Community pipe network intelligent monitoring and early warning system and method based on internet of things

Abstract

The application discloses a community pipe network intelligent monitoring and early warning system and method based on an Internet of Things, and relates to the technical field of pipe network monitoring.The distribution model of a community pipe network is constructed, and the community pipe network is divided into a plurality of unit blocks; an existing monitoring unit and an existing monitoring area are acquired; a supplementary monitoring unit is set, and a supplementary monitoring area is acquired; the monitoring level of different monitoring areas is acquired according to historical accident records of different accident types; the monitoring participation of different monitoring data is acquired according to historical monitoring data of the historical accident records of different accident types; and an accident evaluation model is constructed in combination with real-time monitoring data, the monitoring level and the monitoring participation, so as to judge whether a safety accident exists and the accident type; the whole community pipe network can be brought into an effective monitoring range, and an accurate monitoring mechanism can be formed.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of pipe network monitoring, and particularly relates to a community pipe network intelligent monitoring and early warning system and method based on Internet of Things. BACKGROUND

[0002] With the acceleration of urbanization, the safety problem in the community is increasingly concerned, especially the safety of the pipe network in the community. The traditional pipe network monitoring has problems such as poor real-time performance, slow response speed, high false alarm rate and the like, and often cannot achieve rapid response, and is difficult to meet the safety needs of the modern community, and cannot realize intelligent and dynamic monitoring of the pipe network in the community.

[0003] In the prior art, the safety monitoring of the pipe network in the community is often limited to real-time data, and the effective use of historical data is lacking. Different safety accidents often cause abnormal changes in different parameters. By using this feature, the safety accident can be effectively judged, and the type of the accident can be obtained. The prior art lacks such technical means. Therefore, the present application provides a community pipe network intelligent monitoring and early warning system and method based on Internet of Things. SUMMARY

[0004] The present application provides a community pipe network intelligent monitoring and early warning system and method based on Internet of Things.

[0005] The purpose of the present application can be achieved by the following technical scheme: a community pipe network intelligent monitoring and early warning system based on Internet of Things, comprising the following modules:

[0006] A map construction module is configured to obtain distribution information of the pipe network in the community and construct a distribution model thereof, and divide the pipe network in the community into a plurality of unit blocks in the distribution model.

[0007] A region division module is configured to obtain existing monitoring units and existing monitoring regions, set supplementary monitoring units and obtain supplementary monitoring regions, obtain historical accident records of different accident types, and obtain monitoring levels of different monitoring regions according to the historical accident records.

[0008] A monitoring evaluation module is configured to obtain historical monitoring data of the historical accident records of different accident types, obtain monitoring participation degrees of different monitoring data according to the historical monitoring data, and construct an accident evaluation model according to the historical monitoring data, the monitoring participation degrees and the monitoring levels corresponding to the historical accident records of different accident types.

[0009] An accident determination module is configured to obtain current real-time monitoring data, input the monitoring levels and the monitoring participation degrees into different accident evaluation models to determine whether a safety accident exists and the type of the accident.

[0010] Further, the distribution information of the pipe network in the community is acquired, and a distribution model is constructed, and the process of dividing the pipe network in the community into a plurality of unit blocks in the distribution model comprises:

[0011] The distribution information refers to data related to the distribution of the pipe network for water supply, gas supply and power supply in the community, including pipe purpose, pipe location, pipe direction, pipe burial depth, pipe diameter and material, connection point and interface, valve and control equipment;

[0012] The corresponding distribution model is constructed according to the acquired distribution information by using digital twin technology, the distribution models of pipe networks with different pipe purposes are independent of each other, and the distribution model is divided into a plurality of unit blocks of equal size.

