A multi-parameter fusion gas leakage warning method and device for gas-using buildings

Abstract

The present invention discloses a multi-parameter fusion gas leakage warning method and equipment for gas-using buildings, including: a computer continuously collects real-time flow and real-time pressure information of a gas meter that measures the gas consumption of gas-using equipment within a set time period, and records the collection time, real-time flow and real-time pressure information of the gas meter in chronological order to form an initial data set; sets a pressure threshold, determines a pressure interval division method, classifies the real-time pressure of the initial data set to form an initial data subset, continues to collect real-time pressure data and supplements it to the initial data subset to form a real-time data subset, and constructs a Poisson random variable model; analyzes the distribution of flow feature estimation values ​​in the real-time data subset; determines the distribution mean of the flow feature estimation values ​​in the real-time data subset, and determines whether there is a leak in the gas supply system through the deviation between the means. The present invention improves the accuracy of the warning system, thereby making gas use safer.

Description

Technical Field

[0001] The invention relates to a city gas safety alarm method, in particular to a multi-parameter fusion gas leakage early warning method for gas-using buildings. Background Art

[0002] At present, in addition to conventional gas concentration alarms, there are two other ways of gas safety protection in gas-using buildings in my country, namely pipeline gas self-closing valves and cut-off diaphragm gas meters, which utilize the system flow and pressure conditions during gas use to identify relatively clear abnormal operating conditions such as abnormally small flow, abnormally large flow, constant flow duration, overpressure, underpressure, etc., and provide leakage warning methods.

[0003] Patent publication number CN110566821A, "A method for detecting downstream pipeline leakage by monitoring the pressure state of a pressure regulating device", realizes status monitoring and leakage calculation of the downstream pipeline by calculating the opening flow of the pressure regulator during low-peak gas consumption when a negative pressure wave occurs. Patent publication number CN113606500A, "A method for real-time monitoring of sudden leakage of a gas courtyard pipeline network", achieves the purpose of gas leak identification by constructing a pressure abnormality threshold interval under the healthy operation state of a low-pressure pipeline network and using edge computing equipment to analyze the real-time data of pipeline network pressure. The above patents can only screen for specific pressure abnormalities (negative pressure) and do not consider the interference of pressure fluctuations in the gas supply system of the gas-consuming building itself.

[0004] Patent publication number CN104713610A, "A method for using a safety-cut-off smart gas meter with a self-learning function", solves the problem of users not being able to dynamically determine the maximum gas usage during actual gas usage by automatically learning and updating the maximum gas usage function, thereby achieving safe cutting off of abnormal gas usage by users. Patent publication number CN114878108A, "A method and device for monitoring gas flow", uses the change in absolute pressure of medium and low pressure gas pipelines to calculate the flow increment and decrement of the medium in the pipeline, thereby achieving timely intervention in abnormal phenomena. The above patents can only screen specific flow abnormalities (abnormally large flow, abnormally small flow), and do not consider the interference of flow changes when the normal stove valve is opened at different degrees.

[0005] The above-mentioned control methods and equipment have improved the gas safety of gas-using buildings to a certain extent, but they still cannot effectively solve the leakage warning problem during the normal gas use process (pipeline working pressure fluctuations 1.0~3.0kPa, changes in the valve opening of gas-using equipment) and the coupling of small leaks in the pipeline system, which brings certain risks to the safety of gas-using buildings. Summary of the invention

[0006] The object of the present invention is to overcome the drawbacks of the prior art and provide a gas leakage warning method for gas-using buildings with multi-parameter fusion. By using this method, it is possible to achieve leakage warning when the pipeline system leaks during the normal use of gas in existing gas-using buildings.

[0007] A gas leakage warning method for gas-using buildings with multi-parameter fusion according to the present invention includes the following steps:

[0008] Step 1: When the working pressure of the gas pipeline is 1.0 - 3.0 kPa, change the opening degree of the gas-using equipment valve from small to large in 5 gears within the set time period t0. During this process, within the set time period t0, the computer continuously collects and measures the real-time flow rate and real-time pressure information of the gas meter for the gas consumption of the gas-using equipment, and records the collection time, real-time flow rate, and real-time pressure information of the gas meter in chronological order to form an initial data set.

