Urban heat supply pipe network early warning management method and platform based on artificial intelligence

Abstract

The invention discloses an urban heat supply pipe network early warning management method and platform based on artificial intelligence, and relates to the technical field of heat supply pipe network early warning management, and the method comprises the following steps: generating a flow signal based on a real-time temperature, a real-time flow, a flow prediction function and a prediction flow threshold range; generating a pressure signal based on the real-time pressure and the pressure threshold range; generating an early warning signal based on the flow signal and the pressure signal; the method is used for solving the problem that in an existing heat supply pipe network early-warning management technology, threshold setting is subjective, the threshold cannot be obtained based on historical data, and consequently the accuracy of early-warning heat supply pipe leakage signals is low.

Description

Technical Field

[0001] The present invention relates to the technical field of early warning management of heating pipe networks, and specifically to an artificial intelligence-based early warning management method and platform for urban heating pipe networks. Background Art

[0002] With the acceleration of urbanization, urban heating pipe networks, as an important part of urban infrastructure, shoulder the heavy responsibility of providing stable heating for urban residents and businesses. These pipes are usually buried underground, facing a complex operating environment. They are easily affected by factors such as soil pressure, groundwater erosion, and external construction, which can lead to cracks, deformation, and leakage in the pipes.

[0003] When monitoring thermal pipes, it is usually necessary to monitor the flow and pressure change data of the thermal pipes to generate thermal pipe leakage warnings. However, in the existing technology, the thresholds of flow and pressure changes are usually set subjectively. The flow rate generated at different ambient temperatures is different, which indicates that the subjectively set flow and pressure thresholds are not accurately set. In other words, the threshold setting in the existing thermal pipe network early warning management technology is relatively subjective, and the threshold value cannot be obtained based on historical data, resulting in low accuracy of the early warning thermal pipe leakage signal. Summary of the Invention

[0004] The present invention aims to solve at least one of the technical problems in the prior art to a certain extent. By generating a flow signal based on real-time temperature, real-time flow, flow prediction function and predicted flow threshold range, generating a pressure signal based on real-time pressure and pressure threshold range, and generating an early warning signal based on the flow signal and the pressure signal, the problem of relatively subjective threshold setting in the existing thermal pipe network early warning management technology and the failure to obtain thresholds based on historical data, resulting in low accuracy of the early warning thermal pipe leakage signal, is solved.

[0005] To achieve the above objectives, this application provides an artificial intelligence-based urban heating network early warning management method, comprising the following steps:

[0006] Acquire monitoring values ​​of flow rates in a first number of thermal pipes at different historical temperatures and mark them as historical flow rates;

[0007] Obtain the relationship function between historical flow and historical temperature, and mark it as flow prediction function;

[0008] Obtaining a predicted flow threshold range based on a flow prediction function;

[0009] Obtain the monitored value of the pressure inside the thermal pipe during normal operation and mark it as historical pressure;

[0010] Obtain pressure threshold range based on historical pressure;

[0011] Obtain real-time temperature, real-time flow and real-time pressure in the thermal pipe;

[0012] generating a flow signal based on real-time temperature, real-time flow, a flow prediction function, and a predicted flow threshold range;

[0013] generating a pressure signal based on the real-time pressure and a pressure threshold range;

[0014] Generate early warning signals based on flow signals and pressure signals.

[0015] Furthermore, obtaining the relationship function between historical flow and historical temperature, which is labeled as flow prediction function, includes the following sub-steps:

[0016] With historical temperature as the horizontal coordinate and historical flow as the vertical coordinate, a plane rectangular coordinate system is established and marked as the flow prediction coordinate system;

[0017] The historical flow rate and the corresponding historical temperature are used as the ordinate and abscissa of the flow coordinate point, and all the flow coordinate points are plotted in the flow prediction coordinate system to obtain a flow prediction scatter plot;

[0018] The flow prediction function is obtained by fitting the function of all flow coordinate points in the flow prediction scatter plot.

[0019] Furthermore, obtaining the predicted traffic threshold range based on the traffic prediction function includes the following sub-steps:

[0020] Substitute the historical temperature into the flow prediction function to obtain the historical predicted flow;

[0021] Calculate the difference between the historical predicted flow and the historical flow, and mark it as the historical flow difference;

[0022] Get the range of all historical flow rate differences at the same historical temperature and mark it as the historical difference range;

[0023] Divide the historical difference range into a equal ranges, marked as divided difference ranges;

[0024] Count the frequency of each partition difference range and mark it as the partition difference frequency;

[0025] A histogram is drawn with the historical flow difference as the horizontal axis, the frequency of the difference as the vertical axis, and the difference range as the histogram interval, which is marked as a difference histogram.

