Flow metering automatic calibration system based on wireless remote transmission diaphragm gas meter

Abstract

The present application relates to the technical field of automatic calibration, in particular to a flow measurement automatic calibration system based on wireless remote transmission diaphragm gas meter. It comprises: a configuration and use unit which initializes the settings of the gas meter and performs configuration and use; a battery state monitoring unit which monitors the battery power in real time, gives a warning on the liquid crystal screen when the power is insufficient, and automatically closes the valve to cut off the gas supply when the power is used up; a user interface and control unit which provides relevant information to the user through the liquid crystal display and accepts the user's input instructions to perform corresponding control operations; a flow measurement and automatic calibration unit which monitors the gas flow through the gas meter and uses the normalized least mean square algorithm to periodically check and adjust the measurement accuracy. The present application designs to introduce the normalized least mean square algorithm for periodic flow measurement automatic calibration, which can dynamically adjust its internal parameters to compensate for the deviation caused by environmental changes and hardware aging, ensuring the high accuracy of gas flow measurement over a long period of time.

Description

Technical Field

[0001] This invention relates to the field of automatic calibration technology, and more specifically, to an automatic calibration system for flow measurement based on a wireless remote transmission diaphragm gas meter. Background Technology

[0002] Traditional mechanical or early electronic gas meters may experience a gradual decline in measurement accuracy due to long-term use, environmental changes (such as temperature and pressure variations), and internal component wear. Without regular manual calibration, this can lead to inaccurate billing, harming both users and suppliers. For a large number of distributed gas meters, regular on-site manual calibration is not only costly in terms of manpower and resources but also inefficient, especially in hard-to-reach or remote areas where maintenance costs are even higher. Furthermore, manual calibration carries the risk of error and cannot guarantee consistently achieving the desired accuracy. Traditional gas meters typically lack remote data transmission capabilities, hindering real-time flow monitoring and user information updates. This makes it difficult for gas companies to promptly understand user gas usage and provide timely service feedback, such as low balance alerts and overdue payment notifications. Additionally, they cannot respond quickly to abnormal situations (such as leaks), increasing safety hazards. Therefore, this paper proposes an automatic flow metering calibration system based on a wireless remote-reading diaphragm gas meter. Summary of the Invention

[0003] The purpose of this invention is to provide an automatic calibration system for flow measurement based on a wireless remote transmission diaphragm gas meter, in order to solve the problems mentioned in the background art, such as the decrease in metering accuracy of traditional mechanical or early electronic gas meters over time, high cost of manual calibration, and lack of real-time monitoring and remote management capabilities.

[0004] To achieve the above objectives, the present invention aims to provide an automatic flow metering calibration system based on a wireless remote transmission diaphragm gas meter, comprising:

[0005] A configuration and usage unit is provided, which initializes the settings of the gas meter and performs usage configuration.

[0006] A battery status monitoring unit monitors the battery power in real time, issues a warning on the LCD screen when the power is low, and automatically closes the valve to cut off the gas supply when the power is exhausted.

[0007] The user interface and control unit provide relevant information to the user through an LCD screen, accept user input commands to perform corresponding control operations, and manage the wireless connection with the household combustible gas detector.

[0008] The flow metering and automatic calibration unit monitors the gas flow rate through the gas meter and periodically checks and adjusts the measurement accuracy using a normalized minimum mean square algorithm.

[0009] As a further improvement to this technical solution, the configuration and usage unit includes a barcode management module, a battery initialization module, and a payment and recharge management module;

[0010] The barcode management module reads barcode information and associates it with the user's account.

[0011] The battery initialization module displays the remaining amount approximately 30 seconds after the user installs the battery, and automatically opens the valve to start supplying gas when the balance is greater than 0 yuan.

[0012] The payment and recharge management module uses wireless transmission technology to provide remote recharge and synchronizes the latest remaining amount to the LCD screen for display.

[0013] As a further improvement to this technical solution, the battery status monitoring unit monitors the battery level in real time, issues a warning on the LCD screen when the battery is low, and automatically shuts off the gas supply when the battery is completely depleted, including the following steps:

[0014] S1.1 After the gas meter is powered on, it begins initialization, reads and records the current battery voltage level, and evaluates the remaining battery life based on the extreme gradient boosting model, and sets the initial power threshold.