[0013] Further, the process of acquiring the existing monitoring unit and the existing monitoring area, setting the supplementary monitoring unit and acquiring the supplementary monitoring area comprises:

[0014] The pipe networks with different pipe purposes are respectively marked as water supply pipe network, gas supply pipe network and power supply pipe network, and different parameters of different pipe networks are respectively monitored by using the monitoring unit, including water pressure monitoring unit, flow monitoring unit and flow rate monitoring unit;

[0015] The existing monitoring unit and the supplementary monitoring unit are respectively marked as the existing monitoring unit and the supplementary monitoring unit, and the positions of the existing monitoring unit and the supplementary monitoring unit are uploaded to the distribution model of the corresponding pipe network for synchronization;

[0016] The existing monitoring area of each existing monitoring unit is acquired according to the effective monitoring range of each existing monitoring unit in the distribution model, and the supplementary monitoring area of each supplementary monitoring unit is acquired according to the effective monitoring range of each supplementary monitoring unit, and the effective monitoring range refers to the maximum coverage range of the monitoring unit that can monitor the corresponding parameters.

[0017] Further, the process of acquiring the monitoring level of different monitoring areas according to the historical accident records comprises:

[0018] The historical accident records refer to data related to the safety accidents that have occurred in the pipe network in the community, including accident type, accident time, accident location and accident influence range, and each historical accident record is uploaded to the distribution model of the corresponding pipe network for synchronization according to the accident location;

[0019] The number of historical accident records that have occurred in each monitoring area is acquired, the monitoring area includes the existing monitoring area and the supplementary monitoring area, an accident number standard is set, and the number of historical accident records that have occurred in each monitoring area is compared with the accident number standard to acquire the monitoring level of the corresponding monitoring area, including low monitoring level, high monitoring level and medium monitoring level.

[0020] Further, the historical monitoring data of different accident types is obtained, and the process of obtaining the monitoring participation of different monitoring data according to the historical monitoring data includes:

[0021] Each monitoring unit in the monitoring area where a single historical accident record is located is taken as the historical monitoring data of the historical accident record, including water pressure, flow, flow rate, and each parameter obtained within a preset time length before and after the accident occurrence time of the historical accident record;

[0022] The water pressure, flow, and flow rate in the historical monitoring data corresponding to the single historical accident record are sequentially numbered in time order as C ai , C bi , C ci , i = 1, 2, …, n, n is the number of parameters, and the water pressure participation B a , flow participation B b , and flow rate participation B c at the accident occurrence time are obtained respectively;

[0023] ;

[0024] ;

[0025] ;

[0026] C ax , C bx , C cx are the water pressure, flow, and flow rate at the accident occurrence time respectively, the monitoring participation corresponding to each historical accident record of the same accident type is obtained, including the water pressure participation, the flow participation, and the flow rate participation, and the monitoring participation corresponding to each historical accident record of different accident types is obtained.

[0027] Further, the process of constructing an accident assessment model according to the historical monitoring data, the monitoring participation, and the monitoring level corresponding to the historical accident records of different accident types includes:

[0028] For different types of safety accidents, corresponding accident assessment models are constructed respectively, an accident assessment set is generated according to the historical monitoring data, the monitoring participation, and the monitoring level corresponding to each historical accident record of the same accident type, and the accident assessment set is divided into a training set and a test set;

[0029] A convolutional neural network is constructed, different historical monitoring data, monitoring participation, and monitoring level in the training set are taken as the input data of the convolutional neural network, whether a safety accident will occur is taken as the output data of the convolutional neural network, the convolutional neural network is trained to obtain an initial convolutional neural network;

[0030] The initial convolutional neural network is verified by using the test set, and an initial convolutional neural network with an output less than or equal to a preset test error threshold is output as an accident assessment model corresponding to the safety accident of the accident type, and the accident assessment models corresponding to the safety accidents of different accident types are constructed.

[0031] Further, the current real-time monitoring data is acquired, and the monitoring levels and the monitoring participation degrees are input into different accident assessment models to determine whether a safety accident and the accident type thereof exist, and the process includes:

[0032] The parameters acquired by the monitoring units in each monitoring area within a preset time length are taken as the real-time monitoring data of the monitoring units, and the average values of the monitoring participation degrees corresponding to the historical accident records of the same accident type are taken as the assessment participation degrees of the accident type;

[0033] The real-time monitoring data of a single monitoring area and the monitoring level thereof are input into the accident assessment models of the safety accidents of different accident types to determine whether a safety accident exists in the monitoring area;

[0034] If a safety accident exists, the accident type corresponding to the accident assessment model determining that a safety accident exists is taken as the accident type of the safety accident of the monitoring area, and if a safety accident does not exist, no other operation is performed.

