Fire risk collaborative assessment and early warning method and system

Abstract

This invention discloses a method and system for collaborative fire risk assessment and early warning, relating to the technical field of fire early warning. The method includes: extracting risk factors such as heat accumulation, combustible material release, electrical anomalies, and ventilation diffusion from multi-source sensor data to construct a risk factor vector; analyzing the amplification effect and coupling relationship between factors to establish a risk coupling model; constructing a risk propagation path network containing propagation direction and impedance by combining spatial structure and connectivity; tracking the collaborative evolution of risk factors over time and outputting evolutionary representations; and accumulating and calculating the degree of risk accumulation along the paths to achieve collaborative fire risk assessment and early warning. This invention solves the technical problems of missing coupling relationships among multiple risk factors, disconnect between risk propagation paths and spatial structure, and lack of time-series collaborative evolution tracking in existing technologies. It achieves the technical effects of improving the coupling representation capability of fire risk assessment, the accuracy of path-level risk accumulation calculation, and dynamic collaborative early warning.

Description

Technical Field

[0001] This invention relates to the field of fire early warning technology, specifically to a method and system for collaborative assessment and early warning of fire risks. Background Technology

[0002] Traditional fire detection and early warning systems, such as smoke and heat detectors, are "post-event" or "critical point" alarms. They ignore the collaborative evolution characteristics of multi-physical field information during the incubation, occurrence, and development of fires, making it difficult to achieve accurate early warning in the early stages of a fire, and even more difficult to assess the spread and accumulation trend of fire risks in complex spatial structures. In the field of fire protection, multi-source sensing data (such as temperature, smoke, gas concentration, electrical parameters, airflow velocity, etc.) are introduced for comprehensive monitoring. However, there are many problems. Heat release, combustible material volatilization, electrical faults, and ventilation conditions are usually regarded as independent risk indicators, failing to reveal the nonlinear interactions between these factors, resulting in a serious underestimation of risk. Using global or regional average risk indices ignores the spatial structure and connectivity within the target area, making it difficult to locate the critical paths and nodes of risk accumulation. Instantaneous assessment based on static cross-sectional data fails to track the temporal changes of each risk factor along the propagation path, and cannot analyze the primary and secondary transformations and synergistic enhancement relationships of multiple factors at different stages, resulting in early warning results lagging behind the actual evolution of the risk. In addition, the lack of a quantitative description of the degree of risk accumulation along each propagation path makes it difficult to take targeted smoke extraction, isolation, or evacuation measures.

[0003] Therefore, current technologies suffer from several technical problems, including the lack of coupling relationships among multiple risk factors, the disconnect between risk propagation paths and spatial structures, and the lack of temporal co-evolution tracking. Summary of the Invention

[0004] This application provides a fire risk collaborative assessment and early warning method and system, which solves the technical problems in the prior art such as the lack of coupling relationship of multiple risk factors, the disconnect between risk propagation path and spatial structure, and the lack of time-series collaborative evolution tracking. It achieves the technical effects of improving the coupling representation capability of fire risk assessment, the accuracy of path-level risk accumulation calculation, and dynamic collaborative early warning.

[0005] This application provides a method for collaborative assessment and early warning of fire risks. The method includes: acquiring multi-source sensing data monitored by sensors within a target area, and extracting basic risk factors characterizing heat accumulation processes, combustible material release processes, electrical anomaly processes, and ventilation diffusion capabilities based on the multi-source sensing data, and constructing a risk factor vector; based on the risk factor vector, analyzing the interaction relationships between various risk factors, identifying risk factor combinations with amplification effects, and constructing a risk coupling relationship model reflecting the coupling strength of risk factors; under the constraints of the risk coupling relationship model, and combining the spatial structure and connectivity of the target area, constructing a risk propagation path network including risk propagation direction and propagation impedance; based on the risk propagation path network, performing time-series tracking of the changes of the risk factors on the paths, analyzing the collaborative evolution state of the risk factors, and outputting collaborative evolution characterization results; based on the collaborative evolution characterization results, performing path-level cumulative calculation of risks on the risk propagation path network to obtain the degree of risk accumulation on each propagation path, and establishing a fire risk early warning system.

[0006] In possible implementations, the co-evolution state of risk factors is analyzed, and the co-evolution characterization results are output. This includes: based on the interaction mechanism between different risk factors in the risk coupling relationship model, the risk factors are divided into heat-driven factors, combustible material supply factors, diffusion modulation factors, and triggering excitation factors; and a path evolution sequence reflecting the interaction order of different types of risk factors is established on each path of the risk propagation path network; for each risk propagation path, a segment-level co-action unit is constructed based on the spatial connectivity and environmental attributes between adjacent nodes on the path; within each co-action unit, the release-promoting effect of heat-driven factors on combustible material supply factors and diffusion modulation factors are analyzed. The amplification or inhibition effects of risk factors on heat and combustible gases, and the instantaneous excitation effects of triggering factors, are used to construct a local co-evolution rule for multi-factor coupling. The risk factor response is decomposed and projected along the corresponding risk propagation path, and each risk factor vector is locally co-processed within each co-action unit based on the local co-evolution rule according to the action order defined by the path evolution sequence. The results of the local co-processing are transmitted step by step in the path propagation direction to identify the continuous amplification chain formed between risk factors by heat accumulation, enhanced combustible release, accelerated diffusion, and triggering excitation, and the co-evolution amplification coefficient is calculated on the continuous amplification chain. The co-evolution characterization result is output using the co-evolution amplification coefficient.

[0007] In a possible implementation, the co-evolutionary characterization result is output using the co-evolutionary amplification coefficient, including: constructing a co-evolutionary state sequence along the risk propagation path based on the local co-processing results of each co-working unit in the continuous amplification chain, and extracting the corresponding co-enhancement rate and change gradient for each node in the co-evolutionary state sequence; determining the effectiveness of the continuous amplification chain based on the continuity and monotonicity of the co-enhancement rate in the path propagation direction, identifying effective co-working segments that meet the continuous enhancement condition, and ineffective segments with enhancement interruption or reverse suppression; calculating the chain-level evolution intensity index for the effective co-working segments based on the co-evolutionary amplification coefficient and cumulative growth trend; and identifying co-enhancement based on the distribution characteristics of the change gradient within the effective co-working segments. The process identifies gradient mutation points and gradient sustained rise intervals, using the gradient mutation points as transition boundaries for co-evolution stages and the gradient sustained rise intervals as risk acceleration evolution intervals. Within these intervals, a co-evolution acceleration factor is calculated by combining the evolution intensity index and the growth rate of the changing gradient. Based on this factor, the chain segment-level evolution intensity index is dynamically corrected to obtain a comprehensive evolution index reflecting the coupling relationship between risk evolution speed and intensity. Based on this comprehensive evolution index and the co-evolution amplification coefficient, the effective co-evolution chain segments are classified into initial coupling state, stable enhancement state, and accelerated instability state. Under the accelerated instability state, the corresponding risk propagation path is determined to have entered the co-evolutionary runaway stage, which is output as the co-evolutionary characterization result.