[0026] Furthermore, obtaining the predicted traffic threshold range based on the traffic prediction function further includes the following sub-steps:

[0027] Calculate the sum of all the partition difference frequencies and mark it as the partition sum frequency;

[0028] The calculated division frequency threshold is: H = b*Z / a; where b is the distribution ratio, H is the division frequency threshold, and Z is the total division frequency;

[0029] In the difference histogram, starting from the leftmost partition difference frequency of the difference histogram, determine whether the partition difference frequency is less than the partition frequency threshold. If so, delete the part of the difference histogram corresponding to the partition difference frequency, and stop when it is not. Get the minimum value of the horizontal axis of the partition difference frequency of the difference histogram after deletion, and mark it as the first screening threshold;

[0030] In the difference histogram, starting from the rightmost partition difference frequency of the difference histogram, determine whether the partition difference frequency is less than the partition frequency threshold. If so, delete the part of the difference histogram corresponding to the partition difference frequency, and stop when it is not. Get the maximum value of the horizontal axis of the partition difference frequency of the difference histogram after deletion, and mark it as the second screening threshold;

[0031] The range from the first screening threshold to the second screening threshold is marked as the predicted flow threshold range.

[0032] Furthermore, obtaining the pressure threshold range based on the historical pressure includes the following sub-steps:

[0033] Obtain a second number of historical pressure change values ​​at intervals of the first time, and mark them as historical pressure change values;

[0034] Sort the historical pressure change values ​​from small to large and mark them as Yb1 to Yb i ;

[0035] Determine whether d1*S2 is an integer. If so, set Yb (d1*S2) As the first pressure value; if not, get the integers on the left and right sides of d1*S2, mark them as E1z and E1y respectively, and calculate Yb (E1z) With Yb (E1y) The mean of is taken as the first pressure value; wherein d1 is the first coefficient, the range of d1 is (0, 0.5), and S2 is the second quantity;

[0036] Determine whether 0.5*S2 is an integer. If so, change Yb (0.5*S2) As the second pressure value; if not, get the integers on the left and right sides of 0.5*S2, mark them as E2z and E2y respectively, and calculate Yb (E2z) With Yb (E2y) The mean of is taken as the second pressure value;

[0037] Determine whether d2*S2 is an integer. If so, change Yb (d2*S2) As the third pressure value; if not, get the integers on the left and right sides of d2*S2, mark them as E3z and E3y respectively, and calculate Yb (E3z)With Yb (E3y) The average of is taken as the third pressure value; wherein d2 is the second coefficient, and the range of d2 is (0.5, 1).

[0038] Furthermore, obtaining the pressure threshold range based on the historical pressure also includes the following sub-steps:

[0039] The first pressure range threshold is obtained as: F1 = G2 - 0.5 * (G2 - G1) / (0.5 - d1); where F1 is the first pressure range threshold, G1 is the first pressure value, and G2 is the second pressure value;

[0040] The second pressure range threshold is obtained as: F1 = G2 + 0.5 * (G3 - G2) / (d2 - 0.5); where F2 is the second pressure range threshold, G2 is the second pressure value, and G3 is the third pressure value;

[0041] A range between the first pressure range threshold and the second pressure range threshold is labeled as a pressure threshold range.

[0042] Furthermore, generating a flow signal based on the real-time temperature, the real-time flow, the flow prediction function, and the predicted flow threshold range includes the following sub-steps:

[0043] Substitute the real-time temperature into the flow prediction function to obtain the real-time predicted flow;

[0044] Calculate the difference between the real-time predicted flow and the real-time flow, and mark it as the real-time flow difference;

[0045] Determine whether the real-time flow difference is within the predicted flow threshold range. If so, generate a flow normal signal; if not, generate a flow abnormal signal.

[0046] Furthermore, generating a pressure signal based on the real-time pressure and the pressure threshold range includes the following sub-steps:

[0047] Obtain the real-time pressure change value at the first time of each interval, and mark it as the real-time pressure change value;

[0048] Determine whether the real-time pressure change value is within the pressure threshold range. If so, generate a normal pressure signal; if not, generate an abnormal pressure signal.

[0049] Furthermore, generating an early warning signal based on the flow signal and the pressure signal includes the following steps:

[0050] If an abnormal flow signal and a normal pressure signal are generated, a first-level warning signal will be generated;

[0051] If a normal flow signal and an abnormal pressure signal are generated, a secondary warning signal will be generated;

[0052] If abnormal flow signal and abnormal pressure signal are generated, a third-level warning signal will be generated.

[0053] The present application also provides an artificial intelligence-based urban heat pipe network early warning management platform, including: a historical flow acquisition module, a function acquisition module, a flow threshold acquisition module, a historical pressure acquisition module, a pressure threshold acquisition module, a real-time data acquisition module, a flow signal generation module, a pressure signal generation module, and an early warning module;

[0054] The historical flow acquisition module is used to obtain monitoring values ​​of the flow in the first number of thermal pipes at different historical temperatures, and mark them as historical flow;

[0055] The function acquisition module is used to obtain the relationship function between historical flow and temperature, which is marked as flow prediction function;

[0056] The flow threshold acquisition module is used to obtain a predicted flow threshold range based on a flow prediction function;

[0057] The historical pressure acquisition module is used to obtain the monitoring value of the pressure in the thermal pipe during normal operation, which is marked as historical pressure;

[0058] The pressure threshold acquisition module is used to acquire a pressure threshold range based on historical pressure;

[0059] The real-time data acquisition module is used to obtain the real-time temperature, real-time flow and real-time pressure in the thermal pipe;

[0060] The flow signal generating module is used to generate a flow signal based on the real-time temperature, the real-time flow, the flow prediction function and the predicted flow threshold range;

[0061] The pressure signal generating module is used to generate a pressure signal based on the real-time pressure and the pressure threshold range;

[0062] The early warning module is used to generate an early warning signal based on the flow signal and the pressure signal.

