic wave in the natural gas medium along the flow and against the flow; S113. Based on the downstream propagation time, the upstream propagation time, and the inner diameter of the pipeline, the initial volume flow velocity vector of the natural gas is calculated using the initial volume flow velocity vector calculation formula.

[0025] The initial volume flow velocity vector calculation formula is: ; in, is the initial volume flow rate vector, C is a constant, is the time for downstream propagation, is the time of countercurrent propagation, and D is the inner diameter of the pipe.

[0026] The acquisition of volume flow velocity vector by differential pressure flow sensor includes the following steps: S121. When natural gas flows through the throttling device in the differential pressure flow sensor, the pressure difference generated before and after the throttling device is measured by the pressure sensor. The differential pressure flow sensor is installed at a preset position of the natural gas pipeline (the installation position is selected according to needs); S122. According to the natural gas differential pressure-flow rate characteristic relationship curve set in the database, the measured pressure difference is converted into a volume flow rate vector of the natural gas.

[0027] Common throttling devices include orifice plates, venturi tubes, nozzles, etc. According to Bernoulli's principle, due to the change in the flow rate of natural gas, a pressure difference will be generated before and after the throttling device. Because when natural gas flows through the throttling device, the flow area changes, the flow rate increases, and according to the law of conservation of energy, the pressure will decrease accordingly.

[0028] In S3, generating the operating condition feature vector includes the following steps: S311. Obtain a natural gas operating condition identification model on the IoT data collection node; the IoT data collection node, as the data processing and transmission hub of the entire system, can store or access the natural gas operating condition identification model. The natural gas operating condition identification model is a pre-trained algorithm model used to identify different operating conditions in the natural gas transmission process.

[0029] S312, analyzing the monitoring data through a natural gas operating condition identification model to generate an operating condition feature vector; The natural gas operating condition identification model is obtained as follows: Obtain historical natural gas flow monitoring data sets under various working conditions and divide them into training sets and test sets; The natural gas flow monitoring data group corresponding to each working condition information in the training set is labeled with working condition categories; data with low flow rate, high pressure and low temperature are labeled as "low flow rate-high pressure-low temperature" working conditions, and data with high flow rate, moderate pressure and high temperature are labeled as "high flow rate-medium pressure-high temperature" working conditions, etc. Through this labeling, a clear learning goal is provided for the natural gas working condition recognition model, so that the model can learn the mapping relationship between the characteristics of different working conditions and the corresponding category labels during the training process.

[0030] The labeled natural gas flow monitoring data set is input into the convolutional neural network for training, and a primary natural gas operating condition recognition model is output; The primary natural gas operating condition identification model is tested through the test set, and the natural gas operating condition identification model is output.

[0031] The convolutional neural network may adopt a known architecture or a specialized convolutional neural network. For example, a convolutional neural network architecture specifically applicable to the present embodiment includes a multimodal input fusion module, a dynamic weight adjustment module, a depthwise separable convolution module, a multiscale feature fusion module, and an adaptive working condition classification module, wherein: The multimodal input fusion module structure includes: Input layer: ultrasonic flow sensor data, differential pressure flow sensor data and environmental parameters (temperature, pressure, composition) are used as independent input channels; Feature extraction layer: Design an independent convolution layer for each modality to extract its features separately; for example: Ultrasonic data: Use 1D convolutional layer to extract time series features; Differential pressure data: Use 1D convolutional layer to extract pressure difference change features; Environmental parameters: Use a fully connected layer to convert parameters such as temperature, pressure, and composition into feature vectors; Fusion layer: Through an attention mechanism module, the weights of each modal data under different working conditions are dynamically adjusted to achieve adaptive fusion.

[0032] The dynamic weight adjustment module includes: Working condition feature extraction: Extract working condition feature vectors in the middle layer of CNN; Weight generation network: Input the working condition feature vector into a fully connected network (small) and output the weight of each sensor data; Weight application: The generated weights are applied to the output of the multimodal input fusion module to obtain the weighted feature vector.