[0008] In a possible implementation, the fire risk collaborative assessment and early warning method includes: extracting the response lag characteristics and synchronous change characteristics between risk factors in the time dimension based on the historical change trajectory of each risk factor, and using the response lag characteristics and synchronous change characteristics as the coupling judgment criteria to screen risk factor combinations that exhibit a mutually driving relationship under time correlation conditions; for the screened risk factor combinations, constructing coupling triggering rules describing the triggering conditions of the amplification effect between risk factors in combination with the corresponding environmental state parameters, and defining the state boundary of the transition from weak correlation to strong coupling between risk factors based on the coupling triggering rules; under the constraint of the coupling triggering rules, segmenting the response intensity of the risk factor combination in different state intervals, constructing a coupling intensity mapping relationship that dynamically adjusts with changes in environmental state and factors, and establishing the risk coupling relationship model.

[0009] In possible implementations, a fire risk early warning system is established, including: calculating the path-level risk growth rate based on the degree of risk accumulation and the trend of change of each risk propagation path; when the degree of risk accumulation and the risk growth rate simultaneously meet preset conditions, determining that the corresponding path enters the early warning state and triggering a fire risk early warning.

[0010] Among possible implementation methods, the fire risk collaborative assessment and early warning method includes: conducting time-series monitoring of fire risk early warnings, establishing time-series follow-up feedback, and using the time-series follow-up feedback to manage the early warning update of fire risk early warnings.

[0011] Among possible implementation methods, the fire risk collaborative assessment and early warning method includes: when multiple risk propagation paths exist simultaneously, comparing and analyzing the degree of risk accumulation and collaborative evolution status of each risk propagation path, identifying the dominant path with the highest risk evolution speed, and using the influence relationship between the dominant path and the remaining risk propagation paths to execute collaborative early warning marking of potential risk propagation paths.

[0012] This application also provides a fire risk collaborative assessment and early warning system, the system comprising: a vector construction module, used to acquire multi-source sensing data monitored by sensors within a target area, and extract basic risk factors representing heat accumulation processes, combustible material release processes, electrical anomaly processes, and ventilation diffusion capabilities based on the multi-source sensing data, and construct a risk factor vector; a model construction module, used to analyze the interaction relationships between various risk factors based on the risk factor vector, identify risk factor combinations with amplification effects, and construct a risk coupling relationship model reflecting the coupling strength of risk factors; a network construction module, used to construct a risk propagation path network including risk propagation direction and propagation impedance under the constraints of the risk coupling relationship model and in combination with the spatial structure and connectivity of the target area; a result output module, used to perform time-series tracking of the changes of the risk factors on the path based on the risk propagation path network, analyze the collaborative evolution state of the risk factors, and output collaborative evolution characterization results; and a fire risk early warning module, used to perform path-level cumulative calculation of risks on the risk propagation path network based on the collaborative evolution characterization results, obtain the degree of risk accumulation on each propagation path, and establish a fire risk early warning.

[0013] The proposed fire risk collaborative assessment and early warning method and system extracts risk factors such as heat accumulation, combustible material release, electrical anomalies, and ventilation diffusion from multi-source sensor data to construct a risk factor vector; analyzes the amplification effect and coupling relationship between factors to establish a risk coupling model; combines spatial structure and connectivity to construct a risk propagation path network containing propagation direction and impedance; tracks the collaborative evolution of risk factors over time and outputs evolutionary representations; and accumulates and calculates the degree of risk accumulation along the path to achieve collaborative fire risk assessment and early warning. This method solves the technical problems of missing coupling relationships among multiple risk factors, disconnect between risk propagation paths and spatial structure, and lack of time-series collaborative evolution tracking in existing technologies, achieving the technical effects of improving the coupling representation capability of fire risk assessment, the accuracy of path-level risk accumulation calculation, and dynamic collaborative early warning. Attached Figure Description

[0014] To more clearly illustrate the technical solutions of the embodiments of this disclosure, the accompanying drawings of the embodiments of this disclosure will be briefly described below. Flowcharts are used in this application to illustrate the operations performed by the system according to the embodiments of this application. It should be understood that the preceding or following operations are not necessarily performed precisely in sequence. Instead, various steps can be processed in reverse order or simultaneously as needed. Furthermore, other operations can be added to these processes, or one or more steps can be removed from these processes.

[0015] Figure 1 A schematic diagram of the fire risk collaborative assessment and early warning method provided in the embodiments of this application.

[0016] Figure 2 A schematic diagram of the fire risk collaborative assessment and early warning system provided in this application embodiment.

[0017] Figure labeling: Vector construction module 10, Model construction module 20, Network construction module 30, Result output module 40, Fire risk early warning module 50. Detailed Implementation

[0018] To further illustrate the technical means and effects of the present invention in achieving its intended purpose, the following detailed description is provided in conjunction with the accompanying drawings and preferred embodiments, based on the specific implementation methods, structure, features, and effects of the present invention.

[0019] This application provides a method for collaborative fire risk assessment and early warning, such as... Figure 1 As shown, the method includes: Step S100: Obtain multi-source sensing data monitored by sensors within the target area, and extract basic risk factors representing the heat accumulation process, combustible material release process, electrical anomaly process and ventilation diffusion capacity based on the multi-source sensing data, and construct a risk factor vector.

[0020] Preferably, a data acquisition system is used to acquire multi-source sensing data monitored by various sensors deployed within the target area, including but not limited to temperature values ​​measured by temperature sensors, smoke concentration values ​​measured by smoke sensors, current / voltage / leakage current values ​​measured by electrical circuit monitoring devices, specific gas concentration values ​​measured by gas sensors, and airflow velocity and pressure values ​​measured by wind speed / pressure sensors. Then, through calculation or transformation, four specific categories of physical quantities are extracted from the multi-source sensing data. Each category corresponds to a key physical process in the fire development process, including basic risk factors characterizing the heat accumulation process, combustible material release process, electrical anomaly process, and ventilation diffusion capacity. Specifically, the risk factors characterizing the heat accumulation process refer to physical quantities calculated from temperature data and its rate of change, such as ambient temperature value, temperature rise rate, duration of temperature exceeding a set threshold, and estimated heat release rate, directly reflecting the degree of heat accumulation within the target area; the risk factors characterizing the combustible material release process refer to... Physical quantities calculated from smoke concentration and specific gas concentration data, such as smoke concentration value, smoke concentration growth rate, combustible gas concentration value, and volatile organic compound concentration value, directly reflect the release intensity of pyrolysis products of solid materials and vaporized gases from liquids within the target area. Risk factors characterizing electrical anomalies refer to physical quantities calculated from electrical parameter data, such as residual current value of lines, arc discharge pulse count, conductor temperature value, harmonic distortion rate of current / voltage, and load imbalance, which directly reflect whether there are abnormal operating conditions of electrical equipment or lines within the target area that may cause a fire. Risk factors characterizing ventilation diffusion capacity refer to physical quantities calculated from data such as wind speed, wind pressure, and the opening and closing status of doors and windows, such as regional average wind speed, air exchange rate, pressure difference, and flow resistance coefficient on the ventilation path, which directly reflect the ability of airflow within the target area to dilute, transport, or support the combustion of heat and combustible gases. Finally, the four risk factors are arranged in a predefined order to form a structured risk factor vector.