[0035] A community pipe network intelligent monitoring and early warning method based on the Internet of Things includes the following steps:

[0036] Step S1: The distribution information of the community pipe network is acquired, and a distribution model thereof is constructed, and the community pipe network is divided into a plurality of unit blocks in the distribution model;

[0037] Step S2: The existing monitoring units and the existing monitoring areas are acquired, the supplementary monitoring units are set, the supplementary monitoring areas are acquired, the historical accident records of different accident types are acquired, and the monitoring levels of different monitoring areas are acquired according to the historical accident records;

[0038] Step S3: The historical monitoring data of the historical accident records of different accident types are acquired, the monitoring participation degrees of different monitoring data are acquired according to the historical monitoring data, and the accident assessment models are constructed according to the historical monitoring data, the monitoring participation degrees, and the monitoring levels corresponding to the historical accident records of different accident types;

[0039] Step S4: The current real-time monitoring data is acquired, and the monitoring levels and the monitoring participation degrees are input into different accident assessment models to determine whether a safety accident and the accident type thereof exist.

[0040] Compared with the prior art, the present application has the following advantages:

[0041] The present application can guarantee that the entire pipe network is included in the effective monitoring range by dividing the pipe network in the community into different areas and respectively presetting and adding corresponding monitoring units to monitor them. By utilizing historical accident records, the monitoring level is obtained according to the number of safety accidents in each area, which can reflect the frequent occurrence of accidents in different areas.

[0042] Secondly, the correlation degree of different parameters and safety accidents can be reflected according to the change of each parameter under different safety accidents, the participation degree of each parameter when a safety accident of different types occurs can be obtained, and an accident evaluation model is constructed based on this, which can combine real-time monitoring data and different participation degrees to judge whether a safety accident and its type will occur, which is beneficial to form an accurate and effective monitoring mechanism for the pipe network in the community, and timely discover accidents and their types and feedback. BRIEF DESCRIPTION OF DRAWINGS

[0043] Figure 1 The present application is a schematic diagram of the module. DETAILED DESCRIPTION

[0044] As shown in Figure 1 A community pipe network intelligent monitoring and early warning system based on the Internet of Things includes the following modules:

[0045] A map construction module is used to obtain the distribution information of the pipe network in the community and construct a distribution model, in which the pipe network in the community is divided into a plurality of unit blocks.

[0046] A region division module is used to obtain existing monitoring units and existing monitoring regions, set up supplementary monitoring units and obtain supplementary monitoring regions, obtain historical accident records of different accident types, and obtain the monitoring level of different monitoring regions according to the historical accident records.

[0047] A monitoring evaluation module is used to obtain historical monitoring data of the historical accident records of different accident types, obtain the monitoring participation degree of different monitoring data according to the historical monitoring data, and construct an accident evaluation model according to the historical monitoring data, monitoring participation degree and monitoring level corresponding to the historical accident records of different accident types.

[0048] An accident determination module is used to obtain current real-time monitoring data, combine the monitoring level and the monitoring participation degree, and input them into different accident evaluation models to determine whether a safety accident and its type exist.

[0049] It should be further noted that in the specific implementation process, the process of obtaining the distribution information of the pipe network in the community and constructing a distribution model, in which the pipe network in the community is divided into a plurality of unit blocks, includes:

[0050] The distribution information refers to data related to the distribution of the pipe network in the community, the pipe network refers to a network system composed of various pipelines for water supply, gas supply, and power supply in the community, and the distribution information includes pipeline purpose, pipeline location, pipeline direction, pipeline burial depth, pipeline diameter and material, connection point and interface, valve, and control equipment;

[0051] A digital twin model of the pipe network in the community is constructed according to the obtained distribution information by using a digital twin technology, and is marked as a distribution model, the distribution models of pipe networks with different pipeline purposes are independent of each other, the distribution model is a three-dimensional model, the distribution model is divided into a plurality of unit blocks of equal size, the unit block is a cube, and all unit blocks can contain the entire distribution model.