[0033] The depth-wise separable convolution module includes: Deep convolution: Perform convolution operations on each input channel independently to extract local features; Point-by-point convolution: Use 1×1 convolution to perform channel fusion on the output of depth-wise convolution to generate the final feature map.

[0034] The multi-scale feature fusion module includes: Multi-scale convolution: Use convolution kernels of different sizes (such as 3×3, 5×5, 7×7) to extract features in parallel; Feature fusion: feature maps of different scales are fused through skip connection and concatenation; Feature enhancement: Use a lightweight attention module to enhance key features and suppress redundant features.

[0035] The adaptive condition classification module includes: Working condition classification network: Based on the extracted working condition feature vector, the working condition category is output through a fully connected network; Weight adjustment network: dynamically adjust the weight of sensor data according to the working condition category to generate a fusion weight matrix; Output layer: Combine the operating condition category and weight matrix to output the final operating condition feature vector and weight matrix for subsequent flow calculation.

[0036] Analyzing the volume flow based on the operating condition characteristic vector includes the following steps: S321, according to the generated working condition feature vector, retrieve the fusion weight matrix in the database, the fusion weight matrix includes the weights of the ultrasonic flow sensor and the differential pressure flow sensor for the measurement results under different working conditions; S322, weighted fusion of the initial volume flow velocity vector and the volume flow velocity vector is performed according to the natural gas volume flow velocity formula to obtain the final natural gas volume flow velocity , S323, combined with the cross-sectional area S of the pipe, through the formula Calculate the natural gas volume flow .

[0037] The formula for natural gas volume flow rate is: ; in, For the jth working condition, the ultrasonic flow sensor The weight of For the jth working condition, the differential pressure flow sensor is The weight of j is the working condition type number. is the initial volume flow velocity vector, is the volume flow velocity vector.

[0038] The rows of the fusion weight matrix represent different working condition categories, and the columns represent the weights of the ultrasonic flow sensor and the differential pressure flow sensor respectively; the fusion weight matrix W can be expressed as: ; Where M is the total number of working conditions.

[0039] In S4, analyzing the mass flow rate based on the volume flow rate and environmental monitoring data includes the following steps: S401, obtaining natural gas volume flow, and extracting temperature, pressure and natural gas composition data from environmental monitoring data; S402. Calculate the mass flow rate of natural gas using a mass flow rate calculation model.

[0040] The mass flow calculation model is: ; in, is the mass flow rate, is the volume flow rate, is the molar mass of the i-th component, i is the number of the natural gas component type, n is the total number of natural gas component types, , is the mole fraction of the ith component, Z is the dimensionless compressibility factor of natural gas, obtained by the AGA8 equation, R is the universal gas constant, with a value of 8.314 J / (mol·K), T is the temperature, and P is the pressure.

[0041] Embodiment 2: This embodiment, based on Embodiment 1, further includes: S6, dynamic adaptive model optimization: based on real-time monitoring data and changes in environmental parameters, dynamically update the natural gas operating condition identification model and / or mass flow calculation model through an online learning algorithm, the steps include: S611, real-time collection of natural gas flow monitoring data, environmental monitoring data and corresponding actual working condition labels; S612, using the newly collected data as an incremental training set, inputting it into a natural gas operating condition identification model and / or a mass flow calculation model, and iteratively optimizing the model parameters using an online transfer learning algorithm; S613, dynamically adjusting the fusion weight matrix, the compression factor Z or the mole fraction of the natural gas component according to the optimized model parameters to improve the real-time accuracy of the flow calculation; S614: When a sudden change in environmental parameters is detected or the operating condition characteristic vector exceeds a preset threshold, a global calibration of the model parameters is triggered, and an optimization log is generated and transmitted to a remote monitoring center.

[0042] By introducing the online transfer learning algorithm, the model parameters can be dynamically optimized according to real-time data, solving the problem of the traditional static model's reduced accuracy under complex working conditions or sudden environmental changes, and significantly improving the adaptive ability. Combining incremental training with the global calibration mechanism not only ensures the real-time responsiveness of the model, but also avoids model degradation caused by local data deviations, thereby enhancing the robustness and reliability of flow measurement.