[0015] S1.2 When the battery voltage drops to the preset "low power" threshold, the battery status monitoring unit will display a "low power" warning message on the LCD screen through the user interface and control unit.

[0016] S1.3 If the user fails to replace the battery in time, and the battery voltage drops further below another threshold, the battery status monitoring unit will perform a protection action.

[0017] As a further improvement to this technical solution, in step S1.1, the remaining battery life is evaluated based on the extreme gradient boosting model, and an initial power threshold is set, including the following steps:

[0018] S1.11. Collect historical records of electrical parameters from existing gas meters;

[0019] S1.12. For each historical record, label the corresponding battery status and the time point when the battery was replaced as a tag;

[0020] S1.13. Preprocess the historical data and extract its features;

[0021] S1.14. Divide the historical data dataset into a training set and a test set, and use the training set to train the extreme gradient boosting model;

[0022] S1.15. Pass the currently collected historical information as input to the extreme gradient boosting model. The extreme gradient boosting model outputs the predicted remaining battery life and the suggested initial charge threshold.

[0023] As a further improvement to this technical solution, in S1.14, the extreme gradient boosting model is as follows:

[0024] ;

[0025] in, Represents the loss function; Indicates the index of the training set samples; Indicates the first The actual remaining battery life of a training set sample; The model represents the first Predicted remaining battery life for each training set sample; This represents the number of decision trees that constitute the extreme gradient boosting model; The index representing the decision tree; Indicates the first A decision tree.

[0026] As a further improvement to this technical solution, the user interface and control unit include a display module, a control module, and a management wireless connection module;

[0027] The display module uses an LCD screen to display the current system status, battery alarm information, and recharge prompts.

[0028] The control module executes corresponding control operations by accepting control commands from the user.

[0029] The management wireless connection module continuously monitors the wireless connection status with the home combustible gas detector, allows users to modify detector-related settings through the interface, and uses reinforcement learning algorithms to enable the system to adjust the wireless connection parameters with the detector under different environmental conditions.

[0030] As a further improvement to this technical solution, the reinforcement learning algorithm is as follows:

[0031] ;

[0032] in, Indicates the control system at time step In state Take action at the time The sum of expected future rewards; Indicates the control system at time step The state; Indicates the control system at time step The actions taken; Indicates the learning rate; Indicates an immediate reward; Indicates the discount factor; Indicates the next state The maximum expected future reward for all possible actions; Represents the action value function; Indicates the time step.

[0033] As a further improvement to this technical solution, the flow metering and automatic calibration unit monitors the gas flow rate through the gas meter and periodically checks and adjusts the measurement accuracy using a normalized least mean square algorithm, including the following steps:

[0034] S2.1. Real-time collection of gas flow data passing through the gas meter;

[0035] S2.2, Perform preliminary processing on the collected gas flow data;

[0036] S2.3 Trigger a calibration process according to a preset time interval;

[0037] S2.4. Perform calibration using the normalized least mean square algorithm, comparing the currently collected gas flow data with historical gas flow data to identify deviations.

[0038] As a further improvement to this technical solution, in step S2.4, the normalized least mean square algorithm is performed for calibration, including the following steps:

[0039] S2.41. Set the step size parameter, set it to a positive number, and initialize the filter coefficient vector;

[0040] S2.42. Prepare a buffer for storing gas flow data and desired output;

[0041] S2.43. For each time point, obtain the current gas flow data from the sensor;

[0042] S2.44. Obtain the expected output corresponding to the gas flow data;

[0043] S2.45. Using the current filter coefficient vector and gas flow data, calculate the predicted output and the error between the predicted output and the expected output;

[0044] S2.46. Update the filter coefficients using the normalized minimum mean square algorithm formula based on the error and gas flow data;

[0045] S2.47. Repeat steps S2.41 to S2.46 as a loop until a predetermined number of iterations m have been performed.