[0052] It should be further explained that, in the specific implementation process, the existing monitoring units and the existing monitoring areas are obtained, the supplementary monitoring units are set, and the supplementary monitoring areas are obtained, and the process includes:

[0053] The pipeline purposes include water supply, gas supply, and power supply, and are respectively marked as water supply pipe network, gas supply pipe network, and power supply pipe network, for different pipe networks, the parameters to be monitored are also different, for the water supply pipe network, the water pressure, flow, and flow rate need to be monitored, for the gas supply pipe network, the gas pressure, flow, and concentration need to be monitored, and for the power supply pipe network, the voltage, current, and temperature need to be monitored.

[0054] Taking the water supply pipe network and its distribution model as an example, different monitoring units are needed for monitoring different parameters, including water pressure monitoring units, flow monitoring units, and flow rate monitoring units, and the monitoring units of the current water supply pipe network are marked as existing monitoring units.

[0055] The positions of the existing monitoring units are uploaded to the distribution model of the water supply pipe network for synchronization, and the existing monitoring areas of the existing monitoring units are obtained in the distribution model according to the effective monitoring ranges of the existing monitoring units, the effective monitoring range refers to the maximum coverage range of the monitoring unit that can monitor the corresponding parameter, and the existing monitoring area refers to an area composed of all unit blocks within the effective monitoring range of a single existing monitoring unit.

[0056] In the distribution model, other unit blocks outside the existing monitoring area are marked as a monitoring area, and monitoring units corresponding to different parameters are added to the pipe network in the monitoring area, and the final addition result can make each unit block simultaneously within the effective monitoring range of the monitoring units corresponding to different parameters.

[0057] The added monitoring units are marked as supplementary monitoring units, and the supplementary monitoring areas of the supplementary monitoring units are obtained according to the effective monitoring ranges of the supplementary monitoring units. The supplementary monitoring area refers to an area formed by all unit blocks in the effective monitoring range of a single supplementary monitoring unit. If there is an overlapping area between the existing monitoring area and the supplementary monitoring area, the overlapping area is divided into the existing monitoring area.

[0058] It should be further explained that, in the specific implementation process, the historical accident records of different accident types are obtained, and the monitoring levels of different monitoring areas are obtained according to the historical accident records, which includes:

[0059] The historical accident records refer to data related to safety accidents that have occurred in the community, including accident type, accident occurrence time, accident occurrence location, and accident impact range. The accident type includes pipe rupture, water pressure anomaly, pipe blockage, valve failure, and pump station failure.

[0060] The historical accident records are uploaded to the distribution model of the water supply network according to the accident occurrence location for synchronization. The number of historical accident records that have occurred in a single monitoring area (including the existing monitoring area and the supplementary monitoring area) is obtained in the distribution model and is denoted as S.

[0061] An accident number standard [S min , S max ] is set, and the number of historical accident records that have occurred in a single monitoring area is compared with the accident number standard. If S≤S min , it is marked as a low monitoring level. If S≥S max , it is marked as a high monitoring level. If S min <S<S max , it is marked as a medium monitoring level. The same method is used to obtain the monitoring level of the water supply network of each monitoring area.

[0062] It should be further explained that, in the specific implementation process, the historical monitoring data of the historical accident records of different accident types is obtained, and the monitoring participation of different monitoring data is obtained according to the historical monitoring data, which includes:

[0063] Taking a single historical accident record as an example, the historical monitoring data of each monitoring unit in the monitoring area of the historical accident record is obtained within a preset time period before and after the accident occurrence time of the historical accident record, including water pressure, flow rate, and flow speed.

