Intelligent fire exit door management method and system based on Internet of Things

Abstract

The invention relates to the technical field of fire control management, and discloses an intelligent fire control door management method and system based on the Internet of Things, and the method comprises the following steps: S1, collecting multi-environment data of a fire control door; s2, performing standardization processing to obtain multi-environment data in a unified format, and establishing a feature matrix; s3, determining the importance degree of each item of data in the feature matrix to obtain a comprehensive weight; s4, constructing a membership function, and outputting a fire risk level; s5, determining the fire stage of the current environment; s6, selecting a final risk response strategy through a fuzzy game method; and S7, controlling the action of a fire exit door according to the selected final risk response strategy. According to the method, the information entropy and the analytic hierarchy process algorithm are combined to determine the importance degree of the multi-environment data and calculate the comprehensive weight of the fire risk, so that the assessment is more objective and comprehensive, overlarge deviation of a single model is avoided, and the problems of bias and poor applicability of assessment are solved.

Description

Technical Field

[0001] The present invention relates to the technical field of fire protection management, and in particular to an intelligent fire protection door management method and system based on the Internet of Things. Background Art

[0002] In the safety protection system of modern buildings, fire doors, as important passive fire protection components, play a key role in blocking the spread of fire, slowing down the spread of fire, and ensuring smooth evacuation of personnel. Especially in large shopping malls, hospitals, rail transit, storage facilities and other scenes, fire doors often work together with fire alarm systems, emergency broadcast systems, etc. to provide multiple safety barriers in sudden fires. With the gradual popularization of the Internet of Things and sensing technology, fire doors have been given more and more intelligent functions, realizing real-time perception and dynamic response to environmental conditions, and becoming an important part of the intelligent fire protection system.

[0003] In the existing technology, a relatively extensive technical path is still used. A common practice is to set a fixed threshold: an alarm is triggered when the smoke exceeds a certain concentration, and the door is controlled to open when the temperature is higher than a certain value. Therefore, the judgment logic of a single factor is too idealistic and cannot cover the diverse paths of fire development in reality. In many cases, the carbon monoxide concentration has just increased, but other parameters are still within the normal range. The traditional system cannot perceive the potential risks at all, which will lead to problems such as excessive deviation, biased evaluation, and poor applicability. Summary of the invention

[0004] In view of the shortcomings of the prior art, the present invention provides an intelligent fire door management method and system based on the Internet of Things, which solves the problem in the prior art that the judgment logic of the fire door based on a single factor is too idealistic, which in turn leads to excessive deviation, biased evaluation and poor applicability.

[0005] To achieve the above objectives, the present invention is implemented through the following technical solutions: an intelligent fire door management method based on the Internet of Things, comprising the following steps: Step S1, collecting multi-environment data of fire doors; Step S2, standardizing the multi-environment data to obtain multi-environment data in a unified format and establishing a feature matrix; Step S3, respectively using an information entropy algorithm and a hierarchical analysis algorithm to determine the importance of each data item in the feature matrix, and calculating the comprehensive weight based on the calculation results of the information entropy algorithm and the hierarchical analysis algorithm; Step S4, constructing a membership function based on the unified format multi-environment data and the comprehensive weight as input, wherein the membership function is used to evaluate the current environmental state and output a fire risk level; Step S5: Based on the evaluated fire risk level, use the Bayesian inference method to determine the fire stage of the current environment; Step S6: Based on the fire stage of the current environment, select the final risk response strategy through the fuzzy game method, send an alarm message, and upload the alarm message to the cloud at the same time; Step S7: Control the action of the fire door according to the selected final risk response strategy, and record the fire door data and upload it to the cloud.

[0006] Preferably, in the step S1, the multi-environment data includes the temperature of the fire door, the smoke concentration around the fire door, the carbon monoxide concentration around the fire door, the vibration of the fire door body, and the rate of change of the surface temperature of the fire door; The multi-environment data is obtained through a variety of sensors installed in the fire door area, and the sensors include temperature sensors, smoke sensors, carbon monoxide sensors, vibration sensors, and infrared thermal imaging sensors.

[0007] Preferably, in the step S2, the normalization process adopts the interval scaling method, linearly transforms each item of environmental data according to the corresponding historical maximum and minimum values, so that the multi-environment data values in the unified format are all between 0 and 1.

[0008] Preferably, in the step S3: The information entropy algorithm is used to calculate the objective weights of each item of environmental data; The analytic hierarchy process algorithm calculates the subjective weights through the judgment matrix; The objective weight and the subjective weight are weighted and synthesized according to a preset ratio to synthesize the comprehensive weight for evaluating the fire state.

[0009] Preferably, in the step S4, the fire risk levels include: safe state, early warning state, initial fire state, and developing fire state.

[0010] Preferably, during the evaluation process of the fire risk level, a membership function is constructed based on each item of standardized multi-environment data, and the membership function is used to represent the membership degree of the multi-environment data to different fire risk levels; according to the membership degrees of the multi-environment data in each fire risk level, combined with the comprehensive weight, the fuzzy comprehensive evaluation method is used to calculate the scoring values of each fire risk level, and finally determine the fire risk level corresponding to the current environmental state.

