A data monitoring and alarm method for each point in a building with self-learning function

By setting up digital pressure gauge and pressure reducing valve in the building, and automatically adjusting the alarm threshold using a self-learning algorithm, the false alarm problem caused by manual preset alarm values ​​in the existing technology is solved, and more accurate and timely acquisition of alarm information is achieved. Reduced adjustment workload.

CN115985060BActive Publication Date: 2024-11-08SHANGHAI MINGKONG SENSING TECH CO LTD
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202211135284.6
Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
Filing Date
2022-09-19
Publication Date
2024-11-08
Estimated Expiration
2042-09-19

AI Technical Summary

Technical Problem

The alarm value of the existing building data monitoring system needs to be manually preset, which leads to different fluctuations caused by differences in on-site locations, which can easily lead to false alarms and increase the adjustment workload.

Method used

Design a self-learning function of data monitoring and alarm method for each point in the building. By setting up a digital pressure gauge and pressure reducing valve in the building, and using a self-learning algorithm to calculate the alarm threshold and fluctuation difference, it realizes automatic adjustment of alarm settings and reduces Abnormal data interference, and accurate alarm information is obtained in a timely manner.

Benefits of technology

Reduce false alarms, improve the accuracy and timeliness of alarm information, reduce personnel adjustment workload, and improve the automation level of the system.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN115985060B_ABST
    Figure CN115985060B_ABST
Patent Text Reader

Abstract

The present invention relates to the field of building monitoring technology, and specifically to a method for monitoring and alarming data at various points in a building with a self-learning function. It is characterized in that it includes the following steps: S1, initializing a digital pressure gauge, monitoring the water pressure of each floor in the building, and sampling digital pressure gauge data; S2, entering the alarm algorithm judgment process, S3, according to the judgment of step S2, determining whether to trigger an alarm, if the alarm is triggered, starting the alarm reporting process, and saving and displaying the current pressure data, if the alarm is not triggered, saving and displaying the current pressure data, and repeating steps S1 to S3; S4, after the specified time, ending the scheduled collection task. Compared with the prior art, the method combines the data of various points in the building for self-learning, monitoring and alarming, reduces the interference of abnormal data, and obtains accurate alarm information in a timely manner.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The invention relates to the technical field of building monitoring, in particular to a method for monitoring and alarming data at various points in a building with a self-learning function. Background Art

[0002] At present, the alarm value of building data monitoring is manually preset before the equipment is installed. However, this pre-setting method has disadvantages: the difference in the on-site location causes different fluctuation values, and the pre-set alarm value will be set unreasonably, resulting in false alarms and increasing the number of personnel on-site re-setting and adjustment work. Summary of the invention

[0003] In order to overcome the deficiencies of the prior art, the present invention provides a method for monitoring and alarming data at various points in a building with a self-learning function.

[0004] In order to achieve the above purpose, a data monitoring and alarm method for each point in a building with a self-learning function is designed, which is characterized by: a digital pressure gauge is set on each floor of the building, pressure reducing valves are set between adjacent floors, a water tank is set on the top floor, and a water pump room is set on the bottom floor. The method includes the following steps:

[0005] S1, initialization of digital pressure gauge, monitoring water pressure of each floor in the building, sampling digital pressure gauge data;

[0006] S2, enter the alarm algorithm judgment process,

[0007] S3, according to the judgment of step S2, whether an alarm is triggered is determined. If an alarm is triggered, the alarm reporting process is started, and the current pressure data is saved and displayed. If an alarm is not triggered, the current pressure data is saved and displayed, and steps S1 to S3 are repeated;

[0008] S4, after the specified time, the scheduled collection task is terminated;

[0009] The specific method of the alarm algorithm process in step S2 is as follows:

[0010] S2-1, the digital pressure gauge obtains the current collected data and caches it into the collected data group;

[0011] S2-2, determine whether there are 10 groups of parameters in the collected data group, and proceed to step S2-3, and at the same time determine whether it is in a state of no high or low alarm, and proceed to step S2-7;

[0012] S2-3, if there are 10 sets of parameters in the collected data group, determine whether it is currently in the water release alarm state and enter step S2-4. If there are not 10 sets of parameters, mark it as no alarm and end the alarm algorithm judgment process;