[0064] For water supply pipe network, its monitoring data is water pressure, flow, flow rate, and the monitoring data of other pipes is naturally different, and the application takes the water supply pipe network as an example, each parameter is for the water supply pipe network, and each item of historical accident record is obtained by the same method.

[0065] The water pressure, flow and flow rate in the historical monitoring data corresponding to a single historical accident record are sequentially numbered in time order, respectively C ai , C bi , C ci , wherein i=1, 2, …, n, n is the number of parameters, and the water pressure participation B a , flow participation B b , and flow rate participation B c at the accident occurrence moment are obtained.

[0066] ;

[0067] ;

[0068] ;

[0069] , C ax , C bx , C cx are the water pressure, flow and flow rate at the accident occurrence moment, and the above monitoring participation can reflect the change of the corresponding parameter when the safety accident occurs, and the change of different parameters under different safety accidents is also different.

[0070] The same method is used to obtain the water pressure participation, flow participation and flow rate participation corresponding to each item of historical accident record of the same accident type, and the monitoring participation corresponding to each item of historical accident record of different accident types is obtained, including water pressure participation, flow participation and flow rate participation.

[0071] It should be further pointed out that in the specific implementation process, the process of constructing the accident evaluation model according to the historical monitoring data, monitoring participation and monitoring level corresponding to the historical accident record of different accident types includes:

[0072] For various types of safety accidents, the accident evaluation model is constructed respectively, and for a single accident type, the accident evaluation set is generated according to the historical monitoring data, monitoring participation and monitoring level corresponding to each item of historical accident record of the same accident type, and is divided into training set and test set, and the monitoring level is the monitoring level of the monitoring area where the accident occurrence place of the historical accident record is located.

[0073] The convolutional neural network is constructed, different historical monitoring data, monitoring participation, monitoring levels in the training set are taken as input data of the convolutional neural network, and whether a safety accident occurs is taken as output data of the convolutional neural network, the convolutional neural network is trained to obtain an initial convolutional neural network;

[0074] The initial convolutional neural network is verified by using the test set, and the initial convolutional neural network with an output less than or equal to a preset test error threshold is taken as an accident assessment model corresponding to a safety accident of the accident type, and the same method is used to obtain the accident assessment model corresponding to the safety accident of different accident types.

[0075] It should be further explained that, in the specific implementation process, the current real-time monitoring data is obtained, and the monitoring level and the monitoring participation are input into different accident assessment models to determine whether a safety accident and the accident type thereof exist.

[0076] The parameters obtained by the monitoring units in each monitoring area within a double preset time length are taken as real-time monitoring data, and the average values of the monitoring participations corresponding to the historical accident records of the same accident type are taken as the assessment participations of the accident type, including the water pressure assessment participation, the flow assessment participation and the flow rate assessment participation.

[0077] The real-time monitoring data of a single monitoring area and the monitoring level corresponding thereto are input into the accident assessment model of the safety accident of the corresponding accident type in combination with the assessment participations of different accident types to determine whether a safety accident exists in the monitoring area.

[0078] If a safety accident exists, the accident type corresponding to the accident assessment model that determines that a safety accident exists is taken as the accident type of the safety accident of the monitoring area, and a corresponding alarm signal is generated, and the alarm signal is fed back to the relevant personnel to prompt them to operate and maintain the monitoring area, and if a safety accident does not exist, no other operation is performed.

[0079] In the embodiment of the present application, a community pipe network intelligent monitoring and early warning method based on the Internet of Things is also included, which comprises the following steps:

[0080] Step S1: Obtain the distribution information of the community pipe network and construct a distribution model, and divide the community pipe network into a plurality of unit blocks in the distribution model.

[0081] Step S2: Obtain the existing monitoring units and the existing monitoring areas, set the supplementary monitoring units and obtain the supplementary monitoring areas, obtain the historical accident records of different accident types, and obtain the monitoring levels of different monitoring areas according to the historical accident records.