[0009] Step 2: First, set 4 pressure thresholds P1, P2, P3, P4, and determine the pressure interval division method as P≧P1, P2≦P<P1, P3≦P<P2, P4≦P<P3; where: the value of P1 is 2.8 kPa ± 0.1 kPa, the value of P2 is 2.3 kPa ± 0.1 kPa, the value of P3 is 1.8 kPa ± 0.1 kPa, and the value of P4 is 1 kPa; then, according to the above pressure interval division method, classify the real-time pressure in the initial data set to form four initial data subsets. Each initial data subset includes the real-time pressure within this pressure interval, as well as the sampling time and real-time flow rate corresponding to each real-time pressure.

[0010] Step 3: Continue to collect the real-time flow rate and real-time pressure of the gas meter, divide the interval of the real-time pressure according to the pressure interval division method in Step 2, and supplement the corresponding collection time, real-time flow rate, and real-time pressure into the corresponding initial data subsets respectively to form real-time data subsets. The collection duration is denoted as t1.

[0011] Step 4: Extract the real-time flow rate in the real-time data subset, represent it with flow characteristics, and construct a Poisson random variable model.

[0012] Step 5: Substitute the flow characteristics of the real-time data subset into the Poisson random variable model, and use the Markov chain Monte Carlo algorithm in Bayesian statistics to analyze the distribution of the estimated values of the flow characteristics in the real-time data subset.

[0013] Step 6. Determine the distribution mean of the estimated value of the flow characteristic in the real-time data subset. If the deviation between the means exceeds 15%, it is determined that there is a leak in the gas supply system. The computer sends an alarm signal to the gas meter and shuts down the gas supply system. If the deviation between the means is not greater than 15%, there is no leak in the gas supply system. The computer continues to collect real-time flow and real-time pressure information from the gas meter for a collection period of t1. According to step 3, the information within the collection period is added to the real-time data subset before this moment, the flow characteristic is re-analyzed, and then steps 4 to 6 are repeated.

[0014] A multi-parameter fusion gas leakage warning device for gas-using buildings of the present invention comprises a computer device connected to a signal output end of a gas meter, wherein the computer device comprises the following modules:

[0015] Data collection module: used to collect gas usage information in chronological order, and record the sampling time and the real-time flow and real-time pressure information of the gas meter corresponding to the sampling time;

[0016] Data processing module: used to process the collected data information, wherein the real-time pressure information within the set time period t0 is divided into four initial data subsets according to different pressure intervals; each of the four initial data subsets includes the real-time pressure within the pressure interval, and the sampling time and real-time flow corresponding to each real-time pressure; and after processing time t0, the collection time, real-time flow and real-time pressure information of the gas meter within the collection time length t1 are respectively added to the corresponding initial data subset to form a real-time data subset;

[0017] Data analysis module: used to analyze the traffic characteristics in the real-time data subset, using the Markov chain Monte Carlo algorithm in Bayesian statistics, according to the Poisson random variable model, to analyze and calculate the distribution of the traffic characteristic estimation values ​​corresponding to the traffic characteristics in the real-time data subset and calculate the distribution mean of the traffic characteristic estimation values;

[0018] Leakage warning module: It is used to determine whether there is a leak in the gas supply system through the deviation between the distribution means of the flow characteristic estimation values. If the deviation between the distribution means exceeds 15%, it is determined that there is a leak in the gas supply system, an alarm signal is sent to the gas meter, and the gas supply system is shut down.

[0019] Advantages of the present invention:

[0020] The method and device of the present invention realize the classification of pressure characteristics of the operating state by integrating parameters such as pressure and flow, and effectively reduce the flow overlap problem caused by parallel operations such as working pressure fluctuations and changes in valve opening of gas-using equipment; at the same time, the flow characteristics are analyzed using a Poisson distribution random model to achieve accurate identification of abnormal flow characteristics, and solve the technical problem of leakage warning that may occur during the normal use of the gas pipeline system in the gas-using building. The present invention improves the accuracy of the warning system, thereby making the use of gas safer. BRIEF DESCRIPTION OF THE DRAWINGS

[0021] The accompanying drawings in the specification, which constitute a part of the present invention, are used to provide a further understanding of the present invention. The exemplary embodiments of the present invention and their descriptions are used to explain the present invention and do not constitute improper limitations on the present invention.