[0021] Step S200: Based on the risk factor vector, analyze the interaction relationship between each risk factor, identify risk factor combinations with amplification effect, and construct a risk coupling relationship model that reflects the coupling strength of risk factors.

[0022] Preferably, the input risk factor vector is used to calculate and analyze multiple sets of risk factor vectors collected within the same time period, quantifying the degree of correlation between any two or more factors. Specifically, this includes calculating the correlation coefficient between factors, determining whether another factor shows a regular unidirectional or inverse change when the value of one factor changes, calculating the time-lag correlation between factors, determining whether a change in one factor causes a change in another factor after a fixed time delay, calculating the ratio of the rate of change between factors, determining the response magnitude of other factors corresponding to a unit change in one factor, and outputting the correlation type between each risk factor, such as positive correlation, negative correlation, causal relationship, correlation strength (weak, medium, strong), and response delay time. Then, risk factor combinations with amplification effects are identified, i.e., when two or more risk factors exist simultaneously and interact, their combined effect is greater than the sum of their individual effects. Based on the analysis of interaction relationships, factor combinations are screened, including those where an increase in risk factor A leads to an increase in risk factor B, and an increase in risk factor B further exacerbates the increase in risk factor A; the coupling effect between risk factors causes the rate of change of a certain risk factor to exceed its maximum rate of change when acting independently; and the temporal superposition sequence of risk factors significantly enhances the harmfulness of later-occurring risk factors under the environmental conditions created by preceding risk factors.

[0023] Preferably, the combination of heat accumulation factor and combustible release factor accelerates the pyrolysis or volatilization of combustibles due to temperature increase, and the released combustible gas may be ignited after accumulation, generating more heat; the combination of electrical anomaly factor and heat accumulation factor directly contributes to heat accumulation due to local high temperature caused by electrical leakage or short circuit; the combination of ventilation diffusion factor and the above combinations reduces the dilution rate of heat and combustible gas when ventilation capacity is weakened, resulting in a greater increase in heat accumulation and combustible concentration than under normal ventilation conditions; finally, a list of factor combinations with amplification effect and the amplification trigger conditions for each combination are output. Risk factor coupling strength measures the degree of interdependence and mutual influence among risk factors. The higher the coupling strength, the more significant the amplification effect between factors. For each identified combination of amplification effects, a mathematical function or rule set is established to express the coupling strength as a function of the current value of the relevant factor. Variables include the current value and rate of change of each factor involved in the coupling, as well as the time delay parameter between factors. The output is a coupling coefficient ranging from 0 to 1, where 0 represents no coupling and 1 represents complete coupling. Finally, a risk coupling relationship model is constructed that can receive the risk factor vector at the current moment as input and output the current coupling coefficient of each factor combination, thereby quantitatively describing the mutual reinforcement or inhibition relationship between different risk factors at that moment.

[0024] Furthermore, step S200 also includes: based on the historical change trajectory of each risk factor, extracting the response lag characteristics and synchronous change characteristics between risk factors in the time dimension, and using the response lag characteristics and synchronous change characteristics as the coupling judgment criteria to screen risk factor combinations that exhibit a mutually driving relationship under time correlation conditions; for the screened risk factor combinations, combining the corresponding environmental state parameters, constructing coupling triggering rules to describe the triggering conditions of the amplification effect between risk factors, and defining the state boundary of the transition from weak correlation to strong coupling between risk factors based on the coupling triggering rules; under the constraint of the coupling triggering rules, segmenting the response intensity of the risk factor combination in different state intervals, constructing a coupling intensity mapping relationship that dynamically adjusts with changes in environmental state and factors, and establishing the risk coupling relationship model.

[0025] Preferably, the input is the historical change trajectory of each risk factor, i.e., the historical monitoring data sequence within a continuous time period, such as the numerical sequence of heat accumulation factor, combustible material release factor, electrical anomaly factor, and ventilation diffusion factor recorded every 5 seconds in the past 30 minutes. The consistency of the change direction of two risk factor numerical sequences at the same time point is calculated, and it is determined whether risk factor N also rises / falls at the same time point when risk factor M rises / falls. The synchronous change rate = the number of time points when risk factors M and N change in the same direction / the total number of time points. The synchronous change characteristic value is output, with a value range of 0 to 1, where 1 indicates complete synchronization. At the same time, the pattern of change of one risk factor after another changes after a certain time delay is calculated. The numerical sequence of factor O is time-shifted, with different delay times such as 1 second, 2 seconds, 5 seconds, or 10 seconds. The correlation coefficient after shifting is calculated, and the shift amount that makes the correlation coefficient reach its maximum value is selected as the response lag time value. Then, the synchronous change characteristic value and response lag time value of each risk factor combination are output.

[0026] Preferably, if the synchronous change characteristic value is greater than a preset threshold (e.g., 0.7), it is determined that the two factors have an immediate mutual driving relationship. If the response lag time is within a preset range (e.g., 1-30 seconds) and the lag correlation coefficient is greater than a preset threshold (e.g., 0.6), it is determined that one factor has a unidirectional driving relationship with the other factor. If there is a significant lag correlation in both directions, it is determined that the two factors have a bidirectional mutual driving relationship. Then, all pairwise and tri-factor combinations of risk factors are traversed, and combinations that meet the judgment conditions are retained as candidate coupling combinations. A list of risk factor combinations with mutual driving relationships is output. Combining the corresponding environmental state parameters, such as environmental temperature, environmental humidity, air pressure, oxygen concentration, etc., the numerical range of each factor and the range of environmental state parameters when the amplification effect occurs in historical data are analyzed. The triggering conditions are then expressed as logical judgment statements to determine the triggering rules. When there are multiple environmental parameters or multiple factor thresholds, the AND and OR logical operators are used to connect multiple conditions, and the set of coupling triggering rules corresponding to each risk factor combination is output.

[0027] Preferably, when the value of the risk factor combination does not satisfy any of the coupling triggering rules, it is determined to be a weak correlation state; when the value of the risk factor combination fully satisfies all the conditions of the coupling triggering rules, it is determined to be a strong coupling state. Then, a set of critical conditions between the weak correlation state and the strong coupling state is defined, including treating each threshold parameter in the triggering rule as a boundary value. For example, the temperature threshold of 60℃ is the state boundary between weak correlation and strong coupling in the temperature dimension. When the factor value is exactly equal to the boundary value, a transition rule can be defined to output the state boundary parameters of each risk factor combination.