[0082] Step S3: obtaining historical monitoring data of historical accident records of different accident types, obtaining monitoring participation degrees of different monitoring data according to the historical monitoring data, and constructing an accident assessment model according to historical monitoring data, monitoring participation degrees, and monitoring levels corresponding to historical accident records of different accident types;

[0083] Step S4: obtaining current real-time monitoring data, inputting the monitoring levels and the monitoring participation degrees into different accident assessment models to determine whether a safety accident exists and the type of the accident.

[0084] The above examples are only used to illustrate the technical method of the present application and are not limiting. Although the present application has been described in detail with reference to the preferred embodiments, it should be understood by those skilled in the art that the technical method of the present application can be modified or replaced equivalently without departing from the spirit and scope of the technical method of the present application.

Claims

1. An Internet of Things-based intelligent monitoring and early warning system for in-community pipe networks, characterized in that, The application comprises the following modules: a map construction module for obtaining distribution information of a pipe network in a community and constructing a distribution model thereof, wherein the pipe network in the community is divided into a plurality of unit blocks in the distribution model; a region division module for obtaining existing monitoring units and existing monitoring regions, setting supplementary monitoring units and obtaining supplementary monitoring regions, obtaining historical accident records of different accident types, and obtaining monitoring levels of different monitoring regions according to the historical accident records; a monitoring evaluation module for obtaining historical monitoring data of the historical accident records of different accident types, obtaining monitoring participation degrees of different monitoring data according to the historical monitoring data, and constructing an accident evaluation model according to the historical monitoring data, the monitoring participation degrees and the monitoring levels corresponding to the historical accident records of different accident types; an accident determination module for obtaining real-time monitoring data, inputting the monitoring levels and the monitoring participation degrees into different accident evaluation models to determine whether a safety accident exists and the type of the accident; the process of obtaining the historical monitoring data and the monitoring participation degrees of different monitoring data comprises: obtaining each parameter of each monitoring unit in a monitoring region where a single historical accident record is located within a preset time period before and after the occurrence time of the historical accident record as the historical monitoring data of the monitoring unit, including water pressure, flow rate and flow velocity; The water pressure, flow, and flow rate in the historical monitoring data corresponding to the single historical accident record are sequentially numbered in time order as C ai 、 bi 、 ci , i = 1, 2, …, n, n is the number of parameters, respectively, the water pressure participation B a , the flow participation B b , and the flow rate participation B c at the moment of the accident are obtained. ； ； ； C ax , C bx , C cx are water pressure, flow, and flow rate at the moment of the accident, respectively, the monitoring participation degrees corresponding to each historical accident record of the same accident type are obtained, including water pressure participation degree, flow participation degree, and flow rate participation degree, and the monitoring participation degrees corresponding to each historical accident record of different accident types are obtained, respectively. 2.The community pipe network intelligent monitoring and early warning system based on the Internet of Things according to claim 1, characterized in that, the process of constructing the distribution model and dividing the pipe network in the community into a plurality of unit blocks comprises: the distribution information refers to data related to the distribution of the pipe network for water supply, gas supply and power supply in the community, including pipe purpose, pipe location, pipe direction, pipe burial depth, pipe diameter and material, connection point and interface, valve and control equipment; a corresponding distribution model is constructed according to the obtained distribution information by using digital twinning technology, and the distribution models of pipe networks with different pipe purposes are independent of each other, and the distribution model is divided into a plurality of unit blocks of equal size. 3.The community pipe network intelligent monitoring and early warning system based on the Internet of Things according to claim 2, characterized in that, the process of obtaining the existing monitoring regions and the supplementary monitoring regions comprises: different pipe networks with different pipe purposes are marked as water supply pipe networks, gas supply pipe networks and power supply pipe networks, and different parameters of different pipe networks are monitored by using monitoring units, including water pressure monitoring units, flow rate monitoring units and flow velocity monitoring units; each monitoring unit already set in the current pipe network is marked as an existing monitoring unit, and each monitoring unit added is marked as a supplementary monitoring unit, and the positions of the existing monitoring units and the supplementary monitoring units are uploaded to the distribution model of the corresponding pipe network for synchronization; in the distribution model, the existing monitoring regions of each existing monitoring unit are obtained according to the effective monitoring range of each existing monitoring unit, and the supplementary monitoring regions of each supplementary monitoring unit are obtained according to the effective monitoring range of each supplementary monitoring unit, wherein the effective monitoring range refers to the maximum coverage range of the monitoring unit that can monitor the corresponding parameter.