[0022] Figure 1 It is a flow chart of a multi-parameter fusion gas leakage early warning method for gas-using buildings of the present invention;

[0023] Figure 2 An initial data set of gas consumption data for gas-using buildings;

[0024] Figure 3 It is a subset of real-time data of gas consumption in gas-consuming buildings (1.8kPa≦P<2.3kPa);

[0025] Figure 4 Illustration of the installation and connection of gas leak warning equipment in gas-using buildings. DETAILED DESCRIPTION

[0026] It should be noted that the following detailed descriptions are illustrative and are intended to provide further explanation of the present application. Unless otherwise specified, all technical and scientific terms used in the present application have the same meanings as those commonly understood by those skilled in the art to which the present application belongs.

[0027] like Figure 1 As shown, a multi-parameter fusion gas leakage early warning method for gas-using buildings of the present invention comprises the following steps:

[0028] Step 1: When the working pressure of the gas pipeline is 1.0~3.0kPa, the opening of the valve of the gas-consuming equipment is changed from small to large by 5 gears within the set time period t0. During this process, the computer continuously collects the real-time flow and real-time pressure information of the gas meter measuring the gas consumption of the gas-consuming equipment within the set time period t0, and records the collection time, real-time flow and real-time pressure information of the gas meter in chronological order to form an initial data set. t0 is usually not less than 15 minutes.

[0029] Step 2: First, set four pressure thresholds P1, P2, P3, and P4, and determine the pressure range division method as P≧P1, P2≦P<P1, P3≦P<P2, P4≦P<P3; where: the value of P1 is 2.8 kPa±0.1 kPa, the value of P2 is 2.3 kPa±0.1 kPa, the value of P3 is 1.8 kPa±0.1 kPa, and the value of P4 is 1 kPa; then, according to the pressure range division method described above, classify the real-time pressure in the initial dataset to form four initial data subsets. Each initial data subset includes the real-time pressure within the pressure range, as well as the sampling time and real-time flow rate corresponding to each real-time pressure.

[0030] Step 3: Continue to collect the real-time flow rate and real-time pressure of the gas meter, divide the range of the real-time pressure according to the pressure range division method in Step 2, and supplement the corresponding collection time, real-time flow rate, and real-time pressure into the corresponding initial data subsets respectively to form real-time data subsets. The collection duration is denoted as t1. Preferably, t1 is not greater than 15 min.

[0031] Step 4: Extract the real-time flow rate in the real-time data subset, represent it with flow characteristics, and construct a Poisson random variable model (see P160 of "Basic Statistics (14th Edition) (written by Mario F. Triola, translated by Qian Chenjiang)").

[0032] The specific construction method is as follows:

[0033] Step 401: Use Q i to represent the extracted flow characteristics, and establish a new time index i for Q i , where i = 1, 2, …, n.

[0034] Step 402: Use the Poisson random variable model to describe the probability distribution of the flow characteristics within a certain period of time. , Q i represents the flow characteristics at time index i, λ represents the estimated value of the flow characteristics before time index i that needs to be obtained, is the factorial of Q i , e is the natural exponent; there are two values for the parameter λ: when t≥t s , λ = λ1, when t<t s , λ = λ2; t s is a certain time index point in time index i. P(X = Q i ) represents the probability when the flow characteristic X is Q i .

[0035] Step 403: Set λ to follow an exponential distribution, λ1~Exp(a), λ2~Exp(a), where a is the reciprocal of the mean value of the flow characteristics in the real-time data subset.

[0036] Step 5. Substitute the traffic characteristics of the real-time data subset into the Poisson random variable model, and use the Markov Chain Monte Carlo algorithm in Bayesian statistics (see "Bayesian Statistics (Edited by Wei Laisheng)" P167~171) to analyze the distribution of the traffic characteristic estimates in the real-time data subset, that is, the distribution of the traffic characteristic estimates λ1 and λ2 in the real-time data subset.

[0037] Step 6. Determine the distribution mean of the estimated value of the flow characteristic in the real-time data subset, that is, calculate the mean of λ1 and λ2 respectively. If the deviation between the means (that is, the deviation between the mean of λ1 and the mean of λ2) exceeds 15%, it is determined that there is a leak in the gas supply system, and the computer sends an alarm signal to the gas meter and shuts down the gas supply system. If the deviation between the means is not greater than 15%, there is no leak in the gas supply system, and the computer continues to collect real-time flow and real-time pressure information from the gas meter for a collection time of t1. According to step 3, the information within the collection time is added to the real-time data subset before this moment, the flow characteristic is re-analyzed, and then steps 4 to 6 are repeated.