[0046] As a further improvement to this technical solution, in S2.46, the normalized least mean square algorithm formula is as follows:

[0047] ;

[0048] in, Represents the filter coefficient vector; This represents the updated filter coefficient vector; Indicates a point in time; Indicates the step size parameter; Represents positive numbers; Indicates the error signal; Represents a gas flow data vector

[0049] Compared with the prior art, the beneficial effects of the present invention are as follows:

[0050] 1. This automatic flow metering calibration system based on a wireless remote-reading diaphragm gas meter incorporates the Normalized Least Mean Square (NLMS) algorithm for periodic automatic flow measurement calibration. This system dynamically adjusts its internal parameters to compensate for deviations caused by environmental changes and hardware aging, ensuring high accuracy of gas flow measurement over extended periods. This not only contributes to fair and impartial billing but also provides users with reliable gas usage data, helping them better manage energy consumption and thus achieve energy conservation and emission reduction goals.

[0051] 2. This automatic calibration system for flow metering based on a wireless remote-reading diaphragm gas meter is equipped with a battery status monitoring unit, a user interface and control unit, and wireless connection management optimized by reinforcement learning algorithms. These features work together to improve user convenience and system safety. For example, it monitors battery power in real time and issues warnings when the power is low, avoiding safety hazards caused by power depletion. It also supports remote recharge and provides instant account balance feedback on an LCD screen, simplifying the payment process for users. Furthermore, the system can seamlessly interface with household combustible gas detectors, automatically establishing a wireless connection and continuously optimizing communication performance to ensure timely response when anomalies are detected, further protecting household gas safety. Attached Figure Description

[0052] Figure 1 This is an overall flowchart of the present invention;

[0053] The meanings of the labels in the diagram are as follows:

[0054] 1. Configuration and usage unit; 2. Battery status monitoring unit; 3. User interface and control unit; 4. Flow metering and automatic calibration unit. Detailed Implementation

[0055] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0056] Example: Please refer to Figure 1 As shown, an automatic flow metering calibration system based on a wireless remote transmission diaphragm gas meter is provided, including:

[0057] Configure and use Unit 1 to initialize the gas meter settings and perform usage configuration;

[0058] In this embodiment, the configuration and usage unit 1 includes a barcode management module, a battery initialization module, and a payment and recharge management module;

[0059] The barcode management module reads barcode information and associates it with the user's account;

[0060] The battery initialization module displays the remaining amount approximately 30 seconds after the user installs the battery, and automatically opens the valve to start supplying gas when the balance is greater than 0 yuan.

[0061] The payment and recharge management module uses wireless remote transmission technology to provide remote recharge and synchronizes the latest remaining amount to the LCD screen. Users can recharge remotely by scanning the QR code on the battery box cover with a mobile APP (or other methods, follow the prompts to recharge remotely; after recharging, press and hold the red button on the right side of the meter, or the orange button on the right side of the front of the meter, and the LCD will display the latest remaining amount after about 3 minutes).

[0062] The battery status monitoring unit 2 monitors the battery power in real time, issues a warning on the LCD screen when the power is low, and automatically closes the valve to cut off the gas supply when the power is exhausted.

[0063] In this embodiment, the battery status monitoring unit 2 monitors the battery power in real time, issues a warning on the LCD screen when the power is low, and automatically closes the valve to cut off the gas supply when the power is exhausted, including the following steps:

[0064] S1.1 After the gas meter is powered on, it begins initialization, reads and records the current battery voltage level, and simultaneously assesses the remaining battery life based on the Extreme Gradient Boosting (XGBoost) model and sets the initial power threshold.

[0065] Furthermore, the Extreme Gradient Boosting (XGBoost) model learns the relationship between historical electrical parameters and battery state by integrating multiple decision trees, thereby predicting the current battery's remaining lifespan and dynamically setting a power threshold. XGBoost is an efficient machine learning algorithm that forms a powerful predictive model by constructing multiple decision trees. For battery life prediction, XGBoost can learn complex nonlinear relationships from historical data and performs exceptionally well when handling large-scale datasets, thus achieving high-precision predictions of battery remaining lifespan. XGBoost has a built-in feature selection mechanism that can automatically identify which electrical parameters (such as voltage, current, and temperature) have the greatest impact on battery life. This helps to more accurately understand how various factors work together in the battery aging process and optimize power monitoring strategies accordingly. Since the XGBoost model is trained on historical data, as new data is continuously added, the model can be kept up-to-date through retraining or online updates, ensuring that it always reflects the latest battery behavior patterns. This means that even if battery performance changes over time, the model can adapt to these changes in a timely manner and provide reliable prediction results.