[0013] S2-4, if it is not in the water discharge alarm state, then go to step S2-5, if it is in the water discharge alarm state, then go to step S2-6,

[0014] S2-5, determine whether the water should be released and the alarm should be sounded: S2-51, after subtracting the maximum and minimum values ​​from the 10 groups of parameters in the collected data group, sum up the remaining parameters and then calculate the average value to obtain the average value P; S2-52 compares the parameters in the collected data group with P*X%. If the judgment result is less than 3 times in a row, it is judged as a water release alarm state, triggering the alarm to be sent, recording the sending time, and ending the alarm algorithm judgment process. Otherwise, it is judged that the water should not be released and the alarm should be marked as not alarming, and the alarm algorithm judgment process ends;

[0015] S2-6, determine whether the pressure value of the digital pressure gauge at the pressure reducing valve exceeds the normal water pressure state before the water discharge alarm is triggered for 30S for three consecutive times. If it times out, it is determined to be an abnormal pipeline state, the parameters are cleared, and the water discharge alarm state is canceled after feedback to the platform. If it does not time out, determine whether the data in the collected data group is greater than P*X%. If it is greater and greater than for three consecutive times, it is determined that the water discharge alarm is lifted, the water discharge alarm is canceled after feedback to the platform, marked as no alarm, and the alarm algorithm judgment process ends;

[0016] S2-7, determine whether it is currently in a state of no high or low alarm, if yes, determine whether it is greater than the high alarm threshold and proceed to step S2-8, if not, determine whether it is currently in a state of waiting for confirmation of a high alarm and proceed to step S2-9;

[0017] S2-8, if it is greater than the high alarm threshold, then record the collected data and time, enter the waiting confirmation high alarm state, mark it as no alarm, and end the alarm algorithm judgment process; if it is less than or equal to the high alarm threshold, then determine whether it is less than the low alarm threshold. If it is less than the low alarm threshold, then record the collected data and time, enter the waiting confirmation low alarm state, if it is greater than or equal to the low alarm threshold, then mark it as no alarm, and end the alarm algorithm judgment process;

[0018] S2-9, if it is currently in the high alarm state waiting for confirmation, determine whether the state exceeds 10 seconds and is greater than the high alarm threshold. If so, enter the high alarm state, start the alarm transmission, and end the alarm algorithm judgment process. If not, mark it as no alarm and end the alarm algorithm judgment process; if it is not currently in the high alarm state waiting for confirmation, determine whether it is currently in the high alarm state and enter step S2-10;

[0019] S2-10, if it is in the high alarm state, determine whether it is less than the high alarm return difference value, if so, restore to normal state, start sending, and mark it as no alarm, and end the alarm algorithm judgment process. If not, it is marked as no alarm this time, and the alarm algorithm judgment process is ended. If it is not in the high alarm state, determine whether it is currently in the low alarm state waiting for confirmation, and enter step S2-11;

[0020] S2-11, if it is currently in the low alarm state waiting for confirmation, determine whether the state exceeds 10 seconds and is less than the low alarm threshold. If so, enter the low alarm state, start the alarm transmission, and end the alarm algorithm judgment process. If not, mark it as no alarm, and end the alarm algorithm judgment process. If it is not currently in the low alarm state waiting for confirmation, determine whether it is currently in the low alarm state, and enter step S2-12;

[0021] S2-12, if it is currently in a low alarm state, determine whether it is greater than the low alarm return difference value. If so, restore to normal state, start sending, and end the alarm algorithm judgment process. If not, mark this time as no alarm, and end the alarm algorithm judgment process. If it is not in a low alarm state, enter step S2-13 to determine whether the fluctuation alarm state should be started;

[0022] S2-13, compare the current data with the mean XP. If the difference between the current data and the mean XP is greater than the fluctuation difference FX, the fluctuation alarm is initiated and the alarm algorithm judgment process ends. Otherwise, it is marked as no alarm and the alarm algorithm judgment process ends.