[0028] Preferably, the response strength of risk factor combinations in different state intervals is characterized in segments. Specifically, the numerical range of each risk factor is divided into several continuous intervals according to the state boundaries. Within each state interval, a mapping function between coupling strength and factor value is established, such as a linear function, a piecewise constant, or a nonlinear function. Then, the coupling strength is expressed as a function of environmental parameters, so that the coupling strength mapping relationship can be automatically adjusted with changes in environmental state parameters. For example, coupling strength = basic coupling strength × environmental humidity correction coefficient. When environmental state parameters change, the calculated output of coupling strength changes accordingly. The segmented mapping function, state boundaries, and triggering rules of all risk factor combinations are integrated into the final risk coupling relationship model, which can output the real-time coupling strength coefficient of each factor combination with the current risk factor value and environmental state parameters as input.

[0029] Step S300: Under the constraints of the risk coupling relationship model, and in combination with the spatial structure and connectivity of the target area, a risk propagation path network including risk propagation direction and propagation impedance is constructed.

[0030] Preferably, the coupling strength coefficient and triggering conditions output by the risk coupling relationship model are read to determine the constraints of the risk coupling relationship model, including factor combination constraints, coupling strength constraints, and triggering condition constraints. The planar layout, floor distribution, building structure, and the location, shape, and size of each functional area of ​​the target area are obtained. Spatial structural elements are extracted, including at least nodes, regions, and boundaries. The spatial structure of the target area is output, and all nodes and their spatial attributes are labeled. The connection methods and connection attributes between spatial nodes are determined, such as direct connection, indirect connection, and vertical connection. Simultaneously, the size, status, length, and obstacle distribution of connection openings are recorded. The connectivity matrix between spatial nodes is output, indicating whether a connection exists between any two nodes and the specific attributes of the connection. Then, based on the spatial structure and connectivity, a directed graph network describing the propagation of risk between spatial nodes is established, where each network node is a spatial unit and is assigned attributes. A network edge is a directed edge between two connected nodes. The direction of the edge indicates the direction of risk propagation. The risk propagation direction may include from a high-pressure area to a low-pressure area, from a high-concentration area to a low-concentration area, from a high-temperature area to a low-temperature area, or along the measured or simulated airflow direction. When multiple factors conflict, the final direction is determined according to a preset priority (e.g., pressure difference greater than temperature difference greater than concentration gradient). Propagation impedance represents the degree of obstruction encountered by a risk factor when it propagates from one node to an adjacent node. The larger the impedance value, the more difficult it is for the risk to propagate through that path. Propagation impedance = w1 × (1 / opening area) + w2 × path length + w3 × obstacle density + w4 × fire resistance coefficient, where w1, w2, w3, and w4 are preset weight coefficients, and their sum is 1. The propagation impedance value on each directed edge is then output, and the nodes, directed edges, propagation directions on the edges, and propagation impedance values ​​are integrated into a risk propagation path network.

[0031] Step S400: Based on the risk propagation path network, the change process of the risk factors on the path is tracked in time, the co-evolution state of the risk factors is analyzed, and the co-evolution characterization results are output.

[0032] Step S400 further includes, based on the interaction mechanism between different risk factors in the risk coupling relationship model, classifying the risk factors into heat-driven factors, combustible material supply factors, diffusion modulation factors, and triggering excitation factors, and establishing path evolution sequences reflecting the interaction order of different types of risk factors on each path of the risk propagation path network; for each risk propagation path, constructing segment-level synergistic action units based on the spatial connectivity and environmental attributes between adjacent nodes on the path; within each synergistic action unit, based on the release-promoting effect of heat-driven factors on combustible material supply factors, and the release effect of diffusion modulation factors on heat and combustible gases... The system constructs a local co-evolution rule for multi-factor coupling by considering the effects of large or small inhibition, and the instantaneous excitation effect of triggering factors. The risk factor response is decomposed and projected along the corresponding risk propagation path, and each risk factor vector is locally co-processed within each co-action unit based on the local co-evolution rule according to the action order defined by the path evolution sequence. The results of the local co-processing are then transmitted step-by-step along the path propagation direction to identify the continuous amplification chain formed between risk factors by heat accumulation, enhanced release of combustible materials, accelerated diffusion, and triggering excitation. The co-evolution amplification coefficient is calculated on the continuous amplification chain. The co-evolution characterization result is output using the co-evolution amplification coefficient.

[0033] Preferably, the risk factors and their interactions identified in the input risk coupling relationship model are classified into heat-driven factors, combustible material supply factors, diffusion modulation factors, and triggering excitation factors according to their mechanism of action. Specifically, if they affect other factors through heat generation or heat accumulation, they are classified as heat-driven factors, such as temperature values, temperature rise rates, estimated heat release rates, and thermal radiation flux; if they characterize the release, accumulation, or presence of combustible materials, they are classified as combustible material supply factors, such as smoke concentration, combustible gas concentration, volatile organic compound concentration, and solid material pyrolysis product concentration; if they affect the transport speed and distribution of heat, gas, and smoke in space, they are classified as diffusion modulation factors, such as wind speed, wind direction, air exchange rate, ventilation path flow resistance, and pressure difference; if they are instantaneous, sudden, and can cause step changes in other factors, they are classified as triggering excitation factors, such as arc discharge pulse counts, electrical short circuit signals, open flame detection signals, and abnormal voltage spikes; and then, a category label for each risk factor is output. Heat accumulation leads to an increase in the release of combustibles. After release, the distribution of combustibles is affected by diffusion modulation. The order of action of factors is determined as follows: heat-driven factors → combustible supply factors → diffusion modulation factors → triggering factors. For each propagation path in the risk propagation path network, the risk factors involved in all nodes of the path are sorted according to their order of action to generate the path evolution sequence corresponding to each risk propagation path.

[0034] Preferably, the connection between two adjacent nodes in the risk propagation path is called a road segment. The spatial connectivity and environmental attributes between adjacent nodes in each risk propagation path are used as road segment attributes, including the opening size, opening type and opening status between the two nodes, as well as the environmental temperature, environmental humidity, airflow speed and obstacle distribution in the area where the road segment is located. Then, each road segment is defined as an independent cooperative unit, including the starting node, the ending node, the road segment attributes and the current values ​​of all risk factors on the road segment. Then, within each collaborative unit, a multi-factor coupled local collaborative evolution rule is constructed. Specifically, this includes a rule for the promoting effect of heat-driven factors on combustible material supply, in which the increment of the combustible material supply factor is calculated based on the value of the heat-driven factor within the current collaborative unit; a rule for the amplification or inhibition of heat and combustible gas by diffusion modulation, for example, when the wind speed is below the low-speed threshold (e.g., less than 0.3 m / s), the diffusion modulation factor amplifies the heat and combustible gas, causing them to accumulate within the unit; when the wind speed is above the high-speed threshold (e.g., greater than 1.5 m / s), the diffusion modulation factor inhibits the heat and combustible gas, causing them to be diluted and dispersed; and a rule for the instantaneous excitation effect of triggering factors, in which when the value of the triggering factor is greater than the preset trigger threshold, the heat-driven factors and combustible material supply factors within the collaborative unit instantaneously generate a step increment. Finally, the local collaborative evolution rule is output and executed independently by each collaborative unit.

