CN-122024435-A - Building fire control diagnostic system based on thing networking and artificial intelligence

CN122024435ACN 122024435 ACN122024435 ACN 122024435ACN-122024435-A

Abstract

The invention relates to the field of electric digital data processing, in particular to a building fire-fighting diagnosis system based on the Internet of things and artificial intelligence, which comprises a second-level link electric data processing module, a second-level communication electric data processing module and a second-level communication electric data processing module, wherein the second-level link electric data processing module is used for generating instant Internet of things fire-fighting diagnosis availability; the system comprises a minute-level link electric data processing module, an hour-level link electric data processing module and a fire control communication protocol prediction module, wherein the minute-level link electric data processing module generates session communication electric data into minute Internet of things fire control diagnosis utilization degree, the hour-level link electric data processing module generates electric data accumulation value into hour Internet of things fire control diagnosis utilization degree, and the fire control communication protocol prediction module is used for training a fire control communication protocol prediction model constructed based on a CNN-LSTM framework based on the instant Internet of things fire control diagnosis utilization degree, the minute Internet of things fire control diagnosis utilization degree and the hour Internet of things fire control diagnosis utilization degree. The method and the system realize accurate control of the communication quality of the gateway communication protocol of the Internet of things, and improve the prediction precision and the prediction generalization capability of the communication protocol.

Inventors

GAO QI
TANG KUI
CAO BAOZHU
MOU WEIPENG
WANG CHUNLI

Assignees

潍坊学院

Dates

Publication Date: 20260512
Application Date: 20260415

Claims (10)