[0063] Beneficial effects of the present invention: The present invention generates a flow signal based on real-time temperature, real-time flow, a flow prediction function, and a predicted flow threshold range; generates a pressure signal based on real-time pressure and a pressure threshold range; and generates an early warning signal based on the flow signal and the pressure signal. Advantageously, the present invention can generate appropriate predicted flow threshold ranges and pressure threshold ranges based on historical data and real-time conditions, thereby improving the accuracy of the early warning thermal pipe leakage signal.

[0064] The present invention obtains the pressure threshold range based on historical pressure. The advantage is that it not only obtains the appropriate screening pressure threshold range based on historical data, but also can finely process the data range to filter out abnormal data, thereby improving the accuracy of the early warning thermal pipe leakage signal. BRIEF DESCRIPTION OF THE DRAWINGS

[0065] Figure 1 It is a principle block diagram of the system of the present invention;

[0066] Figure 2 is a schematic diagram of the flow prediction function of the present invention;

[0067] Figure 3 is a schematic diagram of a difference histogram of the present invention;

[0068] Figure 4 is a schematic diagram of the first screening threshold and the second screening threshold of the present invention;

[0069] Figure 5 Flow chart of the steps of the method of the present invention. DETAILED DESCRIPTION

[0070] The following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative efforts are within the scope of protection of the present invention.

[0071] Example 1, please refer to Figure 1 As shown, the present application provides an artificial intelligence-based urban thermal pipe network early warning management platform, including a historical flow acquisition module, a function acquisition module, a flow threshold acquisition module, a historical pressure acquisition module, a pressure threshold acquisition module, a real-time data acquisition module, a flow signal generation module, a pressure signal generation module and an early warning module;

[0072] The historical flow acquisition module is used to obtain a first number of monitoring values ​​of the flow in the thermal pipe at different historical temperatures, which are marked as historical flows; the first number is set to 200, and the first number can be increased according to the collected data, and accurate corresponding monitoring values ​​of the flow in the thermal pipe are obtained based on 200;

[0073] The function acquisition module is used to obtain the relationship function between historical flow and temperature, which is marked as flow prediction function;

[0074] The function acquisition module is configured with a function acquisition strategy, which includes:

[0075] With historical temperature as the horizontal coordinate and historical flow as the vertical coordinate, a plane rectangular coordinate system is established and marked as the flow prediction coordinate system;

[0076] The historical flow rate and the corresponding historical temperature are used as the ordinate and abscissa of the flow coordinate point, and all the flow coordinate points are plotted in the flow prediction coordinate system to obtain a flow prediction scatter plot;

[0077] The flow prediction function is obtained by fitting the function of all flow coordinate points in the flow prediction scatter plot.

[0078] In practical applications, please refer to Figure 2 As shown, the flow prediction function is obtained. Since the use of the thermal pipe is affected by temperature, that is, the lower the temperature, the higher the required heat supply, and the higher the flow in the thermal pipe, the temperature is related to the flow of the thermal pipe, and a function can be fitted;

[0079] The flow threshold acquisition module is used to obtain the predicted flow threshold range based on the flow prediction function;

[0080] The traffic threshold acquisition module is configured with a histogram establishment strategy, which includes:

[0081] Substitute the historical temperature into the flow prediction function to obtain the historical predicted flow;

[0082] Calculate the difference between the historical predicted flow and the historical flow, and mark it as the historical flow difference;

[0083] Get the range of all historical flow rate differences at the same historical temperature and mark it as the historical difference range;

[0084] Divide the historical difference range into a equal ranges, marked as divided difference ranges. The setting of a makes the historical flow difference appear in the divided difference ranges with different distributions, which can further refine the range. However, for the convenience of calculation, the setting should not be too large, for example, a is set to 8.