[0035] Preferably, the risk factor response is decomposed and projected along the corresponding risk propagation path, that is, the overall risk factor vector at the current moment is projected into each collaborative unit according to its spatial node position. Each collaborative unit obtains the risk factor sub-vectors at its starting node and ending node. Then, within each collaborative unit, local collaborative processing is performed sequentially according to the category order specified by the path evolution sequence, and the processed risk factor values ​​of each collaborative unit are output. The results of local collaborative processing are then passed step by step along the path propagation direction. This includes using the risk factor value of the termination node after processing by the first collaborative unit as the input of the starting node of the second collaborative unit, and passing it sequentially along the propagation direction until the end node of the path. During the transmission process, the changes in the risk factor values ​​before and after processing by each collaborative unit are monitored. When the output risk factor value of a collaborative unit is greater than the input value and the increment exceeds a preset threshold (e.g., increment > 5%), the unit is marked as an amplification unit. If multiple consecutive units along the propagation direction are amplification units, they are marked as a continuous amplification chain. The complete continuous amplification chain formed by heat accumulation, enhanced release of combustibles, accelerated diffusion, and triggering excitation between risk factors is identified, and the position, length, and list of collaborative units included in the continuous amplification chain are output.

[0036] Preferably, the input and output values ​​of each cooperating unit in the continuous amplification chain are used as inputs to calculate the co-evolution amplification coefficient. This includes calculating the unit amplification coefficient = (unit output risk value - unit input risk value) / unit input risk value. When the input risk value is 0, the unit amplification coefficient = unit output risk value (absolute value). The chain cumulative amplification coefficient = (chain end output risk value - chain start input risk value) / chain start input risk value is calculated. The unit amplification coefficient list, chain cumulative amplification coefficient, and factor category amplification coefficient of each continuous amplification chain are output. Finally, the co-evolution characterization result is output.

[0037] Furthermore, step S400 also includes: constructing a co-evolutionary state sequence along the risk propagation path based on the local co-processing results of each co-working unit in the continuous amplification chain; extracting the corresponding co-enhancement rate and change gradient for each node in the co-evolutionary state sequence; determining the effectiveness of the continuous amplification chain based on the continuity and monotonicity of the co-enhancement rate in the path propagation direction, identifying effective co-working chain segments that meet the continuous enhancement conditions, and ineffective chain segments with enhancement interruption or reverse suppression; calculating the chain segment-level evolution intensity index for the effective co-working chain segments based on the co-evolutionary amplification coefficient and cumulative growth trend; and identifying gradient mutation points in the co-enhancement process based on the distribution characteristics of the change gradient within the effective co-working chain segments. The gradient continuously rises within a certain range, and the gradient mutation point is used as the switching boundary of the co-evolution stage. The continuously rising gradient range is also used as the risk acceleration evolution range. Within the risk acceleration evolution range, the co-evolution acceleration factor is calculated by combining the evolution intensity index and the growth rate of the changing gradient. Based on the co-evolution acceleration factor, the evolution intensity index at the chain segment level is dynamically corrected to obtain a comprehensive evolution index reflecting the coupling relationship between the speed and intensity of risk evolution. Based on the comprehensive evolution index and the co-evolution amplification coefficient, the effective co-evolution chain segment is classified into three states: initial coupling state, stable enhancement state, and accelerated instability state. Under the accelerated instability state, the corresponding risk propagation path is determined to have entered the co-evolution runaway stage, which is output as the co-evolution characterization result.

[0038] Preferably, the local collaborative processing results of each collaborative unit in the continuous amplification chain are input, and each node on the continuous amplification chain is arranged sequentially according to the risk propagation path to form a collaborative evolution state sequence. Each node corresponds to a state record, including the node number, the value of the heat driving factor, the value of the combustible material supply factor, the value of the diffusion modulation factor, the value of the triggering excitation factor, and the timestamp corresponding to the node. Then, the corresponding collaborative enhancement rate and change gradient are extracted for each node. The collaborative enhancement rate represents the rate of change of the comprehensive value of the risk factors between two adjacent nodes. The collaborative enhancement rate = (comprehensive risk value of node i+1 - comprehensive risk value of node i) / propagation time between node i+1 and node i. The comprehensive risk value is a weighted sum of four risk factors, and the weights are set according to the risk coupling relationship model. The change gradient represents the rate of change of the collaborative enhancement rate with the propagation distance or time. The change gradient = (collaborative enhancement rate of node k+1 - collaborative enhancement rate of node k) / propagation distance or time difference between node k+2 and node k.

[0039] Preferably, the continuity and monotonicity of the cooperative enhancement rate along the path propagation direction are determined. Continuity determination includes checking for missing or invalid data in the cooperative enhancement rate sequence and for spatial discontinuities between adjacent road segments. Monotonicity determination includes checking the trend of the cooperative enhancement rate along the propagation direction. Then, the validity of the continuous amplification chain is determined. If the continuity condition is met and the cooperative enhancement rate is monotonically increasing, it is a valid cooperative chain segment; if the continuity condition is not met or the enhancement rate decreases, it is an invalid chain segment; if the enhancement rate decreases to near 0 or a negative value, it is an enhancement interruption chain segment; if the enhancement rate is negative, indicating a decrease in risk, it is a reverse suppression chain segment. The final output includes valid cooperative chain segments that meet the continuous enhancement condition and invalid chain segments with enhancement interruptions or reverse suppression.

[0040] Preferably, for effective collaborative chain segments, based on the collaborative evolution amplification coefficient and cumulative growth trend, the starting node and ending node of the chain segment are obtained, the number of road segments included in the chain segment is determined, the cumulative amplification coefficient of the chain segment is calculated as (comprehensive risk value at the end of the chain segment - comprehensive risk value at the beginning of the chain segment) / comprehensive risk value at the beginning of the chain segment, the average enhancement rate is calculated, and the evolution intensity index value of each effective collaborative chain segment is calculated. The evolution intensity index is calculated as (cumulative amplification coefficient × average enhancement rate × chain segment length coefficient), where the chain segment length coefficient is calculated as (chain segment length / preset maximum chain segment length value). Then, the difference between adjacent gradient changes is calculated. When the difference exceeds the preset mutation threshold (e.g., more than 3 times the standard deviation), position k+1 is marked as the gradient mutation point, that is, the position where the absolute value of the gradient change changes significantly, indicating that the risk enhancement rate has changed drastically. If the gradient change in the interval is positive and the interval length is greater than or equal to the minimum continuous length (e.g., more than 3 consecutive road segments) and the gradient in the interval shows an overall upward trend, the gradient continuously rising interval is output, that is, the gradient change remains positive for multiple consecutive road segments and gradually increases. Then, each gradient mutation point is used as the switching boundary of the co-evolution stage, and the effective co-evolution chain segment is divided into multiple sub-stages. The gradient continuously rising interval is output as the risk accelerated evolution interval.