1. Building fire control diagnostic system based on thing networking and artificial intelligence, its characterized in that includes: The second-level link electric data processing module is used for enabling the acquired second-level communication electric data of the gateway of the building fire-fighting Internet of things to pass through a second-level fire-fighting diagnosis utility evaluation model to generate the instant fire-fighting diagnosis utility of the Internet of things, wherein the second-level communication electric data comprises an effective signal-to-noise ratio and an error rate, and the second-level fire-fighting diagnosis utility evaluation model comprises a signal-to-noise ratio instant success probability fitting formula; The system comprises a minute-level link electric data processing module, a minute-level fire control diagnosis utility evaluation module and a network management module, wherein the minute-level link electric data processing module is used for enabling the acquired session communication electric data of the gateway of the building fire control Internet of things to pass through a minute-level fire control diagnosis utility evaluation model so as to generate the fire control diagnosis utility of the minute Internet of things, and the session communication electric data comprises a signal-to-interference-and-noise ratio, a packet loss recovery rate and a protocol keep-alive message sending interval; The system comprises an hour-level link electric data processing module, a fire-fighting diagnosis utility evaluation module and a fire-fighting data processing module, wherein the hour-level link electric data processing module is used for passing an acquired electric data accumulation value of a gateway of the building fire-fighting Internet of things through the hour-level fire-fighting diagnosis utility evaluation model so as to generate an hour-level Internet of things fire-fighting diagnosis utility, and the electric data accumulation value comprises uncontrolled routing times, data packet integrity rate and cross-modal data association gain; The fire control communication protocol prediction module is used for constructing a sample set for optimizing and training a fire control communication protocol prediction model based on the instant Internet of things fire control diagnosis utilization degree, the minute Internet of things fire control diagnosis utilization degree and the hour Internet of things fire control diagnosis utilization degree, and using the trained fire control communication protocol prediction model for generating a predicted optimal communication protocol, wherein the fire control communication protocol prediction model is constructed based on a CNN-LSTM architecture.
2. The building fire diagnosis system based on the internet of things and artificial intelligence according to claim 1, wherein the second-level link electric data processing module comprises: the fire control diagnosis basic value calculation unit is used for calculating the fire control diagnosis basic value based on the transmission data type and the fire condition change rate of the building fire control Internet of things gateway; The signal-to-noise ratio instantaneous success probability fitting calculation unit is used for calculating the instantaneous success probability through a signal-to-noise ratio instantaneous success probability fitting formula based on the effective signal-to-noise ratio; The link quality short-term consistency calculating unit is used for calculating the link quality short-term consistency based on the error rate and the error rate threshold; And the fusion calculation unit is used for calculating the fire control diagnosis effectiveness degree of the instant Internet of things based on the product of the fire control diagnosis basic value, the instant success probability and the link quality short-term consistency degree.
3. The building fire diagnosis system based on the internet of things and artificial intelligence according to claim 2, wherein the signal-to-noise ratio instantaneous success probability fitting calculation unit comprises: The fitting formula generation subunit is used for acquiring the data packet receiving success rate of the building fire-fighting Internet of things gateway through a spectrum analyzer, inputting the effective signal-to-noise ratio and the data packet receiving success rate into the signal-to-noise ratio instantaneous success probability fitting formula to fit and determine the slope factor and the target threshold of the signal-to-noise ratio instantaneous success probability fitting formula, wherein the signal-to-noise ratio instantaneous success probability fitting formula is constructed based on a Sigmoid function format; And the fitting formula calculating subunit is used for calculating the instantaneous success probability by passing the effective signal-to-noise ratio through the fitting formula of the signal-to-noise ratio instantaneous success probability after fitting.
4. The building fire diagnosis system based on the internet of things and artificial intelligence according to claim 1, wherein the minute-level link electric data processing module comprises: The anti-interference degree calculation unit is used for calculating the anti-interference degree based on the ratio of the signal-to-interference-plus-noise ratio to the indication function value of the interference threshold value to the burst time length; the packet loss recovery efficiency calculation unit is used for calculating the packet loss recovery efficiency based on the ratio of the packet loss recovery rate to the total time consumption of the packet loss recovery; a protocol maintenance cost calculation unit, configured to calculate a protocol maintenance cost based on an index value of a ratio of the protocol keep-alive message transmission interval and a basic no-interaction timeout reference threshold; And the weighted fusion calculation unit is used for calculating the fire control diagnosis effectiveness degree of the minute Internet of things based on weighted summation calculation of the anti-interference degree, the packet loss recovery efficiency and the protocol maintenance cost.
5. The building fire diagnosis system based on the internet of things and artificial intelligence according to claim 1, wherein the hour-level link electrical data processing module comprises: a long-term security calculation unit for calculating a long-term security based on the number of uncontrolled routes and an index value of a set time window multiplied by a protocol security value; A data asset accumulation value calculation unit configured to calculate a data asset accumulation value based on a product of the packet integrity rate and a cross-modal data association gain; and the long-term weighted fusion calculation unit is used for carrying out weighted calculation on the long-term safety and the data asset accumulation value so as to generate the fire control diagnosis effectiveness degree of the Internet of things.
6. The building fire diagnosis system based on the internet of things and artificial intelligence according to claim 1, wherein the fire control communication protocol prediction module comprises: The score calculating unit is used for calculating an instant score, a session score and a long-term score based on the instant fire-fighting diagnosis availability of the Internet of things, the fire-fighting diagnosis availability of the Internet of things in the protocol routing network and the current value and the extremely poor ratio of the fire-fighting diagnosis availability of the Internet of things in the hour; A sample set optimal protocol calculation unit to determine an optimal protocol for a sample set based on pareto dominance of the instant score, session score, and long-term score; and the loss function calculation unit is used for optimally training the fire control communication protocol prediction model by combining loss functions based on the optimal protocol.
7. The building fire diagnosis system based on the internet of things and artificial intelligence according to claim 6, wherein the loss function calculation unit comprises: A log-likelihood term calculation subunit, configured to calculate a log-likelihood term based on the optimal protocol, second-level communication electric data, session communication electric data, and electric data accumulation value; A KL divergence calculation subunit that calculates, based on the predicted optimal communication protocol and the KL divergence of the optimal protocol, to construct a KL divergence term; A combined loss function calculation subunit, configured to perform weighted summation calculation on the log likelihood term and the KL divergence term to construct the combined loss function; the sample set further comprises second-level communication electric data, session communication electric data and electric data accumulation value.
8. The building fire diagnosis system based on internet of things and artificial intelligence according to any one of claims 1 to 7, wherein the fire control communication protocol prediction module further comprises: The convolution network feature extraction unit is used for enabling the second-level communication electric data, the session communication electric data and the electric data accumulation value to pass through a convolution network so as to generate a fusion time scale feature; the LSTM time sequence modeling unit is used for enabling the fused time scale features to pass through an LSTM network so as to generate time sequence enhancement features; the decision mapping unit is used for enabling the time sequence enhancement characteristic to pass through an output layer so as to generate the predicted optimal communication protocol; the fire control communication protocol prediction model comprises a convolution network, an LSTM network and an output layer.
9. The building fire diagnosis system based on the internet of things and artificial intelligence according to claim 8, wherein the convolutional network feature extraction unit comprises: a multi-scale convolution feature extraction subunit configured to pass the input tensor through the multi-scale convolution layer to generate a multi-time scale extraction feature; the characteristic splicing subunit is used for carrying out characteristic splicing on the multi-time-scale extracted characteristics so as to generate time-scale spliced characteristics; a channel attention subunit, configured to pass the time scale stitching feature through a channel attention mechanism to generate the fused time scale feature; Wherein the input tensor comprises second-level communication electric data, session communication electric data and electric data accumulation value, and the convolution network comprises a multi-scale convolution layer and a channel attention mechanism.
10. The building fire diagnosis system based on the internet of things and artificial intelligence of claim 8, wherein the LSTM timing modeling unit comprises: An LSTM timing feature extraction subunit for passing the fused time scale features through an LSTM layer to generate timing features; A timing attention mechanism subunit configured to pass the timing feature through a timing attention mechanism to generate the timing enhancement feature; Wherein the LSTM network comprises an LSTM layer and a time sequence attention mechanism.

