CN116303690B

CN116303690B - Fire-fighting data processing method and system based on big data

Info

Publication number: CN116303690B
Application number: CN202310599370.0A
Authority: CN
Inventors: 陈麒邦; 景佳妮; 邱文华; 李才贵; 于晶
Original assignee: Shenzhen Lianzheng Tongda Technology Co ltd
Current assignee: Shenzhen Lianzheng Tongda Technology Co ltd
Priority date: 2023-05-25
Filing date: 2023-05-25
Publication date: 2023-07-25
Anticipated expiration: 2043-05-25
Also published as: CN116303690A

Abstract

According to the fire control data processing method and system based on big data, the data representation vector of the target fire control monitoring data set and the data representation vector of the early fire control monitoring data set are loaded to the early warning information mining network, and the environment mapping representation vector of the corresponding target fire control monitoring data set is obtained. Then, the embedded mapping is used for obtaining an embedded mapping characterization vector of the target fire control monitoring data set, the embedded mapping characterization vector and the environment mapping characterization vector are determined to be loading data of the fire control early warning network, an analysis result is obtained through analysis, an early warning label sequence corresponding to the target fire control monitoring data set is obtained, the analysis result obtained by early warning analysis of the current target fire control monitoring data set is prevented from generating deviation, and accuracy of early warning analysis is improved.

Description

Fire-fighting data processing method and system based on big data

Technical Field

The application relates to the fields of data processing, artificial intelligence and intelligent fire control, in particular to a fire control data processing method and system based on big data.

Background

Along with the acceleration of smart city, fire control is as the important guarantee of city safety, and traditional fire control system is difficult to satisfy present fire control demand when coping with the intelligent city of growing day by day, and people attach more importance and urgency to the construction of wisdom fire control to with the rapid development of internet of things, for intelligent fire control system provides wireless communication's technical support, has promoted traditional fire control to intelligent transformation. Aiming at the problems of the traditional fire-fighting system, such as periodic inspection of the building, fire-fighting early warning and the like, the problems are difficult to discover, inspect and process in time. Therefore, based on intelligent analysis of fire control big data generated in the fire control monitoring Internet of things system, the safety state of the fire control monitoring area is automatically determined, the judgment of the necessity of early warning triggering is the focus of the current field, and how to guarantee the accuracy of early warning analysis is the technical problem that needs to be focused.

Disclosure of Invention

The invention aims to provide a fire control data processing method and system based on big data so as to solve the problems.

The implementation manner of the embodiment of the application is as follows:

in a first aspect, an embodiment of the present application provides a fire protection data processing method based on big data, which is applied to an internet of things cloud platform, and the method includes:

Acquiring a data representation vector of a target firefighting monitoring data set and a data representation vector of a pre-firefighting monitoring data set, wherein the pre-firefighting monitoring data set is one or more firefighting monitoring data sets positioned in front of the target firefighting monitoring data set;

determining the data characterization vector of the target firefighting monitoring data set and the data characterization vector of the early firefighting monitoring data set as loading data of an early warning information mining network to obtain an environment mapping characterization vector of the target firefighting monitoring data set;

performing embedded mapping on the data characterization vector of the target firefighting monitoring data set to obtain an embedded mapping characterization vector of the target firefighting monitoring data set;

analyzing the embedded mapping characterization vector of the target firefighting monitoring data set and the environment mapping characterization vector of the target firefighting monitoring data set through a firefighting early warning network to obtain an analysis result of the target firefighting monitoring data set, wherein the analysis result characterizes an early warning tag sequence corresponding to the target firefighting monitoring data set.

As an implementation manner, the determining the data representation vector of the target fire protection monitoring data set and the data representation vector of the early fire protection monitoring data set as the loading data of the early warning information mining network to obtain the environment mapping representation vector of the target fire protection monitoring data set includes:

Determining the data characterization vector of the target firefighting monitoring data set as the loading data of the early warning information mining network to obtain the early warning information characterization vector of the target firefighting monitoring data set;

determining the data characterization vector of the early-stage fire control monitoring data set as the loading data of the early-warning information mining network to obtain the early-warning information characterization vector of the early-stage fire control monitoring data set;

and interacting the early warning information characterization vector of the target firefighting monitoring data set with the early warning information characterization vector of the early firefighting monitoring data set to obtain the environment mapping characterization vector of the target firefighting monitoring data set.

As an implementation manner, the analyzing, by the fire protection early warning network, the embedded mapping characterization vector of the target fire protection monitoring data set and the environment mapping characterization vector of the target fire protection monitoring data set to obtain an analysis result of the target fire protection monitoring data set includes:

loading the embedded mapping characterization vector of the target firefighting monitoring data set and the environment mapping characterization vector of the target firefighting monitoring data set to the firefighting early warning network to obtain the confidence of the reasoning type of each data item in the target firefighting monitoring data set;

And determining the analysis result of the target firefighting monitoring data set through the reasoning type confidence of each data item.

As an implementation manner, the determining the data characterization vector of the target firefighting monitoring data set as the loading data of the early warning information mining network to obtain the early warning information characterization vector of the target firefighting monitoring data set includes:

combining the data characterization vector of the target firefighting monitoring data set with a filling array with the set element number to obtain a combined first data characterization vector;

loading the combined first data characterization vector into the early warning information mining network to obtain an early warning information characterization vector of the target firefighting monitoring data set;

the determining the data characterization vector of the early-stage fire control monitoring data set as the loading data of the early-warning information mining network to obtain the early-warning information characterization vector of the early-stage fire control monitoring data set comprises the following steps:

combining the data representation vector of the early-stage fire control monitoring data set with a filling array with the set element number to obtain a combined second data representation vector;

and loading the combined second data characterization vector into the early warning information mining network to obtain the early warning information characterization vector of the early fire control monitoring data set.