[0023] The X% is calculated by the water release self-learning algorithm, and the specific method is as follows: S2-61, turn on the digital pressure gauge to collect data, collect and record data every 5 minutes after initialization, and calculate the average value PC; S2-62, open the valve 50%, wait for 30 seconds, and calculate the average value P1 of 30 seconds; S2-63, open the valve 100%, wait for 30 seconds, and calculate the average value P2 of 30 seconds; S2-64, close the valve, wait for 60 seconds, and calculate the average value P3 of 60 seconds; S2-65, calculate X%=(P1+P2) / ((P3+PC) / 2).

[0024] The mean value X P The fluctuation difference FX is obtained by the fluctuation self-learning algorithm. The specific method is as follows: S2-131: Turn on the digital pressure gauge to collect data, record the collected data X1 and the collection time; S2-132 determines whether the collection time is greater than 24 hours. If it is greater, proceed to step S2-133. If the collection time is less than or equal to 24 hours, return to step S2-131 to continue collecting; S2-133, calculate the average value X1 of N collected data P; S2-134, calculate the fluctuation difference FX= .

[0025] The calculation method of the high alarm threshold and the low alarm threshold is as follows: High alarm threshold Y 高 =X P +FX / 2, low alarm threshold Y 低 =X P -FX / 2.

[0026] The calculation method of the high alarm hysteresis value and the low alarm hysteresis value is as follows: the high alarm hysteresis value and the low alarm hysteresis value are the same, both of which are FX / 2.

[0027] The method of the alarm reporting process is as follows: S3-1, the data group is sent to the wireless module through the serial port; S3-2, the wireless module transmits data to the data server through the operator base station; S3-3, the data server processes the data and saves it to the database; S3-4, the management platform statistics and displays the current alarm data, the edge server calls the database data to analyze the data, and the mobile phone applet directly calls the database to obtain the alarm data; S3-5, pushes the alarm data to the third-party platform, and sends alarm SMS and emails to customers.

[0028] The specific method in step S1 is as follows: S1-1, turn on the power supply of the digital pressure gauge, after the power supply is stable for 150ms, the digital pressure gauge is initialized, and it is determined whether the initialization is successful; S1-2, if the initialization is successful, sample the data once, if the initialization fails, return to the initialization abnormality parameter and enter step 1-5; S1-3, compare the sampled data of step S1-2 with the maximum and minimum values ​​and save them; S1-4, determine whether to collect n+2 times of data, if so, subtract the maximum and minimum values ​​and calculate the average value of the n times of collected data; if not, return to step S1-2 to continue collecting data; S1-5, turn off the power and initialize the sensor, S1-6, determine whether the sensor returns abnormally, if so, enter the alarm reporting process and end the collection task, if not, enter the alarm algorithm judgment process of step S2.

[0029] In the step S1-2, if the data collection fails, the data is collected again. If the number of failures is greater than 3 times, the process returns to collecting abnormal parameters and proceeds to step S1-5.

[0030] Compared with the prior art, the present invention combines data from various points in the building to perform self-learning, monitoring and alarming, reduces interference from abnormal data, and obtains accurate alarm information in a timely manner. BRIEF DESCRIPTION OF THE DRAWINGS

[0031] Figure 1 This is a layout diagram of the digital pressure gauges at various points in the building of the present invention.

[0032] Figure 2 This is a flow chart of step S2 of the present invention.

[0033] Figure 3 This is a flow chart of the water release self-learning of the present invention.

[0034] Figure 4 This is a flow chart of the fluctuation self-learning of the present invention.

[0035] Figure 5 This is a flow chart of the alarm push of the present invention.

[0036] Figure 6 This is a flow chart of step S1 of the present invention. DETAILED DESCRIPTION

[0037] The present invention will be further described below with reference to the accompanying drawings.

[0038] Embodiment 1:

[0039] This embodiment is a method for monitoring and alarming data at various points in a building with a self-learning function. Figure 1 As shown, a digital pressure gauge 1 is provided on each floor of the building, pressure reducing valves 4 are provided between adjacent floors, a water tank 2 is provided on the top floor, and a water pump 3 is provided on the bottom floor.