Description

Building fire control diagnostic system based on thing networking and artificial intelligence Technical Field The invention relates to the field of electric digital data processing, in particular to a building fire-fighting diagnosis system based on the Internet of things and artificial intelligence. Background Along with acceleration of urban process and complicating building functions, the demands of intelligent and accurate building fire safety are increasingly highlighted, the deep fusion of an internet of things (IoT) technology and an Artificial Intelligence (AI) technology provides core support for upgrading building fire diagnosis, and the intelligent fire diagnosis system has the main stream development direction of the building fire diagnosis system by acquiring electric digital data such as the running state of fire equipment, communication link parameters and the like through an internet of things terminal and combining an artificial intelligent model to realize data analysis and intelligent decision. The application of the Internet of things technology in the building fire diagnosis system is characterized in that the interconnection and intercommunication of the fire control gateway, the sensor and the controller are realized through low-power consumption wide area communication protocols of LoRa and NB-IoT, and a fire control data acquisition network covering the building universe is constructed. However, because the building fire scene has the characteristics of complex environment, distributed equipment, dynamic change of communication requirements and the like, the fire-fighting internet of things gateway is used as a core hub for data transmission, and the communication stability of the fire-fighting internet of things gateway directly determines the real-time performance and reliability of fire-fighting data, so that the data processing is required for the electric digital electric data in the link communication states of packet loss rate, delay, integrity, signal strength, gateway connection equipment quantity, data throughput and the like of fire-fighting communication, and the optimal communication protocol in the current building fire-fighting fire state is further determined, and the timeliness of fire-fighting diagnosis results is further ensured. And further, the problem that key data such as smoke concentration, temperature and the like in the initial stage of a fire disaster cannot be transmitted in time due to communication packet loss or delay is avoided, so that alarm is delayed, and the optimal treatment time is missed. At present, the prior art mostly adopts a fixed communication protocol to deploy a fire gateway of the Internet of things, or adjusts the communication protocol through simple parameter threshold judgment, and the fire gateway does not combine the characteristics of a fire scene of a building. On the one hand, the wireless communication environment difference of different building areas of high-rise buildings and underground garages is obvious, so that signal shielding and interference intensity of different areas are different, a fixed communication protocol cannot adapt to diversified communication scenes, and dynamic balance of packet loss rate, delay and communication integrity is difficult to realize. On the other hand, the existing scheme lacks systematic processing of the electric data of the communication link of the fire prevention, only evaluates the communication quality through a single parameter or a simple algorithm, does not construct an accurate utility evaluation model, and cannot generate an optimal communication protocol label adapting to a real-time scene, so that blindness of protocol selection is caused. In summary, how to perform evaluation processing of adapting to building fire-fighting diagnosis communication utility on electric digital data of internet of things communication, so as to determine an optimal communication protocol of building fire-fighting diagnosis, so as to improve communication stability of internet of things building fire-fighting diagnosis is a technical problem to be solved at present. Disclosure of Invention Therefore, the building fire control diagnosis system based on the Internet of things and the artificial intelligence provided by the invention realizes accurate control of communication quality from instant response to long-term steady state of the gateway communication protocol of the Internet of things through the link electric data evaluation processing model of second level, minute level and hour level, avoids building fire control diagnosis scene adaptation deviation caused by a single time scale, combines CNN-LSTM, realizes the improvement of prediction precision and prediction generalization capability of the communication protocol, and further improves the communication stability of building fire control diagnosis of the Internet of things by determining the optimal communication protocol of buildi