As an embodiment, the acquiring the data characterization vector of the target firefighting monitoring data set includes:

acquiring the target firefighting monitoring data set;

loading the target firefighting monitoring data set into a data representation vector extraction network to obtain a data representation vector of the target firefighting monitoring data set, wherein the data representation vector extraction network is a neural network obtained after debugging by taking a firefighting monitoring data set template as a debugging sample;

the method further comprises the steps of:

acquiring a data characterization vector of a current fire control monitoring data set template and an early warning tag sequence characterization vector corresponding to the current fire control monitoring data set template;

vector erasure or vector shielding processing is carried out on the data representation vector of the current fire control monitoring data set template to obtain a first fire control monitoring data set template representation vector, and vector erasure or vector shielding processing is carried out on the early warning label sequence representation vector corresponding to the current fire control monitoring data set template to obtain a first early warning label sequence representation vector;

acquiring a data characterization vector of a front-stage fire control monitoring data set template and a front-stage early warning tag sequence characterization vector corresponding to the front-stage fire control monitoring data set template;

Vector erasure or vector shielding processing is carried out on the data representation vector of the front-stage fire control monitoring data set template to obtain a first front-stage fire control monitoring data set template representation vector, vector erasure or vector shielding processing is carried out on the front-stage early-warning tag sequence representation vector corresponding to the front-stage fire control monitoring data set template to obtain a first front-stage early-warning tag sequence representation vector;

and repeatedly debugging the early warning information mining network through the first fire control monitoring data set template characterization vector, the first early warning tag sequence characterization vector, the first early fire control monitoring data set template characterization vector and the first early warning tag sequence characterization vector until a preset debugging stop condition is met.

As an embodiment, the method further comprises:

loading the first fire control monitoring data set template characterization vector and the first early warning tag sequence characterization vector into the early warning information mining network to obtain a first early warning information characterization vector;

acquiring iteration cost through errors between the first early warning information characterization vector and the early warning information characterization vector of the current fire control monitoring dataset template; the early warning information characterization vector of the current fire control monitoring data set template is an early warning information characterization vector obtained after the data characterization vector of the current fire control monitoring data set template and the early warning label sequence characterization vector corresponding to the current fire control monitoring data set template are loaded to the early warning information mining network;

Adjusting a network configuration variable of the early warning information mining network through the iteration cost to obtain the early warning information mining network;

the method further comprises the steps of:

respectively loading the data characterization vector of the current fire control monitoring data set template and the characterization vector of the first fire control monitoring data set template into the early warning information mining network to obtain an early warning characterization vector corresponding to the current fire control monitoring data set template and an early warning characterization vector corresponding to the first fire control monitoring data set template characterization vector;

acquiring data cost through errors between the early warning characterization vector corresponding to the current fire control monitoring data set template and the early warning characterization vector corresponding to the first fire control monitoring data set template characterization vector;

and adjusting the network configuration variables of the early warning information mining network through the data cost to obtain the early warning information mining network.

As an implementation manner, the obtaining the data cost through the error between the pre-warning characterization vector corresponding to the current fire protection monitoring data set template and the pre-warning characterization vector corresponding to the first fire protection monitoring data set template characterization vector includes:

Acquiring a first commonality measurement result between an early warning characterization vector corresponding to the current fire control monitoring data set template and an early warning characterization vector corresponding to the first fire control monitoring data set template characterization vector and a second commonality measurement result between the early warning characterization vector corresponding to the current fire control monitoring data set template and a preset data characterization vector, wherein the preset data characterization vector is a data characterization vector which is not related to the current fire control monitoring data set template;

and determining data cost through the first commonality measurement result and the second commonality measurement result.

As an embodiment, the method further comprises:

respectively loading an early warning tag sequence characterization vector and the first early warning tag sequence characterization vector corresponding to the current fire control monitoring data set template into the early warning information mining network to obtain an early warning characterization vector and an early warning characterization vector of the first early warning tag sequence characterization vector of the early warning tag sequence characterization vector corresponding to the current fire control monitoring data set template;

acquiring early warning tag sequence cost through errors between early warning characterization vectors of the early warning tag sequence characterization vectors corresponding to the current fire control monitoring data set template and the early warning characterization vectors of the first early warning tag sequence characterization vectors;

And adjusting the network configuration variables of the early warning information mining network through the early warning tag sequence cost to obtain the early warning information mining network.

As an implementation manner, the obtaining, by the error between the early warning characterization vector of the early warning tag sequence characterization vector corresponding to the current fire protection monitoring dataset template and the early warning characterization vector of the first early warning tag sequence characterization vector, the early warning tag sequence cost includes:

determining a second early warning tag sequence characterization vector from the early warning tag sequence characterization vectors corresponding to the current fire control monitoring dataset template; the second early warning tag sequence characterization vector is a characterization vector except the first early warning tag sequence characterization vector in the early warning tag sequence characterization vectors corresponding to the current fire control monitoring data set template;

loading the second early warning tag sequence characterization vector into the early warning information mining network to obtain an early warning characterization vector of the second early warning tag sequence characterization vector;

and acquiring cross entropy between the early warning characterization vector of the second early warning tag sequence characterization vector and the early warning characterization vector of the first early warning tag sequence characterization vector, and obtaining the early warning tag sequence cost.

On the other hand, the embodiment of the application provides a fire control internet of things system, including thing networking sensor, communication device and thing networking cloud platform, thing networking sensor pass through communication device with thing networking cloud platform communication connection, thing networking cloud platform includes memory and treater, the memory stores computer program, works as the treater operation computer program is when realizing above-mentioned method.

The embodiment of the application at least comprises the following beneficial effects:

according to the fire control data processing method and system based on big data, after the data representation vector of the target fire control monitoring data set and the data representation vector of the early fire control monitoring data set are obtained, the data representation vector of the target fire control monitoring data set and the data representation vector of the early fire control monitoring data set are loaded to the early warning information mining network, so that the environment mapping representation vector of the corresponding target fire control monitoring data set is obtained. Then, the data characterization vector of the target fire control monitoring data set is subjected to embedded mapping to obtain an embedded mapping characterization vector of the target fire control monitoring data set, the embedded mapping characterization vector of the target fire control monitoring data set and the environment mapping characterization vector of the target fire control monitoring data set are determined to be loading data of a fire control early warning network together, so that an analysis result of the target fire control monitoring data set is obtained through analysis, and an early warning tag sequence corresponding to the target fire control monitoring data set is obtained. That is, in the embodiment of the application, the early warning analysis of the current target fire-fighting monitoring dataset is performed, the environmental information is mined from the overall monitoring data corresponding to the target fire-fighting monitoring dataset, the data characterization vector of the current target fire-fighting monitoring dataset and the corresponding environmental mapping characterization vector are included in the analysis, and the analysis result obtained by the early warning analysis of the current target fire-fighting monitoring dataset is prevented from deviating only by the analysis result of the early-stage fire-fighting monitoring dataset, so that the accuracy of the early warning analysis is improved.