4. The community pipe network intelligent monitoring and early warning system based on the Internet of Things according to claim 3, characterized in that, the process of obtaining the historical accident records and the monitoring levels of different monitoring regions comprises: the historical accident records refer to data related to safety accidents that have occurred in the pipe network in the community, including accident type, accident occurrence time, accident occurrence location and accident influence range, and each historical accident record is uploaded to the distribution model of the corresponding pipe network for synchronization according to the accident occurrence location. The number of historical accident records of each monitoring area, including the existing monitoring area and the supplementary monitoring area, is obtained, and an accident number standard is set. The number of historical accident records of each monitoring area is compared with the accident number standard respectively to obtain the monitoring level of the corresponding monitoring area, including a low monitoring level, a high monitoring level, and a medium monitoring level.

5. The community-based intelligent monitoring and early warning system for pipe networks based on the Internet of Things according to claim 4, characterized in that, The process of constructing the accident assessment model includes: For different types of safety accidents, corresponding accident assessment models are constructed respectively. An accident assessment set is generated according to the historical monitoring data, monitoring participation, and monitoring level corresponding to each historical accident record of the same type of accident, and the accident assessment set is divided into a training set and a test set. A convolutional neural network is constructed, different historical monitoring data, monitoring participation, and monitoring level in the training set are taken as input data of the convolutional neural network, whether a safety accident will occur is taken as output data of the convolutional neural network, the convolutional neural network is trained to obtain an initial convolutional neural network, and the initial convolutional neural network is verified by using the test set. The initial convolutional neural network with an output less than or equal to a preset test error threshold is taken as the accident assessment model corresponding to the safety accident of the type of accident, and the accident assessment models corresponding to different types of safety accidents are constructed respectively. The process of determining whether a safety accident and its type exist includes:

6. The community-based intelligent monitoring and early warning system for pipe networks based on the Internet of Things according to claim 5, characterized in that, The parameters obtained by the monitoring unit in each monitoring area within a double preset time length are taken as the real-time monitoring data of the monitoring unit, and the average of each monitoring participation corresponding to each historical accident record of the same type of accident is taken as the evaluation participation of the type of accident. The real-time monitoring data and the monitoring level of a single monitoring area are input into the accident assessment model of the corresponding type of safety accident in combination with the evaluation participation of different types of accidents to determine whether a safety accident exists in the monitoring area. If a safety accident exists, the type of accident corresponding to the accident assessment model that determines that a safety accident exists is taken as the type of accident of the safety accident of the monitoring area, and if a safety accident does not exist, no other operation is performed. The method includes:

7. A community pipe network intelligent monitoring and early warning method based on the Internet of Things, which is implemented based on the community pipe network intelligent monitoring and early warning system in any one of claims 1-6, characterized in that, Step S1: Obtain the distribution information of the pipe network in the community and construct a distribution model, in which the pipe network in the community is divided into a plurality of unit blocks; Step S2: Obtain the existing monitoring unit and the existing monitoring area, set a supplementary monitoring unit, obtain a supplementary monitoring area, obtain historical accident records of different types of accidents, and obtain the monitoring level of different monitoring areas according to the historical accident records; Step S3: Obtain the historical monitoring data of the historical accident records of different types of accidents, obtain the monitoring participation of different monitoring data according to the historical monitoring data, and construct an accident assessment model according to the historical monitoring data, monitoring participation, and monitoring level corresponding to the historical accident records of different types of accidents; Step S4: Obtain the current real-time monitoring data, input the monitoring level and the monitoring participation into different accident assessment models to determine whether a safety accident and its type exist. ​

Images