[0066] Based on the extreme gradient boosting model, the remaining battery life is assessed, and an initial charge threshold is set, including the following steps:

[0067] S1.11. Collect historical records of electrical parameters from existing gas meters, including battery voltage, current, temperature, usage time, ambient temperature, etc.

[0068] S1.12. For each historical record, label the corresponding battery status (such as "normal", "low power", "power exhausted") and the time point when the battery was replaced as a tag.

[0069] S1.13. Preprocess the historical data to remove outliers or incomplete data records to ensure data quality, and extract features from the historical data, including average voltage, maximum and minimum voltage difference, voltage change rate, temperature change trend, etc. These features will be used to train the model.

[0070] S1.14. Divide the historical data dataset into a training set and a test set. Use the training set to train the extreme gradient boosting model and adjust the hyperparameters to optimize performance.

[0071] Furthermore, the extreme gradient boosting model is as follows:

[0072] ;

[0073] in, This represents the loss function, the goal of which is to minimize this value. It combines prediction error and model complexity and is used to evaluate the performance of battery life prediction models. Indicates the index of the training set samples; Indicates the first The actual remaining battery life of a training set sample; The model represents the first Predicted remaining battery life for each training set sample; This represents the number of decision trees that constitute the extreme gradient boosting model; The index representing the decision tree; Indicates the first A series of decision trees, which together determine the prediction of the battery state;

[0074] S1.15. The currently collected battery voltage, temperature and other information are passed as input to the extreme gradient boosting model. The extreme gradient boosting model outputs the predicted remaining battery life and the suggested initial charge threshold.

[0075] S1.2 When the battery voltage drops to a preset "low power" threshold (e.g., close to but not yet fully depleted), the battery status monitoring unit will display a "low power" warning message on the LCD screen through the user interface and control unit 3.

[0076] S1.3 If the user fails to replace the battery in time, and the battery voltage drops further below another threshold (i.e., the "power depletion" state), the battery status monitoring unit 2 will perform a protection action (first, it will reconfirm the warning information on the LCD screen with the control unit 3 through the user interface, and may amplify the warning signal (such as continuous beeping or a red LED on). Then, in order to prevent safety hazards caused by sudden power interruption, the battery status monitoring unit will send an instruction to the flow metering and automatic calibration unit 2, which is responsible for closing the valve to cut off the gas supply. This process ensures that the system can enter a safe state even when the power is completely depleted).

[0077] The user interface and control unit 3 provides relevant information to the user through an LCD screen, accepts user input commands to perform corresponding control operations, and manages the wireless connection with the household combustible gas detector.

[0078] In this embodiment, the user interface and control unit 3 includes a display module, a control module, and a management wireless connection module;

[0079] The display module utilizes an LCD screen to display the current system status (the LCD screen will display the current system status, such as "Connecting", "Connection Successful", or "Connection Failed". If the connection is successful, the screen will further provide information about gas usage, such as the current balance and consumption), battery alarm information (once the battery is low or other abnormal conditions are detected, the LCD screen will display corresponding warning information, such as "Low battery, please replace the battery", and may be accompanied by sound or light alarms to draw attention), and recharge prompts (if the system supports a prepaid mode, when the balance is insufficient to pay the expected gas cost, a "Please recharge" prompt will be displayed. When the LCD screen displays "Please recharge", it means that the amount in the meter is almost used up, reminding the user to recharge in time. When the LCD screen displays that the remaining amount is less than or equal to 0 yuan, the gas meter will automatically close the valve and cut off the gas supply. The valve can be reopened after recharging and pressing the button).

[0080] The control module executes corresponding control operations by accepting user control commands. Users can browse different menu options via buttons or a touch screen (if available), such as viewing history records, setting reminder thresholds, and adjusting system parameters. Users can also input commands to perform specific operations, such as manually opening or closing valves, restarting the system, and confirming receipt of alarm information.