[0041] Preferably, within the risk acceleration evolution interval, the change gradient is linearly fitted with the propagation distance, and the growth rate of the change gradient is equal to the slope of the fitted line. The co-evolution acceleration factor is calculated as 1 + γ × (interval length), which represents the degree to which the risk enhancement rate itself is accelerating. When γ > 0, the acceleration factor > 1; when γ = 0, the acceleration factor = 1; and when γ < 0, the acceleration factor < 1. Then, the evolution intensity index is dynamically corrected. The comprehensive evolution index is equal to the original evolution intensity index × the acceleration factor × the interval weight coefficient, where the interval weight coefficient is equal to the average gradient within the interval / the average gradient of the entire chain segment, which is used to reflect the relative importance of the acceleration interval in the entire chain segment. The comprehensive evolution index is a coupled value that simultaneously reflects the risk evolution speed and intensity. Based on the comprehensive evolution index and the co-evolutionary amplification coefficient, effective co-evolutionary segments are classified into states. If the comprehensive evolution index ≤ E1 and the amplification coefficient ≤ A1, it is determined to be an initial coupling state. If E1 < comprehensive evolution index ≤ E2 or A1 < amplification coefficient ≤ A2, it is determined to be a stable enhancement state. If the comprehensive evolution index > E2 and the amplification coefficient > A2, it is determined to be an accelerated instability state. When an effective co-evolutionary segment is determined to be in an accelerated instability state, the state label of the segment is marked as co-evolutionary runaway, and the runaway start position and initial propagation velocity are recorded. Finally, the state classification results of each effective co-evolutionary segment, the segment identifier and position information entering the accelerated instability state, the determination result of the co-evolutionary runaway stage, and the runaway start position and runaway propagation velocity output in the co-evolutionary runaway stage are output as the co-evolutionary characterization results.

[0042] Step S500: Based on the co-evolutionary characterization results, perform path-level cumulative calculations on the risk propagation path network to obtain the degree of risk accumulation on each propagation path and establish a fire risk early warning system.

[0043] Preferably, path-level cumulative calculation refers to superimposing and accumulating risk factors along the propagation path from the starting point to the end point segment by segment, taking into account the coupling enhancement effect and impedance attenuation effect during the propagation process. Specifically, for each risk propagation path, the initial comprehensive risk value of its starting node is obtained, and calculations are performed on each segment along the propagation direction. The coupling enhancement increment = upstream node risk value × unit amplification coefficient, the impedance attenuation amount = (upstream node risk value + coupling enhancement increment) × impedance attenuation coefficient, and the segment output risk value = upstream node risk value + coupling enhancement increment - impedance attenuation amount. The degree of risk accumulation on each propagation path is determined and the risk accumulation level is divided, including at least the accumulation degree at the end of the path, the average accumulation degree of the path, the peak accumulation degree of the path, and the cumulative increase of the path.

[0044] Furthermore, step S500 also includes calculating the path-level risk growth rate based on the degree of risk accumulation and the trend of change of each risk propagation path; when the degree of risk accumulation and the risk growth rate simultaneously meet the preset conditions, it is determined that the corresponding path enters the early warning state and a fire risk early warning is triggered.

[0045] Preferably, the system acquires historical records of risk accumulation at different time points for each path, time series of the co-evolutionary amplification coefficient for each path, and time series of the comprehensive evolution index for each path to determine the trend of change. Then, based on the risk accumulation degree of each risk propagation path, it calculates the change in the risk accumulation degree of a certain risk propagation path per unit time, and then calculates and determines the instantaneous growth rate and the moving average growth rate, thereby outputting the path-level risk growth rate of each path at the current moment. If the risk accumulation degree and the risk growth rate simultaneously meet preset conditions, the corresponding path is determined to enter the warning state and a fire risk warning is triggered. A warning record containing the warning time, warning path, warning level, triggering conditions, risk accumulation degree, and risk growth rate is generated. Then, the path is displayed on the electronic interface of the monitoring system and the warning level is marked. Different frequencies of sound and different colors of light are triggered according to the warning level, and a warning notification is sent to the management personnel terminal. At the same time, a trigger signal is output to the fire linkage control system, such as starting smoke exhaust and closing fire doors.

[0046] Furthermore, step S500 also includes conducting time-series monitoring of fire risk warnings, establishing time-series feedback, and using the time-series feedback to manage the update of fire risk warnings.

[0047] Preferably, after a fire risk warning is triggered, the risk accumulation level, risk growth rate, co-evolution amplification coefficient, comprehensive evolution index, and state classification results along the warning path are continuously tracked. Specifically, for each triggered warning, an independent monitoring record is established, including the warning path ID, warning level, trigger time, and current status. Different monitoring frequencies are set for different warning levels, with higher frequencies for more severe warnings. The results of each collection are stored in the monitoring record of that warning in chronological order, outputting a time-series monitoring dataset for each triggered warning. This dataset is then compared with preset judgment rules to generate feedback signals and feedback intensity for fire risk warning update management. The feedback intensity is calculated as: current risk accumulation level / risk accumulation level at the time of triggering, representing the relative change in risk compared to the warning trigger time. Finally, the warning status is dynamically adjusted based on the feedback results. When the judgment result is to upgrade, a higher-level alarm signal is issued and the level field in the warning record is updated. When the judgment result is to downgrade, the intensity of the alarm signal is reduced and the level field in the warning record is updated. When the judgment result is to lift the warning, the alarm signal is stopped and the lifting time and reason are marked in the warning record. When the judgment result is to maintain the warning, the current alarm status is kept unchanged and the time-series monitoring continues.

[0048] Furthermore, step S500 also includes, when multiple risk propagation paths exist simultaneously, comparing and analyzing the degree of risk accumulation and the state of collaborative evolution of each risk propagation path, identifying the dominant path with the highest risk evolution speed, and using the influence relationship between the dominant path and the remaining risk propagation paths to perform collaborative early warning marking of potential risk propagation paths.

[0049] Preferably, if multiple risk propagation paths exist simultaneously, a comparative analysis is performed on the risk accumulation degree and co-evolution status of each risk propagation path. This includes ranking all paths from highest to lowest risk accumulation degree, from highest to lowest risk evolution speed, from highest to lowest co-evolution status level, and from highest to lowest comprehensive evolution index. The index values ​​of each path are normalized to relative values ​​between 0 and 1. The comprehensive score is calculated as: ω1 × normalized risk accumulation degree + ω2 × normalized risk evolution speed + ω3 × normalized co-evolution status + ω4 × normalized comprehensive evolution index, where ω1, ω2, ω3, and ω4 are preset weighting coefficients. The sum is 1, such as ω1=0.3, ω2=0.3, ω3=0.2, ω4=0.2. Then, the paths are sorted from high to low according to the comprehensive score, and the ranking list of each path and the comparison results of each indicator are output. Finally, among the multiple risk propagation paths, the dominant path with the highest risk evolution speed and potential impact on other paths is identified. Based on the spatial topology and risk propagation dynamics, the risk transmission or synergistic enhancement effect of the risk evolution of the dominant path on other paths is determined. The impact type, impact coefficient, and risk transmission amount of each remaining path relative to the dominant path are determined. Then, based on the degree of impact and its own risk status, a collaborative early warning label is assigned to achieve multi-path linkage early warning management.

[0050] In the above text, refer to Figure 1 A fire risk collaborative assessment and early warning method according to embodiments of the present invention is described in detail. Next, reference will be made to... Figure 2 A fire risk collaborative assessment and early warning system according to an embodiment of the present invention is described.