[0085] Count the frequency of each partition difference range and mark it as the partition difference frequency;

[0086] Draw a histogram with the historical flow difference as the horizontal axis, the frequency of the difference as the vertical axis, and the difference range as the histogram interval, marked as the difference histogram;

[0087] In practical applications, please refer to Figure 3 As shown, the difference histogram obtained;

[0088] The traffic threshold acquisition module is configured with a traffic threshold acquisition strategy, which includes:

[0089] Calculate the sum of all the partition difference frequencies and mark it as the partition sum frequency;

[0090] The division frequency threshold is calculated as: H = b*Z / a; where b is the distribution ratio, H is the division frequency threshold, and Z is the division total frequency; b is set to filter out the division difference frequency that accounts for a smaller proportion than the rest, and Z / a is the average frequency of each division difference range, so b is less than 1, for example, b is 0.1;

[0091] In the difference histogram, starting from the leftmost partition difference frequency of the difference histogram, determine whether the partition difference frequency is less than the partition frequency threshold. If so, delete the part of the difference histogram corresponding to the partition difference frequency, and stop when it is not. Get the minimum value of the horizontal axis of the partition difference frequency of the difference histogram after deletion, and mark it as the first screening threshold;

[0092] In the difference histogram, starting from the rightmost partition difference frequency of the difference histogram, determine whether the partition difference frequency is less than the partition frequency threshold. If so, delete the part of the difference histogram corresponding to the partition difference frequency, and stop when it is not. Get the maximum value of the horizontal axis of the partition difference frequency of the difference histogram after deletion, and mark it as the second screening threshold;

[0093] Marking the range from the first screening threshold to the second screening threshold as a predicted flow threshold range;

[0094] See also Figure 3 and Figure 4 As shown, when Z is the total frequency of the division, it is 200, and the division frequency threshold is calculated as: H = 0.1*200 / 8 = 2.5, so the first screening threshold is -6m 3 / h, the second screening threshold is 6m 3 / h, that is, the predicted flow threshold range -6m 3 / h to 6m 3 / h.

[0095] The historical pressure acquisition module is used to obtain the monitoring value of the pressure in the thermal pipe during normal operation, which is marked as historical pressure;

[0096] The pressure threshold acquisition module is used to obtain the pressure threshold range based on historical pressure;

[0097] The pressure threshold acquisition module is configured with a pressure value acquisition strategy, which includes:

[0098] Obtain a second number of historical pressure change values ​​at each first time interval, and mark them as historical pressure change values; the second number is set to 200, and the second number can be the same as the first number;

[0099] Sort the historical pressure change values ​​from small to large and mark them as Yb1 to Yb i ;

[0100] Determine whether d1*S2 is an integer. If so, set Yb (d1*S2) As the first pressure value; if not, get the integers on the left and right sides of d1*S2, mark them as E1z and E1y respectively, and calculate Yb (E1z) With Yb (E1y) The mean of is taken as the first pressure value; d1 is the first coefficient, the range of d1 is (0, 0.5), and S2 is the second quantity; a smaller historical pressure change value is selected, so the range of d1 is (0, 0.5), and the middle value between 0 and 0.5 is representative, that is, 0.25, so d1 is preferably set to 0.25;

[0101] Determine whether 0.5*S2 is an integer. If so, change Yb (0.5*S2) As the second pressure value; if not, get the integers on the left and right sides of 0.5*S2, mark them as E2z and E2y respectively, and calculate Yb (E2z) With Yb (E2y) The mean of the two values ​​is taken as the second pressure value; the middle value is selected, so S2 is multiplied by 0.5;

[0102] Determine whether d2*S2 is an integer. If so, change Yb (d2*S2) As the third pressure value; if not, get the integers on the left and right sides of d2*S2, mark them as E3z and E3y respectively, and calculate Yb (E3z) With Yb (E3y) The average of the three pressure values ​​is taken as the third pressure value; d2 is the second coefficient, and the range of d2 is (0.5, 1); a larger historical pressure change value is selected, so the range of d2 is (0.5, 1), and the middle value between 0.5 and 1 is representative, that is, 0.75, so d2 is preferably set to 0.75;

[0103] In practical applications, if 0.25*200 is an integer, then Yb (50) The corresponding specific value is -0.25MPa, that is, -0.26MPa is the first pressure value, and 0.5*200 is judged to be an integer. (100) The corresponding specific value is 0.02MPa, which is used as the second pressure value. It is determined that 0.75*200 is an integer. (150) The corresponding specific value is 0.24MPa, that is, 0.24MPa is the third pressure value;

[0104] The pressure threshold acquisition module is configured with a pressure threshold acquisition strategy, which includes:

[0105] The threshold value of the first pressure range is obtained as follows: F1 = G2 - 0.5 * (G2 - G1) / (0.5 - d1); wherein F1 is the threshold value of the first pressure range, G1 is the first pressure value, and G2 is the second pressure value; when the thermal pipe is working stably, the internal pressure hardly changes, so the historical pressure change values ​​are evenly distributed around 0. The smaller values ​​of the historical pressure change values ​​are regarded as evenly distributed, and the minimum value of the historical pressure change values ​​is obtained by using the first pressure value and the second pressure value, that is, regarded as the minimum value of the historical pressure change values. Instead of directly obtaining the minimum value of the historical pressure change values, it is possible to obtain abnormally small data when obtaining data, so that the threshold value is obtained more accurately;

[0106] The second pressure range threshold is obtained as follows: F1 = G2 + 0.5 * (G3 - G2) / (d2 - 0.5); wherein F2 is the second pressure range threshold, G2 is the second pressure value, and G3 is the third pressure value; the majority of historical pressure change values ​​are considered to be uniformly distributed, and the maximum value of the historical pressure change value is obtained by using the second pressure value and the third pressure value, that is, regarded as the maximum value of the historical pressure change value. The maximum value of the historical pressure change value is not obtained directly, so that abnormally large data can be obtained when obtaining data, making the threshold acquisition more accurate;