[0011] Preferably, in the step S5, the Bayesian inference method calculates the posterior probability of the fire state based on the fire risk level output by the evaluation model for the current environmental state and the prior probability of the historical statistical fire state, and the state with the largest posterior probability is the fire stage of the current environment.

[0012] Preferably, in step S6, the risk response strategy includes opening the fire door, keeping it closed, and activating the alarm device; The fuzzy game method takes the fire stage as the input, establishes the risk assessment value of the risk response strategy, and selects the strategy with the minimum risk value as the final response strategy according to the fire stage in the current environment.

[0013] Preferably, in step S7, the fire door data includes: multi-environment data of the fire door, fire risk level, fire stage, selected response strategy, and the actions of the fire door.

[0014] An intelligent fire door management system based on the Internet of Things, comprising: A multi-environment data acquisition module for acquiring various environment data of the fire door and its surroundings, including temperature, smoke concentration, carbon monoxide concentration, door body vibration, and surface temperature change rate; A data processing module for standardizing the multi-environment data, generating processed data in a unified format, and establishing a feature matrix; A weight calculation module for determining the importance of each item of environment data in the multi-environment data and calculating the comprehensive weight respectively by using the information entropy algorithm and the analytic hierarchy process algorithm based on the feature matrix; A state evaluation module for constructing an evaluation model based on the multi-environment data in the unified format and the comprehensive weight, evaluating the current environment state, and outputting the fire risk level; A fire stage inference module for determining the fire stage in the current environment by using the Bayesian inference method based on the fire risk level; A response strategy decision module for selecting the final risk response strategy by using the fuzzy game method according to the current fire stage and generating the corresponding warning information; A control execution module for controlling the opening or closing of the fire door according to the final risk response strategy, sending out the warning information, and recording the fire door data; A communication module for uploading the warning information and the recorded fire door data to the cloud server.

[0015] The present invention provides an intelligent fire door management method and system based on the Internet of Things. It has the following beneficial effects: 1. The present invention determines the importance of multi-environment data through the information entropy algorithm and the analytic hierarchy process algorithm, and finally calculates the comprehensive weight of the fire risk assessment, making the evaluation result more objective and comprehensive, and avoiding too large deviation of the evaluation result under a single model. Compared with the methods mostly using fixed thresholds or single-factor judgments in the prior art, the accuracy and reliability of risk identification are significantly improved, and the problems of biased judgment results and insufficient practicality in the past are solved.

[0016] 2. The present invention constructs a feature matrix of multi-environment data and uses fuzzy comprehensive evaluation and Bayesian inference to jointly evaluate the fire level and its development stage, enabling dynamic perception in the actual environment, actively identifying state evolution, and avoiding the risks of response lag and misjudgment.

[0017] 3. The present invention provides a decision-making basis for the response strategy of the fire door by adopting the fuzzy game theory, and dynamically selects the optimal action in combination with the real-time fire stage, enabling the device to flexibly adjust the opening and closing states of the fire door, solving the problems of rigid response strategies and inability to cope with changing fire situations in the prior art, and significantly enhancing the intelligent level. BRIEF DESCRIPTION OF THE DRAWINGS

[0018] Figure 1 is a schematic diagram of the method flow of the present invention; Figure 2 is a schematic diagram of the system architecture of the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0019] The following will clearly and completely describe the technical solutions in the embodiments of the present invention with reference to the accompanying drawings in the specification of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all of the embodiments. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of the present invention without creative efforts shall fall within the protection scope of the present invention.

[0020] For a better understanding of the present invention, the above content will be described in detail below with specific embodiments.

[0021] Please refer to the attached Figure 1 , the embodiment of the present invention provides an Internet of Things-based intelligent fire door management method, including the following steps: Step S1, collecting multi-environment data of the fire door; In this embodiment, step S1 is used to complete the multi-dimensional perception of the state of the fire door and its surrounding environment, and is the starting link in the Internet of Things-based intelligent fire door management method. This step provides key input data for subsequent processes such as data standardization, fire risk assessment, and strategy response, and its integrity and accuracy directly affect the overall performance of the system.

[0022] For this reason, in step S1, the multi-environment data collected by the present invention includes but is not limited to the following types: The surface temperature of the fire door body; The smoke concentration in the air around the fire door; The concentration of carbon monoxide (CO) in the air around the fire door; The degree of vibration received by the fire door body; The rate of change of the surface temperature of the fire door over time.

[0023] In a specific embodiment, the above environmental parameters are all collected in real time by a plurality of sensors arranged on the fire door body and its adjacent areas. The types of sensors include but are not limited to: Temperature sensor: used to sense the real-time temperature of the door surface or the surrounding environment; Smoke sensor: used to detect the concentration of smoke particles in the air, reflecting possible combustion behavior; Carbon monoxide sensor: used to monitor the change of CO concentration in the air to identify potential harmful gas accumulation; Vibration sensor: used to detect whether the door body vibrates, combined with external impacts or abnormal structural changes; Infrared thermal imaging sensor: used to capture the thermal distribution image of the door body and further used to calculate the temperature change rate.

[0024] Generally, the sensors should be arranged to cover different heights of the door body and multiple areas of the adjacent wall to improve the integrity of the data and the robustness of the system. For example, a set of multi-functional sensor units can be arranged at the upper, middle, and bottom of the door body respectively to ensure coverage in dimensions such as thermal distribution, gas concentration, and mechanical disturbance.