[0051] The fire risk collaborative assessment and early warning system according to embodiments of the present invention addresses the technical problems in existing technologies, such as the lack of coupling relationships among multiple risk factors, the disconnect between risk propagation paths and spatial structures, and the lack of temporal collaborative evolution tracking. It achieves the technical effects of improving the coupling characterization capability of fire risk assessment, the accuracy of path-level risk accumulation calculation, and dynamic collaborative early warning. Figure 2 As shown, the fire risk collaborative assessment and early warning system includes: vector construction module 10, model construction module 20, network construction module 30, result output module 40, and fire risk early warning module 50.

[0052] The vector construction module 10 is used to acquire multi-source sensing data monitored by sensors within the target area, and extract basic risk factors representing the heat accumulation process, combustible material release process, electrical anomaly process, and ventilation diffusion capacity based on the multi-source sensing data, and construct a risk factor vector. The model construction module 20 is used to analyze the interaction relationship between various risk factors based on the risk factor vector, identify risk factor combinations with amplification effects, and construct a risk coupling relationship model reflecting the coupling strength of risk factors. The network construction module 30 is used to construct a risk propagation path network including risk propagation direction and propagation impedance under the constraints of the risk coupling relationship model and in combination with the spatial structure and connectivity of the target area. The result output module 40 is used to perform time-series tracking of the change process of the risk factors on the path based on the risk propagation path network, analyze the co-evolution state of the risk factors, and output the co-evolution characterization result. The fire risk early warning module 50 is used to perform path-level cumulative calculation of risk on the risk propagation path network based on the co-evolution characterization result, obtain the risk accumulation degree on each propagation path, and establish a fire risk early warning.

[0053] The specific configuration of the result output module 40 will be described in detail below. The result output module 40 further includes: based on the interaction mechanism between different risk factors in the risk coupling relationship model, classifying the risk factors into heat-driven factors, combustible material supply factors, diffusion modulation factors, and triggering excitation factors; and establishing path evolution sequences reflecting the interaction order of different types of risk factors on each path of the risk propagation path network; for each risk propagation path, constructing segment-level collaborative action units based on the spatial connectivity and environmental attributes between adjacent nodes on the path; within each collaborative action unit, based on the release-promoting effect of heat-driven factors on combustible material supply factors, and the effect of diffusion modulation factors on heat and combustible gases... Amplification or inhibition effects, and the instantaneous excitation effects of triggering factors, are used to construct local co-evolution rules for multi-factor coupling. The risk factor responses are decomposed and projected along the corresponding risk propagation paths, and each risk factor vector is locally co-processed within each co-action unit based on the local co-evolution rules according to the action order defined by the path evolution sequence. The results of the local co-processing are then passed step by step along the path propagation direction to identify the continuous amplification chain formed between risk factors by heat accumulation, enhanced release of combustibles, accelerated diffusion, and triggering excitation, and the co-evolution amplification coefficient is calculated on the continuous amplification chain. The co-evolution characterization results are output using the co-evolution amplification coefficient.

[0054] The specific configuration of the result output module 40 will be described in detail below. The result output module 40 further includes: constructing a cooperative evolution state sequence along the risk propagation path based on the local cooperative processing results of each cooperative unit in the continuous amplification chain; extracting the corresponding cooperative enhancement rate and change gradient for each node in the cooperative evolution state sequence; determining the effectiveness of the continuous amplification chain based on the continuity and monotonicity of the cooperative enhancement rate in the path propagation direction, identifying effective cooperative chain segments that meet the continuous enhancement conditions, and ineffective chain segments with enhancement interruption or reverse suppression; calculating the chain segment-level evolution intensity index for the effective cooperative chain segments based on the cooperative evolution amplification coefficient and cumulative growth trend; and identifying gradient mutation points in the cooperative enhancement process based on the distribution characteristics of the change gradient within the effective cooperative chain segments. The gradient continuously rises within a certain range, and the gradient mutation point is used as the switching boundary of the co-evolution stage. The continuously rising gradient range is also used as the risk acceleration evolution range. Within the risk acceleration evolution range, the co-evolution acceleration factor is calculated by combining the evolution intensity index and the growth rate of the changing gradient. Based on the co-evolution acceleration factor, the evolution intensity index at the chain segment level is dynamically corrected to obtain a comprehensive evolution index reflecting the coupling relationship between the speed and intensity of risk evolution. Based on the comprehensive evolution index and the co-evolution amplification coefficient, the effective co-evolution chain segment is classified into three states: initial coupling state, stable enhancement state, and accelerated instability state. Under the accelerated instability state, the corresponding risk propagation path is determined to have entered the co-evolution runaway stage, which is output as the co-evolution characterization result.

[0055] The specific configuration of the model building module 20 will be described in detail below. The model building module 20 further includes: extracting the response lag characteristics and synchronous change characteristics of risk factors in the time dimension based on the historical change trajectories of each risk factor; using these characteristics as the coupling judgment criteria to screen risk factor combinations that exhibit a mutually driving relationship under time-related conditions; constructing coupling triggering rules describing the triggering conditions of the amplification effect between risk factors for the screened risk factor combinations, combined with corresponding environmental state parameters; defining the state boundary of the transition from weak correlation to strong coupling between risk factors based on the coupling triggering rules; and, under the constraints of the coupling triggering rules, segmenting the response intensity of the risk factor combinations in different state intervals to construct a coupling intensity mapping relationship that dynamically adjusts with changes in environmental state and factors, thus establishing the risk coupling relationship model.

[0056] The specific configuration of the fire risk early warning module 50 will be described in detail below. The fire risk early warning module 50 further includes: calculating the path-level risk growth rate based on the degree of risk accumulation and the trend of change of each risk propagation path; when the degree of risk accumulation and the risk growth rate simultaneously meet preset conditions, determining that the corresponding path enters the early warning state and triggering a fire risk early warning.

[0057] The specific configuration of the fire risk early warning module 50 will be described in detail below. The fire risk early warning module 50 further includes: time-series monitoring of fire risk early warnings, establishment of time-series feedback, and management of early warning updates using the time-series feedback.

[0058] The specific configuration of the fire risk early warning module 50 will be described in detail below. The fire risk early warning module 50 further includes: when multiple risk propagation paths exist simultaneously, comparing and analyzing the risk accumulation degree and collaborative evolution status of each risk propagation path, identifying the dominant path with the highest risk evolution speed, and using the influence relationship between the dominant path and the remaining risk propagation paths to execute collaborative early warning marking of potential risk propagation paths.

[0059] The fire risk collaborative assessment and early warning system provided in this embodiment of the invention can execute the fire risk collaborative assessment and early warning method provided in this embodiment of the invention, and has the corresponding functional modules and beneficial effects of the execution method.