[0081] The management wireless connection module continuously monitors the wireless connection status with the home combustible gas detector to ensure the stability and security of data transmission. It allows users to modify detector-related settings, such as sensitivity and response time, via an interface. These changes are immediately synchronized to the detector. Through reinforcement learning algorithms, the system adjusts wireless connection parameters with the detector under different environmental conditions (by continuously monitoring the wireless connection status and taking optimization actions based on current environmental conditions, such as adjusting power or frequency, and updating its strategy based on real-time feedback to achieve long-term stable optimal communication performance), such as power output and frequency selection. The reinforcement learning algorithm can dynamically adjust wireless connection parameters (such as transmit power, frequency selection, modulation method, etc.) according to the current communication environment (such as signal-to-noise ratio, signal strength, bit error rate, etc.) to achieve optimal transmission performance. This adaptive capability allows the system to maintain stable and efficient communication quality in changing environments. The system can continuously receive real-time reward or penalty signals to evaluate the effectiveness of its decisions and gradually optimize its strategy based on this feedback. This means that even when facing new or unknown environmental conditions, the system can evolve over time to find the optimal operating parameter settings.

[0082] Furthermore, the reinforcement learning algorithm is as follows:

[0083] ;

[0084] in, Indicates the control system at time step In state Take action at the time The sum of expected future rewards, which is the so-called action value function or Q-value; Indicates the control system at time step The status includes factors reflecting the quality of the wireless connection, such as the current signal-to-noise ratio, signal strength, and bit error rate. Indicates the control system at time step The actions taken are behaviors such as adjusting the transmission power, changing the frequency, and selecting different modulation methods to optimize the wireless connection; The learning rate is a value between 0 and 1 that determines the extent to which new information affects the old Q value. This refers to an immediate reward, that is, a positive reward given based on improvements in communication quality, or a negative reward due to performance degradation; This represents the discount factor, which is a number between 0 and 1, used to balance the importance of short-term and long-term rewards. Indicates the next state The maximum expected future reward for all possible actions; Represents the action value function; Indicates the time step;

[0085] During initial installation, the battery must first be correctly inserted into the gas meter. The battery initialization module will complete initialization in approximately 30 seconds and display the remaining balance. If the balance is greater than 0 yuan, the valve will automatically open to begin gas supply. Then, connect the household combustible gas detector to a power source. Once the detector is operating stably, it will attempt to automatically pair wirelessly with the gas meter. If the connection is successful, a corresponding indicator will be displayed on the gas meter's LCD screen. If the initial connection fails, the user should press and hold the button on the detector for approximately 10 seconds and then try the Bluetooth connection again until successful. This process may need to be repeated several times until a stable wireless connection is established between the gas meter and the detector.

[0086] The flow metering and automatic calibration unit 4 monitors the gas flow through the gas meter and periodically checks and adjusts the measurement accuracy using the Normalized Least Mean Square (NLMS) algorithm.

[0087] In this embodiment, the Normalized Least Mean Square (NLMS) algorithm continuously updates the filter coefficients to minimize the normalization error between the predicted output and the expected output, thereby automatically adapting to environmental changes and maintaining high measurement accuracy. The NLMS algorithm can automatically adjust its filter coefficients based on the current input data to adapt to the effects of environmental changes or hardware aging, meaning that the system can maintain high-precision measurement performance even during long-term operation. Compared to the traditional LMS algorithm, NLMS improves the convergence speed through a normalization step, enabling it to find the optimal solution in a shorter time, making it particularly suitable for applications requiring real-time adjustments, such as gas flow monitoring. The normalization factor in the NLMS algorithm ensures that the update process remains stable even when the input signal is very small or close to zero. This characteristic enhances the robustness of the system and prevents algorithm failure or instability due to extreme conditions. Regular calibration using the NLMS algorithm can promptly identify and correct measurement deviations, thereby avoiding the accumulation of errors over time, which is crucial for ensuring the accuracy of measurement results over long periods.

[0088] The flow metering and automatic calibration unit 4 monitors the gas flow rate through the gas meter and periodically checks and adjusts the measurement accuracy using the Normalized Least Mean Square (NLMS) algorithm, including the following steps:

[0089] S2.1 Real-time collection of gas flow data passing through the gas meter, accomplished through built-in sensors, such as ultrasonic sensors, turbine flow meters, or diaphragm flow meters;

[0090] S2.2 Perform preliminary processing on the collected gas flow data, such as filtering and compensating for the effects of temperature and pressure changes, to improve data accuracy;

[0091] S2.3 Trigger a calibration process once according to a preset time interval (such as daily, weekly or monthly);

[0092] S2.4. Perform calibration using the normalized least mean square algorithm. Compare the currently collected gas flow data with historical gas flow data to identify deviations. Once it is determined that adjustment is needed, the automatic calibration unit will update its internal calibration parameters to compensate for any detected deviations.