[0040] like Figure 2 As shown, the method comprises the following steps:

[0041] S1, initialization of digital pressure gauge, monitoring water pressure of each floor in the building, sampling digital pressure gauge data;

[0042] S2, enter the alarm algorithm judgment process,

[0043] S3, according to the judgment of step S2, whether an alarm is triggered is determined. If an alarm is triggered, the alarm reporting process is started, and the current pressure data is saved and displayed. If an alarm is not triggered, the current pressure data is saved and displayed, and steps S1 to S3 are repeated;

[0044] S4, after the specified time, the scheduled collection task is terminated;

[0045] The specific method of the alarm algorithm process in step S2 is as follows:

[0046] S2-1, the digital pressure gauge obtains the current collected data and caches it into the collected data group;

[0047] S2-2, determine whether there are 10 groups of parameters in the collected data group, and proceed to step S2-3, and at the same time determine whether it is in a state of no high or low alarm, and proceed to step S2-7;

[0048] S2-3, if there are 10 sets of parameters in the collected data group, determine whether it is currently in the water release alarm state and enter step S2-4. If there are not 10 sets of parameters, mark it as no alarm and end the alarm algorithm judgment process;

[0049] S2-4, if it is not in the water discharge alarm state, then go to step S2-5, if it is in the water discharge alarm state, then go to step S2-6,

[0050] S2-5, determine whether the water should be released and the alarm should be sounded: S2-51, after subtracting the maximum and minimum values ​​from the 10 groups of parameters in the collected data group, sum up the remaining parameters and then calculate the average value to obtain the average value P; S2-52 compares the parameters in the collected data group with P*X%. If the judgment result is less than 3 times in a row, it is judged as a water release alarm state, triggering the alarm to be sent, recording the sending time, and ending the alarm algorithm judgment process. Otherwise, it is judged that the water should not be released and the alarm should be marked as not alarming, and the alarm algorithm judgment process ends;

[0051] S2-6, determine whether the pressure value of the digital pressure gauge at the pressure reducing valve exceeds the normal water pressure state before the water discharge alarm is triggered for 30S for three consecutive times. If it times out, it is determined to be an abnormal pipeline state, the parameters are cleared, and the water discharge alarm state is canceled after feedback to the platform. If it does not time out, determine whether the data in the collected data group is greater than P*X%. If it is greater and greater than for three consecutive times, it is determined that the water discharge alarm is lifted, the water discharge alarm is canceled after feedback to the platform, marked as no alarm, and the alarm algorithm judgment process ends;

[0052] S2-7, determine whether it is currently in a state of no high or low alarm, if yes, determine whether it is greater than the high alarm threshold and proceed to step S2-8, if not, determine whether it is currently in a state of waiting for confirmation of a high alarm and proceed to step S2-9;

[0053] S2-8, if it is greater than the high alarm threshold, then record the collected data and time, enter the waiting confirmation high alarm state, mark it as no alarm, and end the alarm algorithm judgment process; if it is less than or equal to the high alarm threshold, then determine whether it is less than the low alarm threshold. If it is less than the low alarm threshold, then record the collected data and time, enter the waiting confirmation low alarm state, if it is greater than or equal to the low alarm threshold, then mark it as no alarm, and end the alarm algorithm judgment process;

[0054] S2-9, if it is currently in the high alarm state waiting for confirmation, determine whether the state exceeds 10 seconds and is greater than the high alarm threshold. If so, enter the high alarm state, start the alarm transmission, and end the alarm algorithm judgment process. If not, mark it as no alarm and end the alarm algorithm judgment process; if it is not currently in the high alarm state waiting for confirmation, determine whether it is currently in the high alarm state and enter step S2-10;

[0055] S2-10, if it is in the high alarm state, determine whether it is less than the high alarm return difference value, if so, restore to normal state, start sending, and mark it as no alarm, and end the alarm algorithm judgment process. If not, it is marked as no alarm this time, and the alarm algorithm judgment process is ended. If it is not in the high alarm state, determine whether it is currently in the low alarm state waiting for confirmation, and enter step S2-11;

[0056] S2-11, if it is currently in the low alarm state waiting for confirmation, determine whether the state exceeds 10 seconds and is less than the low alarm threshold. If so, enter the low alarm state, start the alarm transmission, and end the alarm algorithm judgment process. If not, mark it as no alarm, and end the alarm algorithm judgment process. If it is not currently in the low alarm state waiting for confirmation, determine whether it is currently in the low alarm state, and enter step S2-12;

[0057] S2-12, if it is currently in a low alarm state, determine whether it is greater than the low alarm return difference value. If so, restore to normal state, start sending, and end the alarm algorithm judgment process. If not, mark this time as no alarm, and end the alarm algorithm judgment process. If it is not in a low alarm state, enter step S2-13 to determine whether the fluctuation alarm state should be started;

[0058] S2-13, compare the current data with the mean X P Compare, if the current data is with the mean X P If the difference is greater than the fluctuation difference FX, the fluctuation alarm is initiated and the alarm algorithm judgment process ends. Otherwise, it is marked as no alarm and the alarm algorithm judgment process ends.