[0060] The above description is merely a preferred embodiment of the present invention and is not intended to limit the present invention in any way. Although the present invention has been disclosed above with reference to preferred embodiments, it is not intended to limit the present invention. Any person skilled in the art can make some modifications or alterations to the above-disclosed technical content to create equivalent embodiments without departing from the scope of the present invention. Any modifications, equivalent changes, and alterations made to the above embodiments based on the technical essence of the present invention without departing from the scope of the present invention shall still fall within the scope of the present invention.

Claims

1. A method for fire risk collaborative assessment and early warning, characterized in that, The method includes: Acquire multi-source sensing data monitored by sensors within the target area, and extract basic risk factors characterizing the heat accumulation process, combustible material release process, electrical anomaly process and ventilation diffusion capacity based on the multi-source sensing data, and construct a risk factor vector; Based on the risk factor vector, the interaction between each risk factor is analyzed, risk factor combinations with amplification effects are identified, and a risk coupling relationship model reflecting the coupling strength of risk factors is constructed. Under the constraints of the risk coupling relationship model, and combined with the spatial structure and connectivity of the target area, a risk propagation path network including risk propagation direction and propagation impedance is constructed. Based on the risk propagation path network, the change process of the risk factors along the path is tracked in time, the co-evolution state of the risk factors is analyzed, and the co-evolution characterization results are output. Based on the co-evolutionary characterization results, path-level cumulative calculations are performed on the risk propagation path network to obtain the degree of risk accumulation on each propagation path, and a fire risk early warning system is established.

2. The fire risk synergistic assessment and early warning method according to claim 1, characterized in that, Analyze the co-evolutionary state of risk factors and output the co-evolutionary characterization results, including: Based on the interaction mechanism between different risk factors in the risk coupling relationship model, the risk factors are divided into heat-driven factors, combustible material supply factors, diffusion modulation factors and triggering excitation factors, and a path evolution sequence reflecting the interaction order of different types of risk factors is established on each path of the risk propagation path network. For each risk propagation path, based on the spatial connectivity and environmental attributes between adjacent nodes on the path, a road segment-level collaborative unit is constructed; Within each synergistic unit, based on the promoting effect of thermal driving factors on the release of combustible material supply factors, the amplifying or inhibiting effect of diffusion modulation factors on heat and combustible gas, and the instantaneous excitation effect of triggering excitation factors, a local synergistic evolution rule of multi-factor coupling is constructed. The risk factor response is decomposed and projected along the corresponding risk propagation path, and each risk factor vector is locally processed in each synergistic unit based on the local synergistic evolution rule according to the action order defined by the path evolution sequence. The results of local collaborative processing are transmitted step by step in the path propagation direction. The continuous amplification chain formed by heat accumulation, enhanced release of combustibles, accelerated diffusion and triggering excitation between risk factors is identified, and the collaborative evolution amplification coefficient is calculated on the continuous amplification chain. The co-evolutionary amplification coefficient is used to output the co-evolutionary characterization results.

3. The fire risk synergistic assessment and early warning method of claim 2, wherein, The co-evolutionary characterization results are output using the co-evolutionary amplification coefficient, including: Based on the local collaborative processing results of each collaborative unit in the continuous amplification chain, a collaborative evolution state sequence along the risk propagation path is constructed, and the corresponding collaborative enhancement rate and change gradient are extracted for each node in the collaborative evolution state sequence. Based on the continuity and monotonicity of the cooperative enhancement rate in the path propagation direction, the effectiveness of the continuous amplification chain is determined, and effective cooperative chain segments that meet the continuous enhancement conditions and ineffective chain segments with enhancement interruption or reverse suppression are identified. For the effective cooperative chain segment, the chain segment-level evolution intensity index is calculated based on the cooperative evolution amplification coefficient and cumulative growth trend; Based on the distribution characteristics of the changing gradient within the effective collaborative chain segment, gradient mutation points and gradient continuous rise intervals in the collaborative enhancement process are identified, and the gradient mutation points are used as the switching boundary of the collaborative evolution stage, and the gradient continuous rise intervals are used as the risk acceleration evolution intervals. Within the risk acceleration evolution range, a co-evolution acceleration factor is calculated by combining the evolution intensity index and the growth rate of the change gradient. Based on the co-evolution acceleration factor, the chain segment-level evolution intensity index is dynamically corrected to obtain a comprehensive evolution index that reflects the coupling relationship between the risk evolution speed and intensity. Based on the comprehensive evolution index and the co-evolution amplification coefficient, the effective co-evolutionary chain segment is classified into initial coupling state, stable enhancement state and accelerated instability state. Under the accelerated instability state, the corresponding risk propagation path is determined to have entered the co-evolutionary runaway stage, which is output as the co-evolutionary characterization result.

4. The fire risk collaborative assessment and early warning method as described in claim 1, characterized in that, include: Based on the historical change trajectory of each risk factor, the response lag characteristics and synchronous change characteristics of the risk factors in the time dimension are extracted, and the response lag characteristics and synchronous change characteristics are used as the coupling judgment criteria to screen the combination of risk factors that show a mutually driving relationship under the time correlation condition. For the selected risk factor combinations, combined with the corresponding environmental state parameters, a coupling triggering rule is constructed to describe the triggering conditions of the amplification effect between risk factors, and the state boundary of the transformation from weak correlation to strong coupling between risk factors is defined based on the coupling triggering rule. Under the constraints of the coupling triggering rule, the response intensity of the risk factor combination in different state intervals is characterized in segments, and a coupling intensity mapping relationship that is dynamically adjusted with changes in environmental state and factors is constructed to establish the risk coupling relationship model.

5. The fire risk collaborative assessment and early warning method as described in claim 1, characterized in that, Establish a fire risk early warning system, including: Calculate the path-level risk growth rate based on the degree of risk accumulation and the changing trend of each risk propagation path. When the risk accumulation level and risk growth rate simultaneously meet the preset conditions, the corresponding path is determined to enter the early warning state and a fire risk warning is triggered.

6. The fire risk collaborative assessment and early warning method as described in claim 5, characterized in that, Fire risk warnings are monitored in real time, and real-time feedback is established. The real-time feedback is then used to manage the update of fire risk warnings.

7. The fire risk collaborative assessment and early warning method as described in claim 5, characterized in that, When multiple risk propagation paths exist simultaneously, the risk accumulation degree and co-evolution status of each risk propagation path are compared and analyzed to identify the dominant path with the highest risk evolution speed. The influence relationship between the dominant path and the remaining risk propagation paths is used to implement the co-early warning labeling of potential risk propagation paths.

8. A fire risk collaborative assessment and early warning system, characterized in that, The system is used to implement the fire risk collaborative assessment and early warning method according to any one of claims 1 to 7, and the system includes: The vector construction module is used to acquire multi-source sensing data monitored by sensors within the target area, and extract basic risk factors that characterize the heat accumulation process, combustible material release process, electrical anomaly process and ventilation diffusion capacity based on the multi-source sensing data, and construct a risk factor vector. The model building module is used to analyze the interaction between risk factors based on the risk factor vector, identify risk factor combinations with amplification effects, and construct a risk coupling relationship model that reflects the coupling strength of risk factors. The network construction module is used to construct a risk propagation path network, including risk propagation direction and propagation impedance, under the constraints of