[0025] In some embodiments, the data collection period is set to not exceed 1 second to meet the rapid response requirements in the early stage of a fire. As an option, this collection period can also be dynamically adjusted according to real-time data fluctuations, system load status, or set security policies. For example, when an abnormal temperature rise trend is identified, the sampling interval is automatically shortened to 0.2 seconds to enhance the response sensitivity.

[0026] To reduce the influence of noise interference or data anomalies on system judgment, a filtering algorithm or preprocessing mechanism is also introduced in the collection process of the present invention. For example: The vibration data is processed by a moving average filter after collection to suppress occasional high-frequency jitters; Before the infrared thermal imaging data is used to calculate the temperature change rate, grayscale image enhancement and edge sharpening are first performed to improve the time series contrast.

[0027] In a specific implementation, the temperature change rate of the fire door surface can be calculated using the following formula: ; Where represents the temperature change rate per unit time (unit: °C / s); is the door surface temperature obtained by the infrared thermal imaging sensor at the current time point; is the temperature value recorded in the previous sampling period; , are the time stamps at the current and previous moments respectively, with the unit of seconds.

[0028] This rate of temperature change is an important dynamic characteristic index, especially suitable for identifying "rapid temperature rise" scenarios, which are commonly seen in sudden fire scenarios such as electrical short circuits and oil combustion; in contrast, "slow temperature rise" may occur in heat source penetration or occlusion environments. Through this index, it can effectively assist in distinguishing the fire risk level and judging the evolution path.

[0029] In some embodiments, the outputs of the above-mentioned multiple sensors can be transmitted to the local edge computing unit through a standard communication interface (such as RS485, I2C or CAN bus) for preliminary data integration and caching, thereby reducing network dependence and enhancing the distributed processing ability of the system.

[0030] In addition, in an extended implementation, the system can also set an event-driven sampling mechanism. When the value collected by a certain sensor exceeds a preset threshold (such as the smoke concentration exceeding a certain limit), it will trigger the entire acquisition subsystem to enter the "high-frequency mode" to monitor parameter changes with a higher time resolution and improve the early warning accuracy for emergencies.

[0031] In summary, step S1 realizes high-frequency acquisition for multi-source environmental elements. Through the collaborative work of heterogeneous sensor fusion deployment, dynamic sampling control, redundant layout and pre-processing algorithms, it greatly improves the recognition sensitivity and real-time performance of the system for early signs of fire, laying a solid perception foundation for subsequent intelligent judgment and response.

[0032] Step S2: Standardize the multi-environment data to obtain multi-environment data in a unified format and establish a feature matrix; In this embodiment, as an important processing link immediately following the multi-environment data acquisition (step S1), the main purpose of step S2 is to perform standardization conversion on the original environmental data from different sources and with different dimensions, so as to facilitate subsequent unified analysis and calculation.

[0033] Specifically, since environmental parameters such as temperature, smoke concentration, carbon monoxide concentration, vibration intensity and rate of temperature change have different physical dimensions and numerical ranges themselves, directly using the original data will lead to inconsistent feature scales, thus affecting the accuracy of subsequent feature weight calculation and risk modeling. Therefore, in step S2 of the present invention, interval scaling processing is performed on all the collected environmental data, so that each feature value is mapped to the same standardized interval, that is, within the range of [0,1], to form a multi-environment data set in a unified format.

[0034] Specifically, the interval scaling standardization method adopts the following transformation formula: ; Among them, represents the value of the th environmental data after standardization, with a range of [0, 1]; represents the original environmental data value actually collected; represents the th historical minimum value of the environmental data, based on long-term observation data or system initialization settings; represents the th historical maximum value of the environmental data, usually a preset limit value or empirical statistical value.

[0035] Generally, the maximum and minimum values of all features are statistically obtained from the historical operation dataset and stored in the system database; in the case of the first deployment of the system or lack of historical data, the limit values can also be preset by expert experience and used as the initial interval.

[0036] In a specific implementation, the standardization process can dynamically adjust the maximum and minimum values based on a sliding time window to improve the adaptability to environmental changes. For example, the system can set a 24-hour sliding window period to update the historical extreme values of each feature value to reflect the short-term change trend.

[0037] As an option, the present invention also supports multi-gear management of the maximum and minimum values based on the context scenario. For example, for temperature data, there are significant differences in the extreme value ranges in the operating environments of winter and summer, and the system can automatically switch the interval settings according to the current season label or external meteorological interface to improve the standardization accuracy.

[0038] After the standardization process is completed, the system synchronizes and integrates each item of data according to the time stamp to form a multi-environment data vector in a unified format. The data at multiple time points can be further combined in sequence to construct a feature matrix in the following form: ; Among them, represents the standardized feature matrix, with a dimension of ; represents the number of samples in the time series; represents the number of dimensions of the environmental features (in this embodiment, the number is 5, corresponding to temperature, smoke concentration, CO concentration, vibration, temperature change rate); represents the standardized value of the th time point and the th item of feature ( , ).