In the following description, other features will be partially set forth. Upon review of the ensuing disclosure and the accompanying figures, those skilled in the art will in part discover these features or will be able to ascertain them through production or use thereof. The features of the present application may be implemented and obtained by practicing or using the various aspects of the methods, tools, and combinations that are set forth in the detailed examples described below.

Drawings

In order to more clearly illustrate the technical solutions of the embodiments of the present application, the drawings that are needed in the embodiments will be briefly described below, it being understood that the following drawings only illustrate some embodiments of the present application and therefore should not be considered limiting the scope, and that other related drawings may be obtained according to these drawings without inventive effort for a person skilled in the art.

The methods, systems, and/or programs in the accompanying drawings will be described further in terms of exemplary embodiments. These exemplary embodiments will be described in detail with reference to the drawings. These exemplary embodiments are non-limiting exemplary embodiments, wherein reference numerals represent similar mechanisms throughout the several views of the drawings.

Fig. 1 is a schematic diagram of the components of a fire monitoring internet of things system according to some embodiments of the present application.

Fig. 2 is a schematic diagram of hardware and software components in an internet of things cloud platform according to some embodiments of the present application.

Fig. 3 is a flow chart illustrating a fire data processing method based on big data according to some embodiments of the present application.

Fig. 4 is a schematic architecture diagram of a fire protection data processing device according to an embodiment of the present application.

Detailed Description

In order to better understand the technical solutions described above, the following detailed description of the technical solutions of the present application is provided through the accompanying drawings and specific embodiments, and it should be understood that the specific features of the embodiments and embodiments of the present application are detailed descriptions of the technical solutions of the present application, and not limit the technical solutions of the present application, and the technical features of the embodiments and embodiments of the present application may be combined with each other without conflict.

In the following detailed description, numerous specific details are set forth by way of examples in order to provide a thorough understanding of the relevant teachings. However, it will be apparent to one skilled in the art that the present application may be practiced without these details. In other instances, well-known methods, procedures, systems, components, and/or circuits have been described at a relatively high-level, without detail, in order to avoid unnecessarily obscuring aspects of the present application.

These and other features, together with the functions, acts, and combinations of parts and economies of manufacture of the related elements of structure, all of which form part of this application, may become more apparent upon consideration of the following description with reference to the accompanying drawings. It is to be expressly understood, however, that the drawings are for the purpose of illustration and description only and are not intended as a definition of the limits of the application. It should be understood that the drawings are not to scale. It is to be expressly understood, however, that the drawings are for the purpose of illustration and description only and are not intended as a definition of the limits of the application. It should be understood that the figures are not to scale.

The flowcharts are used in this application to describe implementations performed by systems according to embodiments of the present application. It should be clearly understood that the execution of the flowcharts may be performed out of order. Rather, these implementations may be performed in reverse order or concurrently. Additionally, at least one other execution may be added to the flowchart. One or more of the executions may be deleted from the flowchart.

Fig. 1 is a schematic diagram of a fire monitoring internet of things system 400 according to some embodiments of the present application, the fire monitoring internet of things system 400 including an internet of things cloud platform 100 and an internet of things sensor 300 communicatively connected to each other through a communication device 200. The internet of things cloud platform 100 is, for example, a server.

In some embodiments, please refer to fig. 2, which is a schematic architecture diagram of an internet of things cloud platform 100, the internet of things cloud platform 100 includes a fire data processing device 110, a memory 120, a processor 130 and a communication unit 140. The memory 120, the processor 130, and the communication unit 140 are electrically connected directly or indirectly to each other to realize data transmission or interaction. For example, the components may be electrically connected to each other via one or more communication buses or signal lines. The fire data processing device 110 includes at least one software function module that may be stored in the memory 120 in the form of software or firmware (firmware) or cured in an Operating System (OS) of the internet of things cloud platform 100. The processor 130 is configured to execute executable modules stored in the memory 120, such as software functional modules and computer programs included in the fire protection data processing device 110.

The Memory 120 may be, but is not limited to, a random access Memory (Random Access Memory, RAM), a Read Only Memory (ROM), a programmable Read Only Memory (Programmable Read-Only Memory, PROM), an erasable Read Only Memory (Erasable Programmable Read-Only Memory, EPROM), an electrically erasable Read Only Memory (Electric Erasable Programmable Read-Only Memory, EEPROM), etc. The memory 120 is used for storing a program, and the processor 130 executes the program after receiving an execution instruction. The communication unit 140 is configured to establish a communication connection between the internet of things cloud platform 100 and the internet of things sensor 300 through a network, and is configured to send and receive data through the network.

The processor may be an integrated circuit chip having signal processing capabilities. The processor may be a general-purpose processor, including a central processing unit (Central Processing Unit, CPU), a network processor (Network Processor, NP), etc.; but also Digital Signal Processors (DSPs)), application Specific Integrated Circuits (ASICs), field Programmable Gate Arrays (FPGAs) or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components. The disclosed methods, steps, and logic blocks in the embodiments of the present invention may be implemented or performed. A general purpose processor may be a microprocessor or the processor may be any conventional processor or the like.

It is to be appreciated that the structure shown in fig. 2 is merely illustrative, and that the internet of things cloud platform 100 may also include more or fewer components than shown in fig. 2, or have a different configuration than shown in fig. 2. The components shown in fig. 2 may be implemented in hardware, software, or a combination thereof.

Fig. 3 is a flowchart of a fire fighting data processing method based on big data according to some embodiments of the present application, where the method is applied to the internet of things cloud platform 100 in fig. 1, and may specifically include the following steps S110 to S140. On the basis of the following steps, alternative embodiments will be described, which should be understood as examples and should not be interpreted as essential features for implementing the present solution.

Step S110: and acquiring the data representation vector of the target firefighting monitoring data set and the data representation vector of the early firefighting monitoring data set.