[0038] like Figure 4 As shown, a multi-parameter fusion gas leakage warning device for gas-using buildings of the present application includes a computer device connected to a signal output end of a gas meter, and the computer device includes the following modules:

[0039] Data collection module: used to collect gas usage information in chronological order, and record the sampling time and the real-time flow and real-time pressure information of the gas meter corresponding to the sampling time. Preferably, the data collection frequency is not less than 1 time / s.

[0040] Data processing module: used to process the collected data information, including dividing the real-time pressure information within the set time period t0 into four initial data subsets according to different pressure intervals; each of the four initial data subsets includes the real-time pressure within the pressure interval, and the sampling time and real-time flow corresponding to each real-time pressure; and after processing time t0, adding the collection time, real-time flow and real-time pressure information of the gas meter within the collection time t1 to the corresponding initial data subset respectively to form a real-time data subset.

[0041] Data analysis module: used to analyze the traffic characteristics in the real-time data subset. It uses the Markov chain Monte Carlo algorithm in Bayesian statistics and the Poisson random variable model to analyze and calculate the distribution of the traffic characteristic estimation values ​​corresponding to the traffic characteristics in the real-time data subset (the distribution of λ1 and λ2) and calculate the distribution mean of the traffic characteristic estimation values.

[0042] Leakage warning module: It is used to determine whether there is a leak in the gas supply system through the deviation between the distribution means (λ1 and λ2 distribution means) of the flow characteristic estimation value. If the deviation between the distribution means exceeds 15%, it is determined that there is a leak in the gas supply system, an alarm signal is sent to the gas meter, and the gas supply system is shut down.

[0043] Embodiment 1:

[0044] Step 1: When the gas system of a gas-using building is in operation for the first time, t0 is no less than 15 minutes. The initial data set is as follows: Figure 2 shown.

[0045] Step 2: The pressure thresholds are determined as P1=2.8kPa, P2=2.3kPa, P3=1.8kP, and P4=1.0kPa. At this time, the data set category division segments are P≧2.8kPa, 2.3kPa≦P<2.8kPa, 1.8kPa≦P<2.3kPa, and 1.0kPa≦P<1.8kPa.

[0046] Step 3: The gas supply pressure is 1.8kPa≦P<2.3kPa, and each acquisition period is t1=5min. Continuous acquisition is performed to form a real-time data subset of 1.8kPa≦P<2.3kPa, such as Figure 3 .

[0047] Step 4: Extract the real-time flow rate from the real-time data subset with 1.8kPa≦P<2.3kPa and set it as the flow feature Q i , construct a Poisson random variable model , set λ to obey exponential distribution, λ1~Exp(a), λ2~Exp(a).

[0048] Step 5: Substitute the flow characteristics of 1.8kPa≦P<2.3kPa, and use the Markov Chain Monte Carlo algorithm in Bayesian statistics to analyze the distribution of the flow characteristic estimation values ​​λ1 and λ2 in the real-time data subset.

[0049] Step 6: Determine the distribution mean of the estimated flow characteristic values ​​λ1 and λ2. Assuming that each acquisition period is 5 minutes, when the deviation between the distribution means exceeds 15% at the 13th acquisition period, it is determined that there is a leak in the gas supply system. The specific situation is as follows.

[0050]

[0051] The above description is only the preferred embodiment of the present application and is not intended to limit the present application. For those skilled in the art, the present application may have various modifications and variations. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present application shall be included in the protection scope of the present application.