[0059] The X% in step S2-6 is calculated by the water release self-learning algorithm, and the specific method is as follows: S2-61, turn on the digital pressure gauge to collect data, collect and record data every 5 minutes after initialization, and calculate the average value PC; S2-62, open the valve 50%, wait for 30 seconds, and calculate the average value P1 of 30 seconds; S2-63, open the valve 100%, wait for 30 seconds, and calculate the average value P2 of 30 seconds; S2-64, close the valve, wait for 60 seconds, and calculate the average value P3 of 60 seconds; S2-65, calculate X%=(P1+P2) / ((P3+PC) / 2).

[0060] In step S2-13, the mean value X PThe fluctuation difference FX is obtained by the fluctuation self-learning algorithm. The specific method is as follows: S2-131: Turn on the digital pressure gauge to collect data, record the collected data X1 and the collection time; S2-132 determines whether the collection time is greater than 24 hours. If it is greater, proceed to step S2-133. If the collection time is less than or equal to 24 hours, return to step S2-131 to continue collecting; S2-133, calculate the average value X1 of N collected data P ; S2-134, calculate the fluctuation difference FX= .

[0061] In steps S2-7 to S2-11, the high alarm threshold and the low alarm threshold are calculated as follows: High alarm threshold Y 高 =X P +FX / 2, low alarm threshold Y 低 =X P -FX / 2.

[0062] In step S10 and step S12, the high alarm hysteresis value and the low alarm hysteresis value are calculated as follows: the high alarm hysteresis value and the low alarm hysteresis value are the same, both of which are FX / 2.

[0063] The method of the alarm reporting process in step S3 is as follows: S3-1, the data group is sent to the wireless module through the serial port; S3-2, the wireless module transmits data to the data server through the operator base station; S3-3, the data server processes the data and saves it to the database; S3-4, the management platform statistics and displays the current alarm data, the edge server calls the database data to analyze the data, and the mobile phone applet directly calls the database to obtain the alarm data; S3-5, push the alarm data to the third-party platform, and send alarm SMS and emails to customers.

[0064] The specific method in step S1 is as follows: S1-1, turn on the power supply of the digital pressure gauge, after the power supply is stable for 150ms, the digital pressure gauge is initialized, and it is determined whether the initialization is successful; S1-2, if the initialization is successful, sample the data once, if the initialization fails, return to the initialization abnormality parameter and enter step 1-5; S1-3, compare the sampled data of step S1-2 with the maximum and minimum values ​​and save them; S1-4, determine whether to collect n+2 times of data, if so, subtract the maximum and minimum values ​​and calculate the average value of the n collected data; if not, return to step S1-2 to continue collecting data; S1-5, turn off the power and initialize the sensor, S1-6, determine whether the sensor returns abnormally, if so, enter the alarm reporting process and end the collection task, if not, enter the alarm algorithm judgment process of step S2.

[0065] In step S12, if data collection fails, data is collected again. If the number of failures is greater than 3 times, the process returns to collecting abnormal parameters and proceeds to step S15.

[0066] In step S2-6, the judgment timeout of 30 seconds is to reduce the possibility of misjudgment caused by pipeline water pressure fluctuation.

[0067] The present invention combines data from various points in a building to perform self-learning, monitoring and alarming, thereby reducing interference from abnormal data and obtaining accurate alarm information in a timely manner.