[0039] In some embodiments, the feature matrix serves as the input data structure for the subsequent weight calculation module and can also be used for the training or online inference of the fire recognition model. In this way, data from different sensor sources is unified and structured, greatly improving the fusion efficiency of cross-modal features.

[0040] To ensure data stability and model convergence, some abnormal or outlier data points can be removed or smoothed and interpolated before entering the standardization module. For example, when the instantaneous jump amplitude of a certain feature value exceeds ±3 times the standard deviation of the historical mean, the data filtering logic can be triggered to automatically postpone its inclusion in the standardization process.

[0041] In a possible system deployment method, the standardization module is deployed on the edge node processing platform to ensure low-latency response and local processing capabilities, especially suitable for scenarios with dense sensor deployment or edge environments with unstable network connections.

[0042] In summary, step S2 not only completes the unification of multi-type environmental features in scale and the structuring in dimension, but also ensures the stability of data input and the compatibility of the subsequent processing link through flexible interval scaling strategies, multi-source historical data fusion, dynamic window mechanisms, and abnormal data processing logics, further guaranteeing the practicality and reliability of the fire recognition system.

[0043] Step S3: Use the information entropy algorithm and the analytic hierarchy process algorithm to determine the importance of each item of data in the feature matrix, and calculate the comprehensive weight by calculating the calculation results of the information entropy algorithm and the analytic hierarchy process algorithm; In this embodiment, step S3 follows immediately after the generation of the multi-environment feature data matrix after standardization (step S2). Its core purpose is to establish a multi-index weight evaluation mechanism that combines data-driven and expert knowledge, so as to provide a comprehensive weight coefficient with discrimination ability for subsequent fire status recognition.

[0044] Specifically, since the distinguishability of different environmental features in different fire scenarios varies, relying solely on objective data statistics is not sufficient to fully reflect their importance in actual judgment. Therefore, the present invention combines the information entropy algorithm and the analytic hierarchy process algorithm (AHP) to calculate the objective weight and the subjective weight respectively, and further forms a unified comprehensive weight vector through weighted fusion as one of the inputs for the subsequent risk recognition model.

[0045] Specifically, the information entropy algorithm is used to calculate the objective weights of each environmental feature in the feature matrix, and the process is as follows: Based on the aforementioned obtained feature matrix It can be seen that , which represents the standardized value of the th feature at the th moment.

[0046] Based on this, first, calculate the probability distribution value of the th feature, and the calculation formula is: ; where represents the number of samples in the time series; represents the number of dimensions of the environmental features (in this embodiment, the number is 5, corresponding to temperature, smoke concentration, CO concentration, vibration, and temperature change rate); Then, calculate the information entropy of the th feature. The specific calculation formula is: ; where is the normalization coefficient to ensure that the information entropy value ; When , define .

[0047] Furthermore, obtain the entropy weight of the th feature. The specific calculation formula is: ; where represents the objective weight value of the th environmental feature; Therefore, the above entropy weight reflects the degree of data variation. The greater the variation, the higher the contribution to the discrimination of the system state, and thus the greater its objective weight.

[0048] In a specific implementation, the analytic hierarchy process is used to calculate the subjective weights of each feature item, which mainly includes the following steps: ; where represents the subjective weight value of the th environmental feature; represents the maximum eigenvector of the th environmental feature in the feature matrix; represents the sum of each maximum eigenvector in the feature matrix; In a possible implementation, the subjective and objective weights are weighted and fused as follows to obtain the final comprehensive weight: ; Among them, represents the comprehensive weight of the th environmental feature; is the proportion of the objective weight, is the proportion of the subjective weight, and it satisfies .

[0049] In summary, in step S3, by introducing two weight evaluation strategies of information entropy and analytic hierarchy process, the importance degree of features is quantified from two aspects of data statistical distribution and expert experience respectively, and a comprehensive weight system for fire status evaluation is established through weighted fusion. While maintaining the sensitivity of the model, this step also improves the rationality and adaptability of weight configuration, providing a scientific basis for subsequent state modeling, alarm threshold judgment and dynamic response strategy.

[0050] Step S4: Construct a membership function with the multi-environment data in a unified format and the comprehensive weight as inputs. The membership function is used to evaluate the current environmental state and output the fire risk level; In this embodiment, step S4 is immediately followed by the generation of the comprehensive weight vector calculated based on information entropy and analytic hierarchy process (step S3). Using this comprehensive weight, fuzzy evaluation is performed on the multi-environment data after standardization, and a fuzzy recognition model is constructed to realize the intelligent discrimination of the fire risk level under the current environmental state.

[0051] Specifically, the core of this step is to realize the quantitative evaluation of different risk states (such as safe, warning, initial fire, developing fire) by establishing a fuzzy mapping relationship between multi-environment features and fire risk levels.

[0052] To achieve this function, the present invention introduces a fuzzy membership function model, constructs the membership degree expression form of each type of environmental feature under each fire level respectively, and finally combines the comprehensive weight to obtain the fire risk level scoring result of the current state through the fuzzy comprehensive evaluation algorithm.

[0053] Specifically, the construction method of the membership function is as follows: For the standardized environmental feature value , its membership degree under each fire risk level is expressed as , where represents the fire risk level, and the value is , corresponding to "safe state", "warning state", "initial fire state", "developing fire state" respectively; represents the membership degree that this feature value belongs to level .