[0043] Embodiment 3: like Figure 2 As shown, a flow metering system based on the Internet of Things is used to implement the method in Example 1 or Example 2, including a monitoring data collection and transmission module, a monitoring data preprocessing module, an initial volume flow analysis module, a mass flow analysis module, and a flow monitoring data display module, wherein: The monitoring data collection and transmission module is used to obtain the monitoring data during the normal natural gas transportation process and transmit the monitoring data to the IoT data collection node. The monitoring data includes the initial volume flow velocity vector collected by the ultrasonic flow sensor, the volume flow velocity vector collected by the differential pressure flow sensor, and the environmental monitoring data collected in real time by the environmental multi-parameter sensor; The monitoring data preprocessing module is used to preprocess the monitoring data after it is received by the IoT data collection node; An initial volume flow analysis module is used to identify the working condition based on the pre-processed monitoring data, generate a working condition feature vector, and analyze the volume flow based on the working condition feature vector; Analytical mass flow module, used to analyze mass flow based on volume flow and environmental monitoring data; The flow monitoring data display module is used to transmit the volume flow, mass flow and monitoring data to the remote monitoring center through the Internet of Things communication module, and perform visual monitoring in the remote monitoring center.

[0044] Embodiment 4: A flow metering device based on the Internet of Things includes a memory, a processor, and a computer program stored in the memory and capable of running on the processor. The system in Example 3 is implemented by executing the computer program by the processor.

Claims

1. A flow metering method based on the Internet of Things, characterized in that: The following steps are involved: S1. Monitoring data collection and transmission: Acquire monitoring data during the normal natural gas transportation process and transmit the monitoring data to the IoT data collection node. The monitoring data includes the initial volume flow velocity vector collected by the ultrasonic flow sensor, the volume flow velocity vector collected by the differential pressure flow sensor, and the environmental monitoring data collected in real time by the environmental multi-parameter sensor; S2. Monitoring data preprocessing: After receiving the monitoring data, the IoT data collection node preprocesses it; S3. Analyze the initial volume flow rate: identify the working condition based on the preprocessed monitoring data, generate a working condition feature vector, and analyze the volume flow rate based on the working condition feature vector; S4. Analyze mass flow: Analyze mass flow based on volume flow and environmental monitoring data; S5. Flow monitoring data display: The volume flow, mass flow and monitoring data are transmitted to the remote monitoring center through the Internet of Things communication module, and visual monitoring is carried out in the remote monitoring center.

2. A flow metering method based on the Internet of Things as claimed in claim 1, characterized in that: In S1, collecting the initial volume flow velocity vector by the ultrasonic flow sensor includes the following steps: S111, transmitting an ultrasonic pulse into the natural gas pipeline through a transmitting transducer of the ultrasonic flow sensor, and receiving the ultrasonic signal after being transmitted through the natural gas medium through a receiving transducer; S112, obtaining the inner diameter of the pipeline, and measuring the propagation time of the ultrasonic wave in the natural gas medium along the flow and against the flow; S113, calculating the initial volume flow velocity vector of the natural gas based on the downstream propagation time, the upstream propagation time, and the inner diameter of the pipeline by using the initial volume flow velocity vector calculation formula; The initial volume flow velocity vector calculation formula is: ； in, is the initial volume flow rate vector, C is a constant, is the time for downstream propagation, is the time of countercurrent propagation, and D is the inner diameter of the pipe.

3. A flow metering method based on the Internet of Things as claimed in claim 1, characterized in that: In S1, collecting the volume flow velocity vector by using a differential pressure flow sensor includes the following steps: S121. When natural gas flows through the throttling device in the differential pressure flow sensor, the pressure difference generated before and after the throttling device is measured by the pressure sensor. The differential pressure flow sensor is installed at a preset position of the natural gas pipeline. S122. According to the natural gas differential pressure-flow rate characteristic relationship curve set in the database, the measured pressure difference is converted into a volume flow rate vector of the natural gas.