In this embodiment of the application, the fire control monitoring dataset is the fire control monitoring dataset that needs to analyze and obtain early warning label sequence, the data that fire control monitoring dataset contained is based on all kinds of thing networking sensor sensing data in the fire control monitoring thing networking system, for example, intelligent smoke sensor, intelligent combustible gas sensor, fire control water source liquid level sensor, fire hydrant water pressure sensor, temperature and humidity sensor etc. the smog concentration, combustible gas concentration, water level, water pressure, temperature, humidity etc. that sense, fire control data such as via communication device, for example NB-IoT sends thing networking cloud platform, obtain the fire control dataset by thing networking cloud platform arrangement. The fire control monitoring data set can comprise a plurality of data classes, the number of the data classes is not limited, for example, the fire control monitoring data set comprises 4 data classes such as smoke, combustible gas, water level, water pressure and the like, each data class comprises at least one data item, and each data item corresponds to the data (such as smoke concentration data) of the Internet of things collected by the sensing equipment at a collection time point. The fire control monitoring data set corresponds to monitoring data of a time period, such as one hour, 10 minutes, and the like, the fire control monitoring data set corresponds to monitoring data of a target area, and is not limited in particular, for the same target area, the corresponding fire control monitoring data set comprises a front fire control monitoring data set and a target fire control monitoring data set, wherein the front fire control monitoring data set is the previous fire control monitoring data set, and the target fire control monitoring data set is the current fire control monitoring data set. The fire monitoring data sets are arranged in sequence according to the generated time, and the number of the front fire monitoring data sets can be one or more, in other words, the front fire monitoring data sets are one or more fire monitoring data sets positioned in front of the target fire monitoring data set. The data characterization vector of the fire control monitoring data set is a vector characterization result obtained by extracting the characterization vector of the fire control monitoring data set, and the dimension of the vector characterization result is related to the number of the fire control monitoring data set. It should be noted that in the fire control monitoring internet of things system, the selection of the hardware architecture is not limited, for example, the fire control data acquisition terminal master controller may be based on low-power consumption embedded equipment (such as STM32L431RCT6 chip), and the access mode may be CoAP protocol or the like.

As an embodiment, the process of acquiring the data characterization vector of the target firefighting monitoring data set may be to acquire the target firefighting monitoring data set first; loading the target fire control monitoring data set into a data representation vector extraction network to obtain a data representation vector of the target fire control monitoring data set, wherein the data representation vector extraction network is a neural network obtained after a fire control monitoring data set template is used as a debugging sample.

Because the data characterization vector extraction network is a characterization vector extraction network which is already debugged in advance, the data characterization vector extraction network is a neural network which is obtained by taking the extracted data characterization vector as a debugging target and taking the fire control monitoring data set template as a debugging sample. And processing the target firefighting monitoring data set through a data representation vector extraction network after the target firefighting monitoring data set is obtained, so as to obtain a data representation vector of the corresponding target firefighting monitoring data set. Correspondingly, after the target firefighting monitoring data set is obtained, the early firefighting monitoring data set of the target firefighting monitoring data set can be obtained through time sequence. Thus, after the early-stage fire control monitoring data set is obtained, the early-stage fire control monitoring data set can be loaded to the data representation vector extraction network, and the data representation vector of the corresponding early-stage fire control monitoring data set is extracted.

Step S120: and determining the data characterization vector of the target fire control monitoring data set and the data characterization vector of the early fire control monitoring data set as loading data of the early warning information mining network to obtain the environment mapping characterization vector of the target fire control monitoring data set.

In the embodiment of the application, the early warning information mining network is a neural network obtained by taking an environmental mapping representation vector of a current fire control monitoring data set template as a debugging target, and taking a data representation vector of the current fire control monitoring data set template, an early warning tag sequence representation vector corresponding to the current fire control monitoring data set template, a data representation vector of a front fire control monitoring data set template and a front early warning tag sequence representation vector corresponding to the front fire control monitoring data set template as debugging samples. For the extraction mode of the data characterization vector of the current fire control monitoring data set template and the data characterization vector of the early fire control monitoring data set template, please refer to the above process. The environment mapping characterization vector characterizes the contextual characteristic information of the current fire monitoring dataset template. And determining a corresponding early warning label sequence after obtaining the current fire control monitoring data set template for the early warning label sequence characterization vector corresponding to the current fire control monitoring data set template. And extracting the characterization vector of the tag information of the current fire control monitoring data set template through a pre-debugged early warning tag sequence characterization vector extraction network, so as to obtain an early warning tag sequence characterization vector corresponding to the corresponding current fire control monitoring data set template.

Then, after obtaining the data representation vector of the current fire-fighting monitoring dataset template, the pre-warning tag sequence representation vector corresponding to the current fire-fighting monitoring dataset template, the data representation vector of the pre-fire-fighting monitoring dataset template and the pre-warning tag sequence representation vector corresponding to the pre-fire-fighting monitoring dataset template, taking the environment mapping representation vector of the current fire-fighting monitoring dataset template as a debugging target, debugging the obtained data representation vector of the current fire-fighting monitoring dataset template, the pre-warning tag sequence representation vector corresponding to the current fire-fighting monitoring dataset template, the data representation vector of the pre-fire-fighting monitoring dataset template and the pre-warning tag sequence representation vector corresponding to the pre-fire-fighting monitoring dataset template, and the pre-warning information mining network which is aided in debugging can extract the environment information (context information) of the current fire-fighting monitoring dataset template in the fire-fighting monitoring dataset sequence.

Optionally, before the extraction of the environment mapping characterization vector by the early warning information mining network, the early warning information mining network needs to be generated and debugged to obtain the early warning information mining network. Specifically, in the network debugging process of the early warning information mining network, vector erasure and/or vector shielding are carried out on the obtained data characterization vector of the current fire control monitoring data set template so as to obtain the first fire control monitoring data set template characterization vector. For example, the data characterization vector of the current fire monitoring dataset template is (V ₁ 、V ₂ 、V ₃ ) For V ₁ Erasing or shielding to obtain a first firefighting monitoring data set template characterization vector (V ₂ 、V ₃ ). And similarly, vector erasure and/or vector shielding are also carried out on the early warning tag sequence characterization vector corresponding to the current fire control monitoring data set template, so as to obtain a first early warning tag sequence characterization vector. Similarly, the data set model for early-stage fire control monitoringAnd carrying out vector erasure and/or vector shielding on the data representation vector of the plate and the front-stage early warning label sequence representation vector corresponding to the front-stage fire control monitoring data set template to obtain a corresponding first front-stage fire control monitoring data set template representation vector and a corresponding first front-stage early warning label sequence representation vector.

Based on the above, after the first fire control monitoring data set template characterization vector, the first early warning tag sequence characterization vector, the first early fire control monitoring data set template characterization vector and the first early warning tag sequence characterization vector are obtained, the first fire control monitoring data set template characterization vector, the first early warning tag sequence characterization vector, the first early fire control monitoring data set template characterization vector and the first early warning tag sequence characterization vector are used as debugging samples, the early warning information mining network is repeatedly debugged until a preset debugging stop condition is met, and the debugging requirement of extracting the environment mapping characterization vector of the current fire control monitoring data set template is met.