[0054] In a specific implementation, the membership function can be modeled using triangular, trapezoidal, or Gaussian functions to improve the system's adaptability to uncertain features. For example, when using a triangular membership function, its general expression is: ; Wherein, represents the left endpoint of the membership function of the th feature at level ; represents the position where the membership degree is 1, corresponding to the maximum membership point; represents the right endpoint; The function graph is a triangle with a peak at .

[0055] And the above parameters, , can be set according to historical monitoring samples and domain knowledge, or obtained through sample fitting algorithms.

[0056] In the fuzzy comprehensive evaluation stage, the following calculation process is adopted in this embodiment: The comprehensive membership degree score value at fire level is calculated as follows: ; Wherein, is the comprehensive weight of the th environmental feature calculated in step S3; is the membership degree of the current feature value at level ; is the fuzzy score value of the current input at the th fire level.

[0057] Finally, the score values of all levels are combined into a fuzzy output vector: ; In a possible implementation, a threshold enhancement mechanism can be introduced to suppress the sensitivity to misjudgment.

[0058] That is, when exceeds the threshold (for example, the threshold is set to 0.5), a flag value of "status uncertain" or "delayed determination required" can be output to improve the robustness of the discrimination system.

[0059] Generally, there should be an overlapping area between the membership functions at each level of the fuzzy comprehensive evaluation to ensure the recognizability of the state in the fuzzy boundary area. For example, a certain overlapping area can be set between the "early warning state" and the "initial fire state" for transitional recognition.

[0060] In some embodiments, according to the fire response requirements for different application scenarios, the present invention can construct multiple sets of membership function templates.

[0061] For example, for residential environments and industrial storage environments, there are significant differences in the distribution of environmental characteristics and the characteristics of fire evolution. The system can automatically switch the membership function configuration according to the equipment deployment scenario to improve adaptability.

[0062] As an option, an adaptive membership function adjustment strategy can also be adopted. Based on continuously collected data, methods such as cluster analysis are used to regularly optimize the membership function parameters to maintain the synchronization of the discrimination ability and environmental evolution.

[0063] To achieve a clear output of the fire level classification, the present invention defines the fire risk level including the following four categories: Safe state (Level-0): indicating that the current environment is in a normal and safe operating state; Early warning state (Level-1): some characteristic values exceed the normal range but do not pose a fire threat yet; Initial fire state (Level-2): there are clear abnormal signs and early fire characteristics; Developing fire state (Level-3): multiple characteristic values deviate significantly from the normal level and have obvious fire signs.

[0064] The system outputs the current fire state level according to the level corresponding to the highest score in the comprehensive membership score vector corresponding to the input data. The result can be used to drive subsequent alarm responses, fire linkage, or control strategies.

[0065] In summary, step S4 realizes the effective fusion and accurate recognition of multi-source data in different fire level discrimination tasks by constructing a membership mapping between multi-environmental characteristics and fire levels, combining the comprehensive weight vector, and introducing the fuzzy comprehensive evaluation method. This mechanism has good robustness in dealing with environmental state uncertainties, provides a clear and progressive risk level output strategy for the system, and significantly enhances the engineering adaptability and practical performance of the intelligent fire recognition system.

[0066] In step S5, based on the evaluated fire risk level, the Bayesian inference method is used to determine the fire stage in which the current environment is located; In this embodiment, based on the fire risk level output by the Bayesian inference method-based evaluation model for the current environmental state evaluation and the prior probability of the fire state statistically obtained in history, the posterior probability of the fire state is calculated, and the state with the maximum posterior probability is the fire stage in which the current environment is located.

[0067] Generally, a single fire risk level output is difficult to reflect the continuity and diversity of fire evolution. Therefore, in the present invention, a Bayesian probability model is used to establish a mapping inference framework of "risk level - fire stage", enabling the system to further identify the stage it is in based on the existing risk level information, such as the stage evolution process from "ignition source generation" to "spread and development" and then to "fully burning".

[0068] Specifically, in the Bayesian model, the required parameters include the following two categories: Prior probability : It represents the occurrence probability of each fire stage in historical statistics and can be obtained by frequency statistics of the stage labels in the historical monitoring dataset; Likelihood function : It represents the probability that the system determines the current fire risk level to be under the fire stage and can be constructed based on the co-occurrence frequency or empirical rules in the training samples. For example, the probability of the fire risk level occurring in step 4 can be statistically calculated.

[0069] ; Among them, is the posterior probability, represents the total number of fire stages divided by the system, such as: no fire, initial fire, continuous combustion, spread and diffusion, etc. Each stage corresponds to a ; The final output stage determination result is as follows: ; Among them, represents the fire stage most likely corresponding currently, which is the stage with the maximum posterior probability value; The result of this stage will be used as the key input for subsequent linkage modules (such as alarm trigger, fire response decision-making).

[0070] As an option, the present invention can further optimize the acquisition methods of the prior probability and the likelihood function.

[0071] In some embodiments, the system can update the prior probability regularly to reflect the dynamic changes in the fire evolution trend under different scenarios or seasons.