Claims

1. A multi-parameter fusion gas leakage warning method for gas-using buildings, characterized in that It includes the following steps: Step 1: When the working pressure of the gas pipeline is 1.0~3.0 kPa, change the opening degree of the gas appliance valve from small to large in 5 gears within the set time period t0. During this process, the computer continuously collects and measures the real-time flow rate and real-time pressure information of the gas meter for the gas consumption of the gas appliance within the set time period t0, and records the collection time, real-time flow rate and real-time pressure information of the gas meter in chronological order to form an initial data set; Step 2: First, set 4 pressure thresholds P1, P2, P3, P4, and determine the pressure interval division method as P≧P1, P2≦P<P1, P3≦P<P2, P4≦P<P3; where: the value of P1 is 2.8 kPa±0.1 kPa, the value of P2 is 2.3 kPa±0.1 kPa, the value of P3 is 1.8 kPa±0.1 kPa, and the value of P4 is 1 kPa; then, according to the above pressure interval division method, classify the real-time pressure in the initial data set to form four initial data subsets. Each initial data subset includes the real-time pressure within this pressure interval, as well as the sampling time and real-time flow rate corresponding to each real-time pressure; Step 3: Continue to collect the real-time flow rate and real-time pressure of the gas meter, divide the interval of the real-time pressure according to the pressure interval division method in Step 2, and supplement the corresponding collection time, real-time flow rate and real-time pressure into the corresponding initial data subset to form a real-time data subset. The collection duration is recorded as t1; Step 4: Extract the real-time flow rate in the real-time data subset, represent it with flow characteristics, and construct a Poisson random variable model; Step 5: Substitute the flow characteristics of the real-time data subset into the Poisson random variable model, and use the Markov chain Monte Carlo algorithm in Bayesian statistics to analyze the distribution of the estimated values of the flow characteristics in the real-time data subset; Step 6: Determine the distribution mean of the estimated values of the flow characteristics in the real-time data subset. If the deviation between the means exceeds 15% or more, it is determined that there is a leakage in the gas supply system. The computer sends a signal to the gas meter to alarm and closes the gas supply system; if the deviation between the means is not more than 15%, there is no leakage in the gas supply system. The computer continues to collect the real-time flow rate and real-time pressure information of the gas meter, and the collection duration is t1. Then, according to Step 3, supplement the information within this collection duration into the real-time data subset before this moment, re-analyze the flow characteristics, and then repeat Steps 4 to 6.

2. The multi-parameter fusion gas leakage early warning method for gas-using buildings according to claim 1 is characterized by: The specific construction method of the above Poisson random variable model is as follows: Step 401: Use Q i represents the extracted traffic features and is Q i Create a new time index i, i=1,2,…,n, Step 402: Use the Poisson random variable model to describe the probability distribution of traffic characteristics within a certain period of time. , Q i represents the traffic characteristics at time index i, λ represents the estimated value of the traffic characteristics before time index i that needs to be obtained, Q i The factorial of , e is the natural exponent; the parameter λ has two values: when t≥t s When λ=λ1, when t<t s When λ=λ2; t s is a time index point in time index i, P (X=Q i ) means when the flow characteristic X is Q i The probability of the situation when Step 403: Set λ to follow an exponential distribution, λ1~Exp(a), λ2~Exp(a), where a is the reciprocal of the mean value of the flow characteristics in the real-time data subset.

3. A multi-parameter fusion gas leakage warning device for gas-using buildings, applied to the multi-parameter fusion gas leakage warning method for gas-using buildings as claimed in claim 1, comprising a computer device connected to the signal output end of the gas meter, characterized in that: The computer device includes the following modules: Data acquisition module: used to collect gas usage information in chronological order, and record the sampling time and the real-time flow rate and real-time pressure information of the gas meter corresponding to the sampling time; Data processing module: used to process the collected data information, wherein the real-time pressure information within the set time period t0 is divided into four initial data subsets according to different pressure intervals; each of the four initial data subsets includes the real-time pressure within the pressure interval, and the sampling time and real-time flow corresponding to each real-time pressure; and after processing time t0, the collection time, real-time flow and real-time pressure information of the gas meter within the collection time length t1 are respectively added to the corresponding initial data subset to form a real-time data subset; Data analysis module: used to analyze the traffic characteristics in the real-time data subset, using the Markov chain Monte Carlo algorithm in Bayesian statistics, according to the Poisson random variable model, to analyze and calculate the distribution of the traffic characteristic estimation values ​​corresponding to the traffic characteristics in the real-time data subset and calculate the distribution mean of the traffic characteristic estimation values; Leakage warning module: It is used to determine whether there is a leak in the gas supply system through the deviation between the distribution means of the flow characteristic estimation values. If the deviation between the distribution means exceeds 15%, it is determined that there is a leak in the gas supply system, an alarm signal is sent to the gas meter, and the gas supply system is shut down.

4. The multi-parameter fusion gas leakage warning device for gas-using buildings according to claim 3 is characterized by: The data acquisition frequency in the data acquisition module is not less than 1 time / s.

Images