[0107] marking a range between the first pressure range threshold and the second pressure range threshold as a pressure threshold range;

[0108] In actual application, the first pressure range threshold is obtained as: F1 = G2-0.5*(G2-G1) / (0.5-d1) = 0.02-0.5*(0.02+0.25) / (0.5-0.25) = -0.52MPa, and the second pressure range threshold is obtained as: F1 = G2+0.5*(G3-G2) / (d2-0.5) = 0.02+0.5*(0.24-0.02) / (0.75-0.5) = 0.46MPa, that is, the pressure threshold range is: [-0.52MPa, = 0.46MPa].

[0109] The real-time data acquisition module is used to obtain the real-time temperature, real-time flow and real-time pressure in the thermal pipe;

[0110] The flow signal generation module is used to generate a flow signal based on real-time temperature, real-time flow, flow prediction function and predicted flow threshold range;

[0111] The traffic signal generation module is configured with a traffic signal generation strategy, which includes:

[0112] Substitute the real-time temperature into the flow prediction function to obtain the real-time predicted flow;

[0113] Calculate the difference between the real-time predicted flow and the real-time flow, and mark it as the real-time flow difference;

[0114] Determine whether the real-time flow difference is within the predicted flow threshold range. If so, generate a flow normal signal; if not, generate a flow abnormal signal;

[0115] In practical applications, please refer to Figure 2 As shown, for example, if the real-time temperature is -15°C, the real-time predicted flow rate is 360m when it is substituted into the flow prediction function. 3 / h, the real-time traffic is 340m 3 / h, the real-time flow difference is 20m 3 / h, the real-time flow difference is not -6m 3 / h to 6m 3 / h, an abnormal flow signal is generated.

[0116] The pressure signal generating module is used to generate a pressure signal based on the real-time pressure and the pressure threshold range;

[0117] The pressure signal generation module is configured with a pressure signal generation strategy, which includes:

[0118] Obtain the real-time pressure change value at the first time of each interval, and mark it as the real-time pressure change value;

[0119] Determine whether the real-time pressure change value is within the pressure threshold range. If so, generate a normal pressure signal; if not, generate an abnormal pressure signal;

[0120] In practical applications, for example, if the real-time pressure change value is -0.9 MPa, a pressure abnormality signal is generated.

[0121] The early warning module is used to generate an early warning signal based on the flow signal and the pressure signal;

[0122] The early warning module is configured with early warning strategies, which include:

[0123] If an abnormal flow signal and a normal pressure signal are generated, a first-level warning signal will be generated;

[0124] If a normal flow signal and an abnormal pressure signal are generated, a secondary warning signal will be generated;

[0125] If abnormal flow signal and abnormal pressure signal are generated, a three-level warning signal will be generated;

[0126] In actual applications, if an abnormal flow signal and an abnormal pressure signal are generated, a third-level warning signal is generated. This is because the abnormal flow signal is greatly affected by user usage, which may be due to damage to the thermal pipe or increased usage by users. The change in pressure in a short period of time may be due to damage to the thermal pipe, resulting in internal pressure changes. Therefore, when an abnormal flow signal and a normal pressure signal are generated, a first-level warning signal is generated. When a normal flow signal and an abnormal pressure signal are generated, a second-level warning signal is generated. From the first-level warning signal to the third-level warning signal, the urgency increases, indicating that the probability of damage increases.

[0127] Example 2, please refer to Figure 5 As shown, this application provides an artificial intelligence-based urban heating pipe network early warning management method, including the following steps:

[0128] Step S1, obtaining monitoring values ​​of flow rates in a first number of thermal pipes at different historical temperatures, and marking them as historical flow rates;

[0129] Step S2, obtaining a relationship function between historical flow and historical temperature, marked as a flow prediction function; Step S2 includes the following sub-steps:

[0130] Step S201: Establish a plane rectangular coordinate system with historical temperature as the horizontal coordinate and historical flow rate as the vertical coordinate, and mark it as the flow rate prediction coordinate system;

[0131] Step S202: Using historical flow rates and corresponding historical temperatures as the ordinates and abscissas of flow coordinate points, all flow coordinate points are plotted in a flow prediction coordinate system to obtain a flow prediction scatter plot;

[0132] Step S203: Perform function fitting on all the flow coordinate points in the flow prediction scatter plot to obtain a flow prediction function.