[0072] In addition, the likelihood function It can also be established based on Bayesian network structure learning or conditional probability table estimation methods, and the inference accuracy can be improved through a continuous learning mechanism. For example, in an industrial environment, the grade distribution corresponding to stage 3 may be biased towards the "initial fire state", while in a residential environment, it may be biased towards the "warning state", so a sub-scenario modeling mechanism needs to be introduced.

[0073] In some embodiments, to enhance the model's interpretability and robustness, the present invention supports outputting a complete posterior probability vector : ; This posterior probability distribution can be further used by the upper-layer policy module. For example: If the highest probability is only slightly higher than the second-highest value, the "transition period" state can be output; Or a risk interval can be set to control the uncertainty of the posterior distribution fluctuation.

[0074] Specifically, combining the risk level output in the aforementioned step S4, this step takes this value as an observation variable and inputs it into the Bayesian inference module, and links it with the built-in fire stage probability model in the system to realize the construction of the inference link from the level to the stage.

[0075] For example: If an input is judged as the "initial fire state" (level 2) after step S4, and historical statistics show that the probability of this level in stage 3 (continuous fire state) is the highest, then the system outputs that it is currently in stage 3.

[0076] This inference process realizes the enhancement of the mapping from the risk level to the fire stage while maintaining the low complexity of the input data, forming a complete and hierarchical fire judgment system.

[0077] In summary, step S5 constructs a risk-level-driven fire stage probability inversion mechanism by introducing a Bayesian inference model. On the basis of making full use of historical statistical knowledge and current state information, this step realizes the dynamic estimation of the fire state evolution path, provides scientific support for the accurate matching of subsequent response strategies, and improves the overall intelligent level and discrimination credibility of the system.

[0078] In step S6, based on the fire stage in the current environment, the final risk response strategy is selected through a fuzzy game method, and an alarm message is sent, and at the same time, the alarm message is uploaded to the cloud; In this embodiment, step S6 follows immediately after the determination result of the fire stage in step S5. Based on the fire development stage of the current environment, a fuzzy game analysis mechanism is further introduced to select the optimal response path from multiple possible risk response strategies and generate and upload the alarm information in real time. This step aims to achieve strategy adaptation and risk avoidance in different environmental scenarios while ensuring response timeliness, based on the response priority and risk-return relationship corresponding to different fire stages. It forms a closed-loop intelligent response system with the aforementioned fuzzy comprehensive evaluation (step S4) and Bayesian inference stage recognition (step S5), jointly constructing an intelligent fire response logic link of "level recognition - stage judgment - strategy linkage".

[0079] Generally, the judgment result of the fire stage cannot be directly mapped to a single response operation. Especially in the situation where multiple strategies coexist and their effects are mutually exclusive, a single decision criterion is difficult to effectively weigh the response priority in a complex scenario. Therefore, in the present invention, a fuzzy game model is further constructed, taking the current fire stage as the decision input variable, combining the risk assessment function of different strategies at the current stage, and selecting the response strategy based on the minimum risk principle.

[0080] Specifically, the response strategy selection method driven by fuzzy game mainly includes the following contents: Define the fire response strategy set as: ; Where: represents the th optional response strategy, ; In a possible implementation, the response strategies include: : Keep the current state (do not react); : Activate the alarm device; : Automatically open the fire door or switch to the fire extinguishing mode; Define the fire stage set as: ; And the stage of the current environment is determined by step S5 as; And on this basis, introduce the fuzzy risk assessment function , indicating the risk intensity brought by executing the strategy at the fire stage .

[0081] At this time, the core formula for fuzzy game strategy selection is: ; Where, Is the finally selected response strategy; For the current stage The risk value corresponding to the execution of the strategy .

[0082] Regarding the risk assessment function The construction method is as follows: In some embodiments, the risk function can be constructed by a fuzzy rule inference system, and the following fuzzy rules are set: If in the "initial fire state", the risk value of "alarm" is low, the risk of "opening the fire door" is medium, and the risk value of "maintaining the current" is high; If in the "high-risk spread state", the risk value of "maintaining the current" is extremely high, and the risks of "alarm" and "opening the fire door" are relatively low; If in the "fire-free state", the risk values of all strategies are relatively high, but the "maintaining the current" is the lowest.

[0083] The triangular membership function or Gaussian membership function can be used to describe the fuzzy degree of the strategy risk intensity, construct a fuzzy rule base, perform fuzzy reasoning to obtain the fuzzy risk value of each strategy, and then perform defuzzification through the centroid method or the maximum membership degree method to obtain the specific risk quantification value.

[0084] Specifically, in a possible implementation, the fuzzy game system includes the following elements: Input item: The currently determined fire stage ; Rule base: The set of strategy risk fuzzy rules for each stage; Inference engine: Adopt the fuzzy composition method to perform risk assessment between strategies; Output item: The selected optimal response strategy .

[0085] For example: If the current determination is the stage (continuous fire state), the fuzzy rule system evaluates and obtains: ; ; ; Then the final response strategy at this time is: ; That is, execute the "activate the alarm device" strategy.

[0086] As an option, the execution result of the response strategy will synchronously generate an alarm message and upload it to the cloud management platform.

[0087] In some embodiments, the alarm information includes, but is not limited to: Current fire stage ; Response policy number ; Auxiliary information such as system timestamp, monitoring location coordinates, and fire level etc.; The alarm information is organized in a structured data format (such as JSON or XML) and uploaded to the cloud platform through a standard communication protocol (such as MQTT or HTTPS) for remote response deployment or platform-level data archiving.