The costs used in debugging the early warning information mining network include, for example, iteration costs, data costs, and tag information costs, for example:

when the network is mined based on iteration cost debugging early warning information:

optionally, the fire protection data processing method based on big data further comprises: loading a first fire control monitoring data set template characterization vector and a first early warning tag sequence characterization vector into an early warning information mining network to obtain a first early warning information characterization vector; and acquiring iteration cost through errors between the first early warning information characterization vector and the early warning information characterization vector of the current fire control monitoring dataset template. The early warning information characterization vector of the current fire control monitoring data set template is an early warning information characterization vector obtained after loading the data characterization vector of the current fire control monitoring data set template and the early warning label sequence characterization vector corresponding to the current fire control monitoring data set template into an early warning information mining network; and adjusting network configuration variables of the early warning information mining network through iteration cost to obtain the early warning information mining network.

For example, when an output result of iteration cost is obtained, a first fire control monitoring data set template characterization vector and a first early warning label sequence characterization vector are loaded to an early warning information mining network to obtain a first early warning information characterization vector. The first early warning information characterization vector is used for erasing or shielding the characterization vector after the partial fire control monitoring data set template characterization vector and the early warning label sequence characterization vector, and then the iteration cost is calculated based on the error between the acquired first early warning information characterization vector and the early warning information characterization vector of the current fire control monitoring data set template. Based on the method, the network configuration variables (namely network parameters) of the early warning information mining network can be continuously adjusted through iteration cost, and then the early warning information mining network is obtained through debugging. The early warning information characterization vector of the current fire control monitoring data set template is an early warning information characterization vector obtained after the data characterization vector of the current fire control monitoring data set template and the early warning label sequence characterization vector corresponding to the current fire control monitoring data set template are loaded to the early warning information mining network. That is, the warning information characterization vector of the current fire monitoring dataset template does not erase or mask the characterization vector of the partial fire monitoring dataset template characterization vector and the characterization vector before the warning tag sequence characterization vector.

When the network is mined based on data cost debugging early warning information:

as one implementation mode, the network configuration variables of the early warning information mining network are adjusted according to the output result of the data cost, and the early warning information mining network is obtained. The method may further comprise: respectively loading the data characterization vector of the current fire control monitoring data set template and the first fire control monitoring data set template characterization vector into an early warning information mining network to obtain an early warning characterization vector corresponding to the current fire control monitoring data set template and an early warning characterization vector corresponding to the first fire control monitoring data set template characterization vector; acquiring data cost through errors between the early warning characterization vector corresponding to the current fire control monitoring data set template and the early warning characterization vector corresponding to the first fire control monitoring data set template characterization vector; and adjusting network configuration variables of the early warning information mining network through data cost to obtain the early warning information mining network. In the embodiment of the application, the early warning characterization vectors corresponding to the current fire control monitoring data set templates, namely the early warning characterization vectors before the data characterization vectors of all the current fire control monitoring data set templates are not erased or shielded. The early warning characterization vector corresponding to the characterization vector of the first fire control monitoring data set template is the early warning characterization vector after the partial data characterization vector of the current fire control monitoring data set template is erased or shielded.

And obtaining data cost by obtaining errors between the early warning characterization vector corresponding to the characterization vector of the first fire control monitoring data set template and the early warning characterization vector of the current fire control monitoring data set template, continuously adjusting network configuration variables of the early warning information mining network through the data cost, and debugging to obtain the early warning information mining network.

For example, for obtaining the data cost, the data cost may also be obtained by an error between an early warning characterization vector corresponding to the current fire protection monitoring dataset template and an early warning characterization vector corresponding to the first fire protection monitoring dataset template characterization vector, including: acquiring a first commonality measurement result between an early warning characterization vector corresponding to the current fire control monitoring data set template and an early warning characterization vector corresponding to the first fire control monitoring data set template characterization vector and a second commonality measurement result between the early warning characterization vector corresponding to the current fire control monitoring data set template and a preset data characterization vector, wherein the preset data characterization vector is a data characterization vector which is not related to the current fire control monitoring data set template; and determining the data cost through the first commonality measurement result and the second commonality measurement result. In the embodiment of the application, a first commonality measurement result between the early warning characterization vector corresponding to the current fire control monitoring data set template and the early warning characterization vector corresponding to the first fire control monitoring data set template characterization vector can be obtained in a manner of calculating the Euclidean distance, and a second commonality measurement result between the early warning characterization vector corresponding to the current fire control monitoring data set template and the preset data characterization vector is obtained. By a first commonality measure S ₁ And a second commonality measurement S ₂ And determining the data cost. For example, the data cost L is calculated based on the following formula:

L=-log(S ₁ /(S ₁ +S ₂ ))

when the network is mined based on tag information cost debugging early warning information:

as one implementation mode, the network configuration variables of the early warning information mining network are adjusted through the output result of the tag information cost, and the early warning information mining network is obtained. The method may further comprise the steps of: respectively loading the early warning tag sequence characterization vector and the first early warning tag sequence characterization vector corresponding to the current fire control monitoring data set template into an early warning information mining network to obtain the early warning characterization vector and the first early warning tag sequence characterization vector of the early warning tag sequence characterization vector corresponding to the current fire control monitoring data set template; acquiring early warning tag sequence cost through errors between early warning characterization vectors of early warning tag sequence characterization vectors corresponding to the current fire control monitoring data set template and early warning characterization vectors of the first early warning tag sequence characterization vectors; and adjusting network configuration variables of the early warning information mining network through early warning label sequence cost to obtain the early warning information mining network.

In the embodiment of the application, the early warning characterization vectors of the early warning tag sequence characterization vectors corresponding to the current fire monitoring data set templates, namely the characterization vectors before all the early warning tag sequence characterization vectors corresponding to the current fire monitoring data set templates are not erased or shielded. And the pre-warning characterization vector corresponding to the first pre-warning tag sequence characterization vector is the characterization vector after the partial pre-warning tag sequence characterization vector of the current fire control monitoring data set template is erased or shielded. And then, acquiring the early warning tag sequence cost by acquiring an error between an early warning characterization vector of the early warning tag sequence characterization vector corresponding to the current fire control monitoring data set template and an early warning characterization vector corresponding to the first early warning tag sequence characterization vector, and continuously adjusting a network configuration variable of the early warning information mining network through the early warning tag sequence cost, and debugging to obtain the early warning information mining network.