Claims

1. A method for monitoring and alarming data at various points in a building with a self-learning function, characterized in that: A digital pressure gauge (1) is provided on each floor of the building, pressure reducing valves (4) are provided between adjacent floors, a water tank (2) is provided on the top floor, and a water pump (3) is provided on the bottom floor. The method comprises the following steps: S1, initialization of digital pressure gauge, monitoring water pressure of each floor in the building, sampling digital pressure gauge data; S2, enter the alarm algorithm judgment process, S3, according to the judgment of step S2, whether an alarm is triggered is determined. If an alarm is triggered, the alarm reporting process is started, and the current pressure data is saved and displayed. If an alarm is not triggered, the current pressure data is saved and displayed, and steps S1 to S3 are repeated; S4, after the specified time, the scheduled collection task is terminated; The specific method of the alarm algorithm process in step S2 is as follows: S2-1, the digital pressure gauge obtains the current collected data and caches it into the collected data group; S2-2, determine whether there are 10 groups of parameters in the collected data group, and proceed to step S2-3, and at the same time determine whether it is in a state of no high or low alarm, and proceed to step S2-7; S2-3, if there are 10 sets of parameters in the collected data group, determine whether it is currently in the water release alarm state and enter step S2-4. If there are not 10 sets of parameters, mark it as no alarm and end the alarm algorithm judgment process; S2-4, if it is not in the water discharge alarm state, then go to step S2-5, if it is in the water discharge alarm state, then go to step S2-6, S2-5, determine whether the water should be released and the alarm should be sounded: S2-51, after subtracting the maximum and minimum values ​​from the 10 groups of parameters in the collected data group, sum up the remaining parameters and then calculate the average value to obtain the average value P; S2-52 compares the parameters in the collected data group with P*X%. If the judgment result is less than 3 times in a row, it is judged as a water release alarm state, triggering the alarm to be sent, recording the sending time, and ending the alarm algorithm judgment process. Otherwise, it is judged that the water should not be released and the alarm should be marked as not alarming, and the alarm algorithm judgment process ends; S2-6, determine whether the pressure value of the digital pressure gauge at the pressure reducing valve exceeds the normal water pressure state before the water discharge alarm is triggered for 30S for three consecutive times. If it times out, it is determined to be an abnormal pipeline state, the parameters are cleared, and the water discharge alarm state is canceled after feedback to the platform. If it does not time out, determine whether the data in the collected data group is greater than P*X%. If it is greater and greater than for three consecutive times, it is determined that the water discharge alarm is lifted, the water discharge alarm is canceled after feedback to the platform, marked as no alarm, and the alarm algorithm judgment process ends; S2-7, determine whether it is currently in a state of no high or low alarm, if yes, determine whether it is greater than the high alarm threshold and proceed to step S2-8, if not, determine whether it is currently in a state of waiting for confirmation of a high alarm and proceed to step S2-9; S2-8, if it is greater than the high alarm threshold, then record the collected data and time, enter the waiting confirmation high alarm state, mark it as no alarm, and end the alarm algorithm judgment process; if it is less than or equal to the high alarm threshold, then determine whether it is less than the low alarm threshold. If it is less than the low alarm threshold, then record the collected data and time, enter the waiting confirmation low alarm state, if it is greater than or equal to the low alarm threshold, then mark it as no alarm, and end the alarm algorithm judgment process; S2-9, if it is currently in the high alarm state waiting for confirmation, determine whether the state exceeds 10 seconds and is greater than the high alarm threshold. If so, enter the high alarm state, start the alarm transmission, and end the alarm algorithm judgment process. If not, mark it as no alarm and end the alarm algorithm judgment process; if it is not currently in the high alarm state waiting for confirmation, determine whether it is currently in the high alarm state and enter step S2-10; S2-10, if it is in the high alarm state, determine whether it is less than the high alarm return difference value, if so, restore to normal state, start sending, and mark it as no alarm, and end the alarm algorithm judgment process. If not, it is marked as no alarm this time, and the alarm algorithm judgment process is ended. If it is not in the high alarm state, determine whether it is currently in the low alarm state waiting for confirmation, and enter step S2-11; S2-11, if it is currently in the low alarm state waiting for confirmation, determine whether the state exceeds 10 seconds and is less than the low alarm threshold. If so, enter the low alarm state, start the alarm transmission, and end the alarm algorithm judgment process. If not, mark it as no alarm, and end the alarm algorithm judgment process. If it is not currently in the low alarm state waiting for confirmation, determine whether it is currently in the low alarm state, and enter step S2-12; S2-12, if it is currently in a low alarm state, determine whether it is greater than the low alarm return difference value. If so, restore to normal state, start sending, and end the alarm algorithm judgment process. If not, mark this time as no alarm, and end the alarm algorithm judgment process. If it is not in a low alarm state, enter step S2-13 to determine whether the fluctuation alarm state should be started; S2-13, compare the current data with the mean X P Compare, if the current data is with the mean X P If the difference is greater than the fluctuation difference FX, the fluctuation alarm is initiated and the alarm algorithm judgment process ends. Otherwise, it is marked as no alarm and the alarm algorithm judgment process ends.