[0133] Step S3, obtaining a predicted flow threshold range based on the flow prediction function; Step S3 includes the following sub-steps:

[0134] Step S301, substituting historical temperature into the flow prediction function to obtain historical predicted flow;

[0135] Step S302, calculating the difference between the historical predicted flow and the historical flow, and marking it as the historical flow difference;

[0136] Step S303, obtaining the range of all historical flow rate differences at the same historical temperature, and marking it as the historical difference range;

[0137] Step S304, dividing the historical difference range into a equal ranges, marked as divided difference ranges;

[0138] Step S305, counting the frequency of each partition difference range, and marking it as the partition difference frequency;

[0139] Step S306: Draw a histogram with the historical flow difference as the horizontal axis, the frequency of the difference as the vertical axis, and the difference range as the histogram interval, and mark it as a difference histogram.

[0140] Step S307, calculating the sum of all the partition difference frequencies and marking it as the partition total frequency;

[0141] Step S308, calculating the partition frequency threshold as: H = b*Z / a; where b is the distribution ratio, H is the partition frequency threshold, and Z is the partition total frequency;

[0142] Step S309: Starting from the leftmost partition difference frequency in the difference histogram and proceeding to the right, determine whether the partition difference frequency is less than the partition frequency threshold. If so, delete the portion of the difference histogram corresponding to the partition difference frequency, and stop until it is not. Obtain the minimum value of the horizontal axis of the partition difference frequency of the difference histogram after deletion, and mark it as the first screening threshold.

[0143] Step S310: Starting from the rightmost partition difference frequency in the difference histogram, determine whether the partition difference frequency is less than the partition frequency threshold. If so, delete the portion of the difference histogram corresponding to the partition difference frequency. This process stops when it is not, and obtains the maximum value of the horizontal axis of the partition difference frequency in the difference histogram after deletion, marking it as the second screening threshold.

[0144] Step S311: Mark the range from the first screening threshold to the second screening threshold as the predicted flow threshold range.

[0145] Step S4, obtaining the monitored value of the pressure in the thermal pipe during normal operation, and marking it as historical pressure;

[0146] Step S5, obtaining a pressure threshold range based on historical pressure; Step S5 includes the following sub-steps:

[0147] Step S501, obtaining a second number of historical pressure change values ​​at first time intervals, and marking them as historical pressure change values;

[0148] Step S502: sort the historical pressure change values ​​from small to large and mark them as Yb1 to Yb i ;

[0149] Step S503, determine whether d1*S2 is an integer, if so, change Yb (d1*S2) As the first pressure value; if not, get the integers on the left and right sides of d1*S2, mark them as E1z and E1y respectively, and calculate Yb (E1z) With Yb(E1y) The mean of is taken as the first pressure value; wherein d1 is the first coefficient, the range of d1 is (0, 0.5), and S2 is the second quantity;

[0150] Step S504, determine whether 0.5*S2 is an integer, if so, change Yb (0.5*S2) As the second pressure value; if not, get the integers on the left and right sides of 0.5*S2, mark them as E2z and E2y respectively, and calculate Yb (E2 z ) With Yb (E2y) The mean of is taken as the second pressure value;

[0151] Step S505, determine whether d2*S2 is an integer, if so, change Yb (d2*S2) As the third pressure value; if not, get the integers on the left and right sides of d2*S2, mark them as E3z and E3y respectively, and calculate Yb (E3z) With Yb (E3y) The average of is taken as the third pressure value; wherein d2 is the second coefficient, and the range of d2 is (0.5, 1).

[0152] Step S506 , obtaining the first pressure range threshold value as: F1 = G2 - 0.5*(G2 - G1) / (0.5 - d1); wherein F1 is the first pressure range threshold value, G1 is the first pressure value, and G2 is the second pressure value;

[0153] Step S507 , obtaining the second pressure range threshold value as: F1=G2+0.5*(G3-G2) / (d2-0.5); wherein F2 is the second pressure range threshold value, G2 is the second pressure value, and G3 is the third pressure value;

[0154] Step S508: Mark the range between the first pressure range threshold and the second pressure range threshold as a pressure threshold range.

[0155] Step S6, obtaining the real-time temperature, real-time flow rate and real-time pressure in the thermal pipe;

[0156] Step S7, generating a flow signal based on the real-time temperature, the real-time flow, the flow prediction function, and the predicted flow threshold range; Step S7 includes the following sub-steps:

[0157] Step S701, substituting the real-time temperature into the flow prediction function to obtain the real-time predicted flow;

[0158] Step S702: Calculate the difference between the real-time predicted flow rate and the real-time flow rate, and mark it as the real-time flow rate difference;

[0159] Step S703: determine whether the real-time flow difference is within the predicted flow threshold range. If so, generate a flow normal signal; if not, generate a flow abnormal signal.

[0160] Step S8, generating a pressure signal based on the real-time pressure and the pressure threshold range; Step S8 includes the following sub-steps:

[0161] Step S801, obtaining a real-time pressure change value at each first time interval, and marking it as a real-time pressure change value;

[0162] Step S802 , determining whether the real-time pressure change value is within the pressure threshold range, if so, generating a normal pressure signal, otherwise generating an abnormal pressure signal.