[0088] In some embodiments, to improve the flexibility and scalability of response policy selection, the system supports dynamically adjusting the policy set and the parameter set of the risk assessment function .

[0089] For example: For large industrial sites, policy items can be added, such as "trigger sprinkler system" and "isolate power module"; For residential buildings, the risk function structure can be dynamically adjusted according to scenarios such as night / day or population density to make the response policy more adaptable to the scenario.

[0090] In summary, based on the determination of the fire stage in step S6, the fuzzy game method is introduced, and by constructing a multi-strategy response risk model, the selection of the response policy driven by the minimum risk criterion is achieved. This mechanism takes into account both response efficiency and risk control, ensures the adoption of the most appropriate response actions at different fire stages, and at the same time realizes the systematic upload of alarm information, providing a timely and accurate basis for platform-level response deployment, and significantly enhancing the intelligent linkage control ability of the system.

[0091] Step S7, control the action of the fire door according to the selected final risk response policy, and record the fire door data and upload it to the cloud.

[0092] In this embodiment, step S7 follows immediately after the risk response policy selection in the previous step S6, and further executes the response policy through an intelligent control system to achieve the automated operation of the fire door, and record and upload the corresponding data to the cloud. The core of this step is to accurately control the open / closed state of the fire door according to the selected final risk response policy, and ensure that each operation record and environmental data are timely and accurately uploaded to the cloud platform for data archiving and subsequent remote monitoring.

[0093] In general, the response strategy for fire scenarios should not only consider the real-time development of the fire but also take into account the flexibility and stability of the system. Especially in high-risk environments, the opening and closing of fire doors must be precisely controlled according to the requirements of specific fire stages and response strategies. Therefore, through the interpretation of the selected risk response strategy in step S6 of the present invention , combined with the fire door control system, an accurate execution and real-time feedback mechanism are realized to ensure that each response behavior can meet the predetermined safety standards and system requirements.

[0094] Specifically, the execution of the fire door action needs to be linked with the built-in sensors and execution modules in the system to ensure the correct action of the fire door.

[0095] In some embodiments, the system can also dynamically adjust the action of the fire door according to the specific fire scenario and its changes. For example, during the spread of the fire, certain areas may automatically close the fire door to prevent the spread of the fire, while in other stages, the fire door may be automatically opened to remove smoke or gas. Through precise control, each operation can respond according to the needs of the real-time environment, improving the timeliness and safety of the response.

[0096] Regarding the technical implementation of fire door data recording and uploading: In this embodiment, each action of the fire door will generate corresponding data records, including but not limited to: Operation timestamp; Current fire stage; Operation type (such as opening or closing); Current environmental fire rating; Operation execution status (whether it is successfully executed), etc.

[0097] These data will be organized in a structured data format (such as JSON or XML) through a secure encryption method and uploaded to the cloud management platform. The specific data uploading process includes: Pack the data into a standard format; Use a standard communication protocol (such as MQTT, HTTPS, etc.) to send it to the cloud platform; Store and monitor the data in the cloud to provide basic data for subsequent remote response and decision-making.

[0098] Specifically, in a possible implementation, the specific technical solution of the data uploading process is as follows: Data collection: The fire door control system records the details of each operation, such as the timestamp, current fire stage, operation type, etc.; Data formatting: Convert all data into a unified format (such as JSON) to ensure the consistency and parsability of the data during the uploading process; Data encryption: Encrypt the data to ensure the privacy and security of the data; Data upload: Through communication protocols supported by the cloud platform (such as MQTT, HTTPS, etc.), upload the data to the remote monitoring system in real time; Cloud platform processing: The cloud platform stores and analyzes the uploaded data, and supports functions such as real-time display, trend prediction, and alarm linkage.

[0099] As an option, when uploading data, the system can synchronously transmit the real-time fire stage and response strategy information, which is convenient for the cloud platform to perform big data analysis and intelligent decision-making.

[0100] For example, when a certain fire door is opened, the cloud platform can receive the status information of the door opening and the current fire stage (such as "high-risk spread state") at the same time. Through historical data analysis, the platform can dynamically optimize and adjust the response strategy to ensure that the system is more efficient and intelligent in the next response.

[0101] In some embodiments, to improve the adaptability of the system, the system can adjust the fire door control strategy according to the changes in the fire stage.

[0102] For example, when it is determined that the fire is in the "continuous fire state", the system may automatically close some fire doors to prevent the spread of the fire, but in the "initial fire state", it may open the fire doors to exhaust smoke in time.

[0103] Through real-time monitoring and dynamic adjustment, ensure that the system can respond flexibly according to the specific changes in the fire scene.

[0104] In summary, step S7 introduces a fire door control and data upload mechanism, which, while ensuring the accuracy of the fire door actions, records and uploads relevant data to the cloud in a timely manner. This mechanism not only improves the efficiency of the response system, but also provides valuable data support for subsequent decision-making analysis, further enhancing the intelligence level and adaptability of the system.