For example, in the process of acquiring the early warning tag sequence cost, the early warning tag sequence cost may also be acquired based on the following process, specifically, through an error between an early warning characterization vector of the early warning tag sequence characterization vector corresponding to the current fire control monitoring data set template and an early warning characterization vector of the first early warning tag sequence characterization vector, including: determining a second early warning tag sequence characterization vector in the early warning tag sequence characterization vector corresponding to the current fire monitoring data set template, wherein the second early warning tag sequence characterization vector is a characterization vector except the first early warning tag sequence characterization vector in the early warning tag sequence characterization vector corresponding to the current fire monitoring data set template; loading the second early warning tag sequence characterization vector into an early warning information mining network to obtain an early warning characterization vector of the second early warning tag sequence characterization vector; and acquiring cross entropy between the early warning characterization vector of the second early warning tag sequence characterization vector and the early warning characterization vector of the first early warning tag sequence characterization vector, and obtaining early warning tag sequence cost.

In the embodiment of the application, the first early warning tag sequence characterization vector is the residual early warning tag sequence characterization vector after vector erasure or shielding is performed on the early warning tag sequence characterization vector corresponding to the current fire control monitoring data set template. The second warning tag sequence characterization vector may be determined from the warning tag sequence characterization vectors corresponding to the current fire monitoring dataset template, in other words, the second warning tag sequence characterization vector is a characterization vector other than the first warning tag sequence characterization vector in the warning tag sequence characterization vectors corresponding to the current fire monitoring dataset template. The second warning label sequence representation vector is a warning label sequence representation vector which is erased or shielded in the warning label sequence representation vector corresponding to the current fire control monitoring data set template. And then loading the first early warning tag sequence characterization vector into an early warning information mining network to obtain an early warning characterization vector of the corresponding first early warning tag sequence characterization vector. And similarly, loading the second early warning tag sequence characterization vector into an early warning information mining network to obtain an early warning characterization vector of the corresponding second early warning tag sequence characterization vector. Therefore, the cross entropy between the early warning characterization vector of the second early warning label sequence characterization vector and the early warning characterization vector of the first early warning label sequence characterization vector is obtained based on the preset label information cost, and the corresponding early warning label sequence cost can be obtained.

Based on the early warning information mining network obtained by the process debugging, the environment mapping characterization vector of the target firefighting monitoring data set can be extracted.

In other embodiments, step S120 determines a data representation vector of the target fire protection monitoring data set and a data representation vector of the early fire protection monitoring data set as loading data of the early warning information mining network, so as to obtain an environment mapping representation vector of the target fire protection monitoring data set, which may specifically include:

step S121: and determining the data characterization vector of the target firefighting monitoring data set as loading data of the early warning information mining network to obtain the early warning information characterization vector of the target firefighting monitoring data set.

In the embodiment of the application, after the data characterization vector of the target fire control monitoring data set is obtained, the data characterization vector of the target fire control monitoring data set can be directly determined to be the loading data of the early warning information mining network, so that the early warning information characterization vector of the target fire control monitoring data set is obtained, and the early warning information characterization vector of the target fire control monitoring data set is the characterization vector of the current target fire control monitoring data set.

For example, determining the data characterization vector of the target firefighting monitoring data set as the loading data of the early warning information mining network, to obtain the early warning information characterization vector of the target firefighting monitoring data set includes: combining (e.g., stitching) the data characterization vector of the target firefighting monitoring data set and the filling array of the set element number to obtain a combined first data characterization vector; and loading the combined first data characterization vector into an early warning information mining network to obtain an early warning information characterization vector of the target firefighting monitoring data set. The padding array may be a one-dimensional array (i.e., vector) in nature. In the embodiment of the application, in the debugging process of the early warning information mining network, the early warning information mining network takes a sample group generated by the data characterization vector and the early warning tag sequence characterization vector as a debugging sample, but when the early warning information mining network is called, only the data characterization vector of the target firefighting monitoring data set is loaded, so that the number of elements corresponding to the early warning tag sequence characterization vector in the initial debugging process can be filled by filling an array. For example, combining the data characterization vector of the target firefighting monitoring data set with a filling array with the number of set elements to obtain a combined first data characterization vector, and loading the combined first data characterization vector into an early warning information mining network to obtain an early warning information characterization vector of the target firefighting monitoring data set.

The early warning information characterization vector of the target fire control monitoring data set can embody characterization information independently learned in the target fire control monitoring data set, and the element quantity of the filling array is real.

Step S122: and determining the data characterization vector of the early-stage fire control monitoring data set as loading data of the early-warning information mining network to obtain the early-warning information characterization vector of the early-stage fire control monitoring data set.

In the embodiment of the application, after the data characterization vector of the early-stage fire control monitoring data set is obtained, the data characterization vector of the early-stage fire control monitoring data set can be directly determined as the loading data of the early-warning information mining network, so that the early-warning information characterization vector of the early-stage fire control monitoring data set is obtained. For example, determining the data characterization vector of the early-stage fire control monitoring data set as the loading data of the early-stage fire control monitoring data set to obtain the early-stage fire control monitoring data set may specifically include: combining the data characterization vector of the early-stage fire control monitoring data set and a filling array with the set element number to obtain a combined second data characterization vector; and loading the combined second data characterization vector into an early warning information mining network to obtain an early warning information characterization vector of the early fire control monitoring data set.

Step S123: and interacting the early warning information characterization vector of the target fire control monitoring data set with the early warning information characterization vector of the early fire control monitoring data set to obtain the environment mapping characterization vector of the target fire control monitoring data set.

In the embodiment of the application, after the early warning information representation vector of the target fire control monitoring data set and the early warning information representation vector of the early fire control monitoring data set are obtained, the early warning information representation vector of the target fire control monitoring data set and the early warning information representation vector of the early fire control monitoring data set can be interacted. And the interacted early warning information characterization vector is used as an environment mapping characterization vector of the target firefighting monitoring data set. For example, the pre-warning information characterization vector of the target fire control monitoring data set is combined with the pre-warning information characterization vector of the early-stage fire control monitoring data set according to the time sequence, so as to obtain the environment mapping characterization vector of the target fire control monitoring data set.

Based on this, an environmental map representation vector of the target fire monitoring dataset is obtained.

Step S130: and carrying out embedded mapping on the data characterization vector of the target firefighting monitoring data set to obtain the embedded mapping characterization vector of the target firefighting monitoring data set.