[0093] The calibration process, which involves performing a normalized least mean square algorithm, includes the following steps:

[0094] S2.41. Set the step size parameter, which determines the magnitude of each update. It is usually a small positive number. Setting a positive number is used to prevent the denominator from being zero. Initialize the filter coefficient vector, which is usually set to a zero vector or a small random value.

[0095] S2.42, Prepare for storing gas flow data and expected output buffer;

[0096] S2.43. For each time point, acquire the current gas flow data from the sensor. The input signals include relevant measurement data such as flow rate, temperature, and pressure.

[0097] S2.44 Obtain the expected output corresponding to the gas flow data, which is an ideal value derived from historical data and model prediction;

[0098] S2.45. Using the current filter coefficient vector and gas flow data, calculate the predicted output. And calculate the error between the predicted output and the expected output. ;

[0099] S2.46. Update the filter coefficients using the normalized minimum mean square algorithm formula based on the error and gas flow data;

[0100] Furthermore, the formula for the normalized least mean square algorithm is:

[0101] ;

[0102] in, This represents the filter coefficient vector, which represents the set of parameters currently used to predict the output. This represents the updated filter coefficient vector, which is achieved by adjusting... To achieve results that are closer to the ideal value; Indicates a point in time; The step size parameter (also known as the learning rate) determines the magnitude of each update. This value is a small positive number used to balance the relationship between convergence speed and stability. Representing a positive number is used to prevent the denominator from being zero, which helps ensure that the update process remains stable even when the input signal is very small or close to zero; Indicates the error signal; Represents a vector of gas flow data;

[0103] S2.47. Repeat steps S2.41 to S2.46 (data acquisition to updating filter coefficients) as a loop to adapt to environmental changes and improve measurement accuracy until a predetermined number of iterations m has been performed.

[0104] The foregoing has shown and described the basic principles, main features, and advantages of the present invention. Those skilled in the art should understand that the present invention is not limited to the above embodiments. The embodiments and descriptions in the specification are merely preferred examples and are not intended to limit the invention. Various changes and modifications can be made to the invention without departing from its spirit and scope, and all such changes and modifications fall within the scope of the claimed invention.

Claims

1. An automatic calibration system for flow measurement based on a wireless remote transmission diaphragm gas meter, characterized in that, include: Configuration and usage unit (1), which initializes the settings of the gas meter and performs usage configuration; Battery status monitoring unit (2), the battery status monitoring unit (2) monitors the battery power in real time, issues a warning on the LCD screen when the power is low, and automatically closes the valve to cut off the gas supply when the power is exhausted; User interface and control unit (3), the user interface and control unit (3) provides relevant information to the user through the LCD screen, and accepts the user input instructions to perform corresponding control operations, and manages the wireless connection with the household combustible gas detector; The flow metering and automatic calibration unit (4) monitors the gas flow through the gas meter and periodically checks and adjusts the measurement accuracy using the normalized minimum mean square algorithm.

2. The automatic flow metering calibration system based on a wireless remote transmission diaphragm gas meter according to claim 1, characterized in that: The configuration unit (1) includes a barcode management module, a battery initialization module, and a payment and recharge management module; The barcode management module reads barcode information and associates it with the user's account. The battery initialization module displays the remaining amount approximately 30 seconds after the user installs the battery, and automatically opens the valve to start supplying gas when the balance is greater than 0 yuan. The payment and recharge management module uses wireless transmission technology to provide remote recharge and synchronizes the latest remaining amount to the LCD screen for display.

3. The automatic calibration system for flow measurement based on a wireless remote transmission diaphragm gas meter according to claim 2, characterized in that: The battery status monitoring unit (2) monitors the battery power in real time, issues a warning on the LCD screen when the power is low, and automatically shuts off the gas supply when the power is depleted, including the following steps: S1.1 After the gas meter is powered on, it begins initialization, reads and records the current battery voltage level, and evaluates the remaining battery life based on the extreme gradient boosting model, and sets the initial power threshold. S1.2 When the battery voltage drops to the preset "low power" threshold, the battery status monitoring unit will display a "low power" warning message on the LCD screen through the user interface and the control unit (3); S1.3 If the user fails to replace the battery in time, and the battery voltage drops further to below another threshold, the battery status monitoring unit (2) performs a protection action.