[0105] Although the embodiments of the present invention have been shown and described, it will be understood by those of ordinary skill in the art that various changes, modifications, substitutions and variations can be made in these embodiments without departing from the principles and spirit of the present invention, and the scope of the present invention is defined by the appended claims and their equivalents.

Claims

1. An intelligent fire door management method based on the Internet of Things, characterized in that, Including the following steps: Step S1, collecting multi-environment data of the fire door; Step S2, performing standardization processing on the multi-environment data to obtain multi-environment data in a unified format, and establishing a feature matrix; Step S3, respectively using the information entropy algorithm and the analytic hierarchy process algorithm to determine the importance degree of each item of data in the feature matrix, and calculating the comprehensive weight by calculating the calculation results of the information entropy algorithm and the analytic hierarchy process algorithm; Step S4, constructing a membership function with the multi-environment data in the unified format and the comprehensive weight as inputs, where the membership function is used to evaluate the current environmental state and output the fire risk level; Step S5, based on the evaluated fire risk level, using the Bayesian inference method to determine the fire stage in which the current environment is located; Step S6, based on the fire stage in which the current environment is located, selecting the final risk response strategy through the fuzzy game method, sending out an alarm message, and uploading the alarm message to the cloud at the same time; Step S7, controlling the action of the fire door according to the selected final risk response strategy, and recording the fire door data and uploading it to the cloud.

2. The intelligent fire door management method based on the Internet of Things according to claim 1, wherein, In the step S1, the multi-environment data includes the temperature of the fire door, the smoke concentration around the fire door, the carbon monoxide concentration around the fire door, the vibration of the fire door body, and the surface temperature change rate of the fire door; The multi-environment data is obtained by various sensors installed in the fire door area, and the sensors include a temperature sensor, a smoke sensor, a carbon monoxide sensor, a vibration sensor, and an infrared thermal imaging sensor.

3. The intelligent fire door management method based on the Internet of Things according to claim 1, wherein, In the step S2, the standardization processing adopts the interval scaling method, linearly transforming each item of environmental data according to the corresponding historical maximum value and minimum value, so that the multi-environment data values in the unified format are all between 0 and 1.

4. The intelligent fire door management method based on the Internet of Things according to claim 1, wherein, In the step S3: The information entropy algorithm is used to calculate the objective weight of each item of environmental data; The analytic hierarchy process algorithm calculates the subjective weight through the judgment matrix; The objective weight and the subjective weight are weighted and synthesized according to a preset ratio to synthesize the comprehensive weight for evaluating the fire state.

5. The intelligent fire door management method based on the Internet of Things according to claim 1, characterized in that, In the step S4, the fire risk levels include: safe state, early warning state, initial fire state, and developing fire state.

6. The intelligent fire door management method based on the Internet of Things according to claim 5, characterized in that, In the evaluation process of the fire risk level, a membership function is constructed based on each item of standardized multi-environment data, and the membership function is used to represent the membership degree of the multi-environment data to different fire risk levels; According to the membership degrees of the multi-environment data in each fire risk level, combined with the comprehensive weight, the fuzzy comprehensive evaluation method is used to calculate the scoring values of each fire risk level, and finally the fire risk level corresponding to the current environmental state is determined.

7. The intelligent fire door management method based on the Internet of Things according to claim 6, characterized in that, In the step S5, the Bayesian inference method calculates the posterior probability of the fire state based on the fire risk level output after evaluating the current environmental state by the evaluation model and the prior probability of the historical statistical fire state, and the state with the largest posterior probability is the fire stage in which the current environment is located.

8. The intelligent fire door management method based on the Internet of Things according to claim 1, characterized in that, In the step S6, the risk response strategies include opening the fire door, keeping it closed, and activating the alarm device; The fuzzy game method takes the fire stage as the input, establishes the risk assessment value of the risk response strategy, and selects the strategy with the minimum risk value as the final response strategy according to the fire stage in the current environment.

9. The intelligent fire door management method based on the Internet of Things according to claim 1, wherein In step S7, the fire door data includes: multi-environment data of the fire door, fire risk level, fire stage, selected response strategy, and actions of the fire door.

10. An intelligent fire door management system based on the Internet of Things, based on the intelligent fire door management method based on the Internet of Things according to any one of claims 1-9, characterized in that, It includes: A multi-environment data acquisition module for collecting various environmental data around the fire door and itself, including temperature, smoke concentration, carbon monoxide concentration, door body vibration, and surface temperature change rate; A data processing module for normalizing the multi-environment data, generating processed data in a unified format, and establishing a feature matrix; A weight calculation module for determining the importance of each environmental data item in the multi-environment data and calculating the comprehensive weight respectively by using the information entropy algorithm and the analytic hierarchy process algorithm based on the feature matrix; A state evaluation module for constructing an evaluation model based on the multi-environment data in the unified format and the comprehensive weight, evaluating the current environmental state, and outputting the fire risk level; A fire stage inference module for determining the fire stage in the current environment by using the Bayesian inference method based on the fire risk level; A response strategy decision module for selecting the final risk response strategy by using the fuzzy game method according to the current fire stage and generating corresponding warning information; A control execution module for controlling the opening or closing of the fire door according to the final risk response strategy, sending out warning information, and recording the fire door data; A communication module for uploading the warning information and the recorded fire door data to the cloud server.

Images