In the embodiment of the application, after the data representation vector of the target firefighting monitoring data set is obtained, the data representation vector of the target firefighting monitoring data set is embedded and mapped to finish encoding. For example, the embedded mapping is performed on the data characterization vector of the target firefighting monitoring data set through an Encoder in the firefighting early warning network, so as to obtain the embedded mapping characterization vector (high-dimensional characterization vector) of the target firefighting monitoring data set. The structure of the Encoder may be general and may include a convolution layer, a mapping integration layer (multi-head attention), a multi-layer perceptron (FFN), etc., and after obtaining the data characterization vector of the target fire monitoring dataset, the data characterization vector of the target fire monitoring dataset is loaded into the mapping integration layer to obtain the characterization vector V _n And then characterize vector V _n Loading the vector into a convolution layer to perform convolution calculation to obtain a characterization vector V _c The characterization vector V will then _c The embedded mapping characterization vector V of the target firefighting monitoring data set is obtained by inputting a multi-layer sensor _f 。

Step S140: analyzing the embedded mapping characterization vector of the target firefighting monitoring data set and the environment mapping characterization vector of the target firefighting monitoring data set through the firefighting early warning network to obtain an analysis result of the target firefighting monitoring data set, wherein the analysis result characterizes an early warning tag sequence corresponding to the target firefighting monitoring data set.

In order to rapidly determine the early warning information in the fire-fighting internet of things data, the extracted fire-fighting monitoring data early warning points need to be subjected to label mapping to obtain an early warning label sequence, the distribution positions of the early warning labels contained in the early warning label sequence and the data items corresponding to the fire-fighting monitoring data set are corresponding, so that the early warning data items can be locked in time, it can be understood that the early warning labels in the early warning label sequence can comprise various types and categories, such as distribution according to early warning levels, different early warning levels correspond to different label contents, such as I, II and III correspond to no early warning, general early warning, serious early warning and the like, and the early warning label sequence is beneficial to rapid and accurate positioning of the early warning information of the internet of things cloud platform and timely and accurate early warning. In the embodiment of the application, the fire-fighting early-warning network is a neural network obtained by taking a tag information analysis result of an acquired target fire-fighting monitoring data set as a debugging target and taking a data characterization vector of a current fire-fighting monitoring data set template and an early-warning tag sequence characterization vector corresponding to the current fire-fighting monitoring data set template as debugging samples. That is, the debugging sample of the fire protection early warning network needs to focus the data characterization vector of the current fire protection monitoring data set template and the early warning label sequence characterization vector of the corresponding label information of the current fire protection monitoring data set template. And taking a tag information analysis result of the current fire control monitoring data set template as a debugging target, taking a data characterization vector of the current fire control monitoring data set template and an early warning tag sequence characterization vector corresponding to the current fire control monitoring data set template as debugging samples, debugging the debugging samples, and generating and debugging to obtain the fire control early warning network. The step S120 is referred to as a data characterization vector of the current fire protection monitoring dataset template and an early warning tag sequence characterization vector corresponding to the current fire protection monitoring dataset template.

When the serious early warning label is contained in the early warning label sequence, the system can give an alarm based on a wireless audible and visual alarm in a fire control monitoring Internet of things system, can control the relay to prevent fire and water, and can realize automatic water spraying fire extinguishing, and the functions of monitoring environmental information, fire control early warning and emergency treatment are realized by combining the data acquisition terminal and the control unit.

After the embedded mapping characterization vector and the environment mapping characterization vector of the target firefighting monitoring data set are obtained, the embedded mapping characterization vector and the environment mapping characterization vector of the target firefighting monitoring data set can be determined to be loading data of the firefighting early-warning network so as to analyze based on a Decoder in the firefighting early-warning network, and an analysis result of the target firefighting monitoring data set is obtained.

As an implementation manner, analyzing, by a fire protection early warning network, an embedded mapping characterization vector of a target fire protection monitoring data set and an environment mapping characterization vector of the target fire protection monitoring data set to obtain an analysis result of the target fire protection monitoring data set, including: loading the embedded mapping characterization vector of the target firefighting monitoring data set and the environment mapping characterization vector of the target firefighting monitoring data set into a firefighting early warning network to obtain the confidence level of the reasoning type of each data item in the target firefighting monitoring data set; and determining the analysis result of the target firefighting monitoring data set through the reasoning type confidence of each data item. In the embodiment of the application, because the target fire-fighting monitoring data set is one fire-fighting monitoring data set in the fire-fighting monitoring data set sequence, the embedded mapping characterization vector of the target fire-fighting monitoring data set and the environment mapping characterization vector of the target fire-fighting monitoring data set are loaded to the fire-fighting early warning network, so that the reasoning type confidence of each data item in the target fire-fighting monitoring data set can be obtained. And then identifying the analysis result of the target firefighting monitoring data set in the reasoning type confidence of all the data items. For example, the highest confidence level of the inference type is selected from the confidence levels of the inference types of all the data items, and the early warning label corresponding to the highest confidence level of the inference type is used as the analysis result.

Referring to fig. 4, a schematic diagram of a functional module architecture of a fire-fighting data processing device 110 according to an embodiment of the present invention is provided, where the fire-fighting data processing device 110 may be used to execute a fire-fighting data processing method based on big data, and the fire-fighting data processing device 110 includes:

a vector extraction module 111, configured to obtain a data representation vector of a target fire protection monitoring dataset and a data representation vector of a pre-fire protection monitoring dataset, where the pre-fire protection monitoring dataset is one or more fire protection monitoring datasets located in front of the target fire protection monitoring dataset;

the environment mapping module 112 is configured to determine a data representation vector of the target firefighting monitoring data set and a data representation vector of the early firefighting monitoring data set as loading data of an early warning information mining network, so as to obtain an environment mapping representation vector of the target firefighting monitoring data set;

the embedding mapping module 113 is configured to perform embedding mapping on the data characterization vector of the target firefighting monitoring data set to obtain an embedding mapping characterization vector of the target firefighting monitoring data set;

the vector analysis module 114 is configured to analyze, through a fire protection early warning network, the embedded mapping characterization vector of the target fire protection monitoring data set and the environment mapping characterization vector of the target fire protection monitoring data set, to obtain an analysis result of the target fire protection monitoring data set, where the analysis result characterizes an early warning tag sequence corresponding to the target fire protection monitoring data set.