4. A flow metering method based on the Internet of Things as claimed in claim 1, characterized in that: In S3, generating the operating condition feature vector includes the following steps: S311. Obtaining a natural gas operating condition identification model on an IoT data collection node; S312, analyzing the monitoring data through a natural gas operating condition identification model to generate an operating condition feature vector; The natural gas operating condition identification model is obtained as follows: Obtain historical natural gas flow monitoring data sets under various working conditions and divide them into training sets and test sets; Mark the working condition category of the natural gas flow monitoring data group corresponding to each working condition information in the training set; The labeled natural gas flow monitoring data set is input into the convolutional neural network for training, and a primary natural gas operating condition recognition model is output; The primary natural gas operating condition identification model is tested through the test set, and the natural gas operating condition identification model is output.

5. A flow metering method based on the Internet of Things as claimed in claim 1, characterized in that: In S3, analyzing the volume flow rate based on the operating condition characteristic vector includes the following steps: S321, according to the generated working condition feature vector, retrieve the fusion weight matrix in the database, the fusion weight matrix includes the weights of the ultrasonic flow sensor and the differential pressure flow sensor for the measurement results under different working conditions; S322, weighted fusion of the initial volume flow velocity vector and the volume flow velocity vector is performed according to the natural gas volume flow velocity formula to obtain the final natural gas volume flow velocity , S323, combined with the cross-sectional area S of the pipe, through the formula Calculate the natural gas volume flow .

6. A flow metering method based on the Internet of Things as claimed in claim 5, characterized in that: The natural gas volume flow rate formula is: ； in, For the jth working condition, the ultrasonic flow sensor The weight of For the jth working condition, the differential pressure flow sensor is The weight of, j is the working condition type number, is the initial volume flow velocity vector, is the volume flow velocity vector.

7. A flow metering method based on the Internet of Things as claimed in claim 1, characterized in that: In S4, analyzing the mass flow rate based on the volume flow rate and environmental monitoring data includes the following steps: S401, obtaining natural gas volume flow, and extracting temperature, pressure and natural gas composition data from environmental monitoring data; S402. Calculate the mass flow rate of natural gas using a mass flow rate calculation model.

8. A flow metering method based on the Internet of Things as claimed in claim 7, characterized in that: The mass flow calculation model is: ； in, is the mass flow rate, is the volume flow rate, is the molar mass of the i-th component, i is the number of the natural gas component type, n is the total number of natural gas component types, , is the mole fraction of the ith component, Z is the compressibility factor of natural gas, R is the universal gas constant, T is the temperature, and P is the pressure.

9. A flow metering system based on the Internet of Things, used to implement the method described in any one of claims 1 to 8, characterized in that: It includes monitoring data collection and transmission module, monitoring data preprocessing module, initial volume flow analysis module, mass flow analysis module, and flow monitoring data display module, among which: The monitoring data collection and transmission module is used to obtain the monitoring data during the normal natural gas transportation process and transmit the monitoring data to the IoT data collection node. The monitoring data includes the initial volume flow velocity vector collected by the ultrasonic flow sensor, the volume flow velocity vector collected by the differential pressure flow sensor, and the environmental monitoring data collected in real time by the environmental multi-parameter sensor; The monitoring data preprocessing module is used to preprocess the monitoring data after it is received by the IoT data collection node; An initial volume flow analysis module is used to identify the working condition based on the pre-processed monitoring data, generate a working condition feature vector, and analyze the volume flow based on the working condition feature vector; Analytical mass flow module, used to analyze mass flow based on volume flow and environmental monitoring data; The flow monitoring data display module is used to transmit the volume flow, mass flow and monitoring data to the remote monitoring center through the Internet of Things communication module, and perform visual monitoring in the remote monitoring center.

10. A flow metering device based on the Internet of Things, characterized in that: The system comprises a memory, a processor, and a computer program stored in the memory and capable of running on the processor, and the system described in claim 9 is implemented by executing the computer program by the processor.

Images