4. The automatic flow metering calibration system based on a wireless remote transmission diaphragm gas meter according to claim 3, characterized in that: In step S1.1, the remaining battery life is evaluated based on the extreme gradient boosting model, and an initial battery capacity threshold is set, including the following steps: S1.

11. Collect historical records of electrical parameters from existing gas meters; S1.

12. For each historical record, label the corresponding battery status and the time point when the battery was replaced as a tag; S1.

13. Preprocess the historical data and extract its features; S1.

14. Divide the historical data dataset into a training set and a test set, and use the training set to train the extreme gradient boosting model; S1.

15. Pass the currently collected historical information as input to the extreme gradient boosting model. The extreme gradient boosting model outputs the predicted remaining battery life and the suggested initial charge threshold.

5. The automatic calibration system for flow measurement based on a wireless remote transmission diaphragm gas meter according to claim 4, characterized in that: In S1.14, the extreme gradient boosting model is as follows: ； in, Represents the loss function; Indicates the index of the training set samples; Indicates the first The actual remaining battery life of a training set sample; The model represents the first Predicted remaining battery life for each training set sample; This represents the number of decision trees that constitute the extreme gradient boosting model; The index representing the decision tree; Indicates the first A decision tree.

6. The automatic flow metering calibration system based on a wireless remote transmission diaphragm gas meter according to claim 5, characterized in that: The user interface and control unit (3) includes a display module, a control module, and a management wireless connection module; The display module uses an LCD screen to display the current system status, battery alarm information, and recharge prompts. The control module executes corresponding control operations by accepting control commands from the user. The management wireless connection module continuously monitors the wireless connection status with the home combustible gas detector, allows users to modify detector-related settings through the interface, and uses reinforcement learning algorithms to enable the system to adjust the wireless connection parameters with the detector under different environmental conditions.

7. The automatic flow metering calibration system based on a wireless remote transmission diaphragm gas meter according to claim 6, characterized in that: The reinforcement learning algorithm is as follows: ； in, Indicates the control system at time step In state Take action at the time The sum of expected future rewards; Indicates the control system at time step The state; Indicates the control system at time step The actions taken; Indicates the learning rate; Indicates an immediate reward; Indicates the discount factor; Indicates the next state The maximum expected future reward for all possible actions; Represents the action value function; Indicates the time step.

8. The automatic flow metering calibration system based on a wireless remote transmission diaphragm gas meter according to claim 7, characterized in that: The flow metering and automatic calibration unit (4) monitors the gas flow through the gas meter and periodically checks and adjusts the measurement accuracy using the normalized minimum mean square algorithm, including the following steps: S2.

1. Real-time collection of gas flow data passing through the gas meter; S2.2, Perform preliminary processing on the collected gas flow data; S2.3 Trigger a calibration process according to a preset time interval; S2.

4. Perform calibration using the normalized least mean square algorithm, comparing the currently collected gas flow data with historical gas flow data to identify deviations.

9. The automatic flow metering calibration system based on a wireless remote transmission diaphragm gas meter according to claim 8, characterized in that: In step S2.4, the normalized least mean square algorithm is used for calibration, including the following steps: S2.

41. Set the step size parameter, set it to a positive number, and initialize the filter coefficient vector; S2.

42. Prepare a buffer for storing gas flow data and desired output; S2.

43. For each time point, obtain the current gas flow data from the sensor; S2.

44. Obtain the expected output corresponding to the gas flow data; S2.

45. Using the current filter coefficient vector and gas flow data, calculate the predicted output and the error between the predicted output and the expected output. S2.

46. Update the filter coefficients using the normalized minimum mean square algorithm formula based on the error and gas flow data; S2.

47. Repeat steps S2.41 to S2.46 as a loop until a predetermined number of iterations m have been performed.

10. The automatic flow metering calibration system based on a wireless remote transmission diaphragm gas meter according to claim 9, characterized in that: In S2.46, the formula for the normalized least mean square algorithm is: ； in, Represents the filter coefficient vector; This represents the updated filter coefficient vector; Indicates a point in time; Indicates the step size parameter; Represents positive numbers; Indicates the error signal; This represents a vector of gas flow data.

Images