[0163] Step S9: generating an early warning signal based on the flow signal and the pressure signal; Step S9 includes the following sub-steps:

[0164] Step S901: If a flow abnormality signal and a pressure normal signal are generated, a first-level warning signal is generated;

[0165] Step S902: If a normal flow signal and an abnormal pressure signal are generated, a secondary warning signal is generated;

[0166] Step S903: If a flow abnormality signal and a pressure abnormality signal are generated, a third-level warning signal is generated.

[0167] Those skilled in the art will appreciate that embodiments of the present invention may be provided as methods, systems, or computer program products. Therefore, the present invention may take the form of a fully hardware embodiment, a fully software embodiment, or an embodiment combining software and hardware. Furthermore, the present invention may take the form of a computer program product implemented on one or more computer-usable storage media containing computer-usable program code. The storage medium may be implemented by any type of volatile or non-volatile storage device or a combination thereof, such as static random access memory (SRAM), electrically erasable programmable read-only memory (EEPROM), erasable programmable read-only memory (EPROM), programmable read-only memory (PROM), read-only memory (ROM), magnetic storage, flash memory, magnetic disk, or optical disk. These computer program instructions may also be stored in a computer readable memory that can direct a computer or other programmable data processing device to work in a specific manner, so that the instructions stored in the computer readable memory produce an article of manufacture comprising an instruction device, which implements the process Figure 1 a process or multiple processes and / or boxes Figure 1 The function specified in one or more boxes.

[0168] In the embodiments provided in this application, it should be understood that the disclosed devices and methods can be implemented in other ways. The device embodiments described above are merely schematic. For example, the division of the units is only a logical function division. There may be other division methods in actual implementation. For example, multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the mutual coupling or direct coupling or communication connection shown or discussed can be through some communication interface, indirect coupling or communication connection of devices or units, which can be electrical, mechanical or other forms.

Claims

1. An artificial intelligence-based early warning management method for urban heating pipe networks, characterized in that: The steps include: Acquire monitoring values ​​of flow rates in a first number of thermal pipes at different historical temperatures and mark them as historical flow rates; Obtain the relationship function between historical flow and historical temperature, and mark it as flow prediction function; Obtaining a predicted flow threshold range based on a flow prediction function; Obtain the monitored value of the pressure inside the thermal pipe during normal operation and mark it as historical pressure; Obtain pressure threshold range based on historical pressure; Obtain real-time temperature, real-time flow and real-time pressure in the thermal pipe; generating a flow signal based on real-time temperature, real-time flow, a flow prediction function, and a predicted flow threshold range; generating a pressure signal based on the real-time pressure and a pressure threshold range; Generate early warning signals based on flow signals and pressure signals.

2. The artificial intelligence-based early warning management method for urban heating pipe networks according to claim 1 is characterized in that: Obtaining the relationship function between historical flow and historical temperature, labeled as flow prediction function, includes the following sub-steps: With historical temperature as the horizontal coordinate and historical flow as the vertical coordinate, a plane rectangular coordinate system is established and marked as the flow prediction coordinate system; The historical flow rate and the corresponding historical temperature are used as the ordinate and abscissa of the flow coordinate point, and all the flow coordinate points are plotted in the flow prediction coordinate system to obtain a flow prediction scatter plot; The flow prediction function is obtained by fitting the function of all flow coordinate points in the flow prediction scatter plot.

3. The artificial intelligence-based early warning management method for urban heating pipe networks according to claim 2 is characterized in that: Obtaining the predicted traffic threshold range based on the traffic prediction function includes the following sub-steps: Substitute the historical temperature into the flow prediction function to obtain the historical predicted flow; Calculate the difference between the historical predicted flow and the historical flow, and mark it as the historical flow difference; Get the range of all historical flow rate differences at the same historical temperature and mark it as the historical difference range; Divide the historical difference range into a equal ranges, marked as divided difference ranges; Count the frequency of each partition difference range and mark it as the partition difference frequency; A histogram is drawn with the historical flow difference as the horizontal axis, the frequency of the difference as the vertical axis, and the difference range as the histogram interval, which is marked as a difference histogram.

4. The artificial intelligence-based early warning management method for urban heating pipe networks according to claim 3 is characterized in that: Obtaining the predicted traffic threshold range based on the traffic prediction function also includes the following sub-steps: Calculate the sum of all the partition difference frequencies and mark it as the partition sum frequency; The calculated frequency threshold for division is: H = b*Z / a; where b is the distribution ratio, H is the frequency threshold for division, and Z is the total frequency for division; In the difference histogram, starting from the leftmost partition difference frequency of the difference histogram, determine whether the partition difference frequency is less than the partition frequency threshold. If so, delete the part of the difference histogram corresponding to the partition difference frequency, and stop when it is not. Get the minimum value of the horizontal axis of the partition difference frequency of the difference histogram after deletion, and mark it as the first screening threshold; In the difference histogram, starting from the rightmost partition difference frequency of the difference histogram, determine whether the partition difference frequency is less than the partition frequency threshold. If so, delete the part of the difference histogram corresponding to the partition difference frequency, and stop when it is not. Get the maximum value of the horizontal axis of the partition difference frequency of the difference histogram after deletion, and mark it as the second screening threshold; The range from the first screening threshold to the second screening threshold is marked as the predicted flow threshold range.