Since in the above embodiments, the fire fighting data processing method based on big data provided in the embodiments of the present invention has been described in detail, and the principle of the fire fighting data processing device 110 is the same as that of the method, the execution principle of each module of the fire fighting data processing device 110 will not be described in detail here.

In the several embodiments provided in this application, it should be understood that the disclosed apparatus and method may be implemented in other manners as well. The apparatus embodiments described above are merely illustrative, for example, of the flowcharts and block diagrams in the figures that illustrate the architecture, functionality, and operation of possible implementations of apparatus, methods and computer program products according to various embodiments of the present invention. In this regard, each block in the flowchart or block diagrams may represent a module, segment, or portion of code, which comprises one or more executable instructions for implementing the specified logical function(s). It should also be noted that in some alternative implementations, the functions noted in the block may occur out of the order noted in the figures. For example, two blocks shown in succession may, in fact, be executed substantially concurrently, or the blocks may sometimes be executed in the reverse order, depending upon the functionality involved. It will also be noted that each block of the block diagrams and/or flowchart illustration, and combinations of blocks in the block diagrams and/or flowchart illustration, can be implemented by special purpose hardware-based systems which perform the specified functions or acts, or combinations of special purpose hardware and computer instructions.

In addition, functional modules in the embodiments of the present invention may be integrated together to form a single part, or each module may exist alone, or two or more modules may be integrated to form a single part.

The functions, if implemented in the form of software functional modules and sold or used as a stand-alone product, may be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention may be embodied essentially or in a part contributing to the prior art or in a part of the technical solution, in the form of a software product stored in a storage medium, comprising several instructions for causing a computer device (which may be a personal computer, an internet of things data server, or a network device, etc.) to perform all or part of the steps of the method according to the embodiments of the present invention. And the aforementioned storage medium includes: a U-disk, a removable hard disk, a Read-Only Memory (ROM), a random access Memory (RAM, random Access Memory), a magnetic disk, or an optical disk, or other various media capable of storing program codes.

It is noted that relational terms such as first and second, and the like are used solely to distinguish one entity or action from another entity or action without necessarily requiring or implying any actual such relationship or order between such entities or actions. Moreover, the terms "comprises," "comprising," or any other variation thereof, are intended to cover a non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements does not include only those elements but may include other elements not expressly listed or inherent to such process, method, article, or apparatus. Without further limitation, an element defined by the phrase "comprising one … …" does not exclude the presence of other like elements in a process, method, article, or apparatus that comprises the element.

It is to be understood that the terminology which does not make a noun interpretation with respect to the above description is not to be interpreted as a noun interpretation, and that the skilled person can unambiguously ascertain the meaning to which it refers from the above disclosure. The foregoing of the disclosure of the embodiments of the present application will be apparent to and complete with respect to those skilled in the art. It should be appreciated that the process of deriving and analyzing technical terms not explained based on the above disclosure by those skilled in the art is based on what is described in the present application, and thus the above is not an inventive judgment of the overall scheme.

While the basic concepts have been described above, it will be apparent to those skilled in the art that the foregoing detailed disclosure is by way of example only and is not intended to be limiting. Although not explicitly described herein, various modifications, improvements, and adaptations may occur to one skilled in the art. Such modifications, improvements, and modifications are intended to be suggested within this application, and are therefore within the spirit and scope of the exemplary embodiments of this application.

It should also be appreciated that in the foregoing description of the embodiments of the present application, various features are sometimes grouped together in a single embodiment, figure, or description thereof for the purpose of streamlining the disclosure aiding in the understanding of at least one of the embodiments of the invention. This method of disclosure, however, is not intended to imply that more features than are presented in the claims are required for the subject application. Indeed, less than all of the features of a single embodiment disclosed above.

Claims

1. The fire control data processing method based on big data is characterized by being applied to an Internet of things cloud platform, and comprises the following steps:

Determining the data representation vector of the target fire control monitoring data set and the data representation vector of the early-stage fire control monitoring data set as loading data of an early-warning information mining network to obtain an environment mapping representation vector of the target fire control monitoring data set, wherein the environment mapping representation vector represents the context characteristic information of a current fire control monitoring data set template;

analyzing the embedded mapping characterization vector of the target firefighting monitoring data set and the environment mapping characterization vector of the target firefighting monitoring data set through a firefighting early warning network to obtain an analysis result of the target firefighting monitoring data set, wherein the analysis result characterizes an early warning tag sequence corresponding to the target firefighting monitoring data set;

the determining the data characterization vector of the target fire control monitoring data set and the data characterization vector of the early fire control monitoring data set as loading data of an early warning information mining network to obtain an environment mapping characterization vector of the target fire control monitoring data set comprises the following steps:

the early warning information representation vector of the target firefighting monitoring data set and the early warning information representation vector of the early firefighting monitoring data set are interacted to obtain an environment mapping representation vector of the target firefighting monitoring data set;

analyzing the embedded mapping characterization vector of the target firefighting monitoring data set and the environment mapping characterization vector of the target firefighting monitoring data set through the firefighting early warning network to obtain an analysis result of the target firefighting monitoring data set, wherein the analysis result comprises the following steps:

2. The method of claim 1, wherein determining the data characterization vector of the target firefighting monitoring data set as the loading data of the early warning information mining network, to obtain the early warning information characterization vector of the target firefighting monitoring data set, comprises:

3. The method of claim 1, wherein the obtaining a data characterization vector for a target firefighting monitoring data set comprises:

acquiring the target firefighting monitoring data set;

the method further comprises the steps of:

4. A method according to claim 3, characterized in that the method further comprises:

the method further comprises the steps of:

5. The method of claim 4, wherein the obtaining the data cost by an error between the pre-warning token vector corresponding to the current fire monitoring dataset template and the pre-warning token vector corresponding to the first fire monitoring dataset template token vector comprises:

6. A method according to claim 3, characterized in that the method further comprises:

7. The method of claim 6, wherein the obtaining the early warning tag sequence cost by an error between an early warning token vector of the early warning tag sequence token vector corresponding to the current fire monitoring dataset template and an early warning token vector of the first early warning tag sequence token vector comprises:

8. The fire control internet of things system is characterized by comprising an internet of things sensor, a communication device and an internet of things cloud platform, wherein the internet of things sensor is in communication connection with the internet of things cloud platform through the communication device, the internet of things cloud platform comprises a memory and a processor, the memory stores a computer program, and when the processor runs the computer program, the method of any one of claims 1-7 is achieved.