2. According to the method for monitoring and alarming data at each point in a building with self-learning function as described in claim 1, it is characterized in that: The X% is obtained by the water release self-learning algorithm, and the specific method is as follows: S2-61, turn on the digital pressure gauge to collect data, collect and record data for 5 minutes after initialization, and calculate the average value PC; S2-62, open the valve 50%, wait for 30 seconds, and calculate the average value P1 of 30 seconds; S2-63, open the valve 100%, wait for 30 seconds, and calculate the average value P2 of 30 seconds; S2-64, close the valve, wait for 60 seconds, and calculate the average value P3 of 60 seconds; S2-65, calculate X%=(P1+P2) / ((P3+PC) / 2).

3. The method for monitoring and alarming data at each point in a building with a self-learning function according to claim 1 is characterized in that: The mean value X P The fluctuation difference FX is obtained by the fluctuation self-learning algorithm. The specific method is as follows: S2-131: Turn on the digital pressure gauge to collect data, record the collected data X1 and the collection time; S2-132 determines whether the collection time is greater than 24 hours. If it is greater, proceed to step S2-133. If the collection time is less than or equal to 24 hours, return to step S2-131 to continue collecting; S2-133, calculate the average value X1 of N collected data P ; S2-134, calculate the fluctuation difference FX= .

4. The method for monitoring and alarming data at each point in a building with a self-learning function according to claim 3 is characterized in that: The calculation method of the high alarm threshold and the low alarm threshold is as follows: High alarm threshold Y 高 =X P +FX / 2, low alarm threshold Y 低 =X P -FX / 2.

5. The method for monitoring and alarming data at each point in a building with a self-learning function according to claim 3 is characterized in that: The calculation method of the high alarm hysteresis value and the low alarm hysteresis value is as follows: the high alarm hysteresis value and the low alarm hysteresis value are the same, both of which are FX / 2.

6. The method for monitoring and alarming data at each point in a building with a self-learning function according to claim 1 is characterized in that: The method of the alarm reporting process is as follows: S3-1, the data group is sent to the wireless module through the serial port; S3-2, the wireless module transmits data to the data server through the operator base station; S3-3, the data server processes the data and saves it to the database; S3-4, the management platform statistics and displays the current alarm data, the edge server calls the database data to analyze the data, and the mobile phone applet directly calls the database to obtain the alarm data; S3-5, pushes the alarm data to the third-party platform, and sends alarm SMS and emails to customers.

7. The method for monitoring and alarming data at each point in a building with a self-learning function according to claim 1 is characterized in that: The specific method in step S1 is as follows: S1-1, turn on the power supply of the digital pressure gauge, after the power supply is stable for 150ms, the digital pressure gauge is initialized, and it is determined whether the initialization is successful; S1-2, if the initialization is successful, sample the data once, if the initialization fails, return to the initialization abnormality parameter and enter step 1-5; S1-3, compare the sampled data of step S1-2 with the maximum and minimum values ​​and save them; S1-4, determine whether to collect n+2 times of data, if so, subtract the maximum and minimum values ​​and calculate the average value of the n times of collected data; if not, return to step S1-2 to continue collecting data; S1-5, turn off the power and initialize the sensor, S1-6, determine whether the sensor returns abnormally, if so, enter the alarm reporting process and end the collection task, if not, enter the alarm algorithm judgment process of step S2.

8. The method for monitoring and alarming data at each point in a building with a self-learning function according to claim 7 is characterized in that: In the step S1-2, if the data collection fails, the data is collected again. If the number of failures is greater than 3 times, the process returns to collecting abnormal parameters and proceeds to step S1-5.