5. The artificial intelligence-based early warning management method for urban heating pipe networks according to claim 4 is characterized in that: Obtaining the pressure threshold range based on historical pressure includes the following sub-steps: Obtain a second number of historical pressure change values ​​at intervals of the first time, and mark them as historical pressure change values; Sort the historical pressure change values ​​from small to large and mark them as Yb1 to Yb i ; Determine whether d1*S2 is an integer. If so, set Yb (d1*S2) As the first pressure value; if not, get the integers on the left and right sides of d1*S2, mark them as E1z and E1y respectively, and calculate Yb (E1z) With Yb (E1y) The mean of is taken as the first pressure value; wherein d1 is the first coefficient, the range of d1 is (0, 0.5), and S2 is the second quantity; Determine whether 0.5*S2 is an integer. If so, change Yb (0.5*S2) As the second pressure value; if not, get the integers on the left and right sides of 0.5*S2, mark them as E2z and E2y respectively, and calculate Yb (E2z) With Yb (E2y) The mean of is taken as the second pressure value; Determine whether d2*S2 is an integer. If so, change Yb (d2*S2) As the third pressure value; if not, get the integers on the left and right sides of d2*S2, mark them as E3z and E3y respectively, and calculate Yb (E3z) With Yb (E3y) The average of is taken as the third pressure value; wherein d2 is the second coefficient, and the range of d2 is (0.5, 1).

6. The artificial intelligence-based early warning management method for urban heating pipe networks according to claim 5 is characterized in that: Obtaining the pressure threshold range based on historical pressure also includes the following sub-steps: The first pressure range threshold is obtained as: F1 = G2 - 0.5*(G2 - G1) / (0.5 - d1); Where F1 is the first pressure range threshold, G1 is the first pressure value, and G2 is the second pressure value; The second pressure range threshold is obtained as: F1 = G2 + 0.5 * (G3 - G2) / (d2 - 0.5); where F2 is the second pressure range threshold, G2 is the second pressure value, and G3 is the third pressure value; A range between the first pressure range threshold and the second pressure range threshold is labeled as a pressure threshold range.

7. The artificial intelligence-based early warning management method for urban heating pipe networks according to claim 6 is characterized in that: Generating a flow signal based on real-time temperature, real-time flow, flow prediction function, and predicted flow threshold range includes the following sub-steps: Substitute the real-time temperature into the flow prediction function to obtain the real-time predicted flow; Calculate the difference between the real-time predicted flow and the real-time flow, and mark it as the real-time flow difference; Determine whether the real-time flow difference is within the predicted flow threshold range. If so, generate a flow normal signal; if not, generate a flow abnormal signal.

8. The artificial intelligence-based early warning management method for urban heating pipe networks according to claim 7 is characterized in that: Generating a pressure signal based on the real-time pressure and the pressure threshold range includes the following sub-steps: Obtain the real-time pressure change value at the first time of each interval, and mark it as the real-time pressure change value; Determine whether the real-time pressure change value is within the pressure threshold range. If so, generate a normal pressure signal; if not, generate an abnormal pressure signal.

9. The artificial intelligence-based early warning management method for urban heating pipe networks according to claim 8 is characterized in that: Generating an early warning signal based on the flow signal and the pressure signal includes the following steps: If an abnormal flow signal and a normal pressure signal are generated, a first-level warning signal will be generated; If a normal flow signal and an abnormal pressure signal are generated, a secondary warning signal will be generated; If abnormal flow signal and abnormal pressure signal are generated, a third-level warning signal will be generated.

10. An artificial intelligence-based urban heating pipe network early warning management platform, used to implement the artificial intelligence-based urban heating pipe network early warning management method according to any one of claims 1 to 9, characterized in that: It includes a historical flow acquisition module, a function acquisition module, a flow threshold acquisition module, a historical pressure acquisition module, a pressure threshold acquisition module, a real-time data acquisition module, a flow signal generation module, a pressure signal generation module and an early warning module; The historical flow acquisition module is used to obtain monitoring values ​​of the flow in the first number of thermal pipes at different historical temperatures, and mark them as historical flow; The function acquisition module is used to obtain the relationship function between historical flow and temperature, which is marked as flow prediction function; The flow threshold acquisition module is used to obtain a predicted flow threshold range based on a flow prediction function; The historical pressure acquisition module is used to obtain the monitoring value of the pressure in the thermal pipe during normal operation, which is marked as historical pressure; The pressure threshold acquisition module is used to acquire a pressure threshold range based on historical pressure; The real-time data acquisition module is used to obtain the real-time temperature, real-time flow and real-time pressure in the thermal pipe; The flow signal generating module is used to generate a flow signal based on the real-time temperature, the real-time flow, the flow prediction function and the predicted flow threshold range; The pressure signal generating module is used to generate a pressure signal based on the real-time pressure and the pressure threshold range; The early warning module is used to generate an early warning signal based on the flow signal and the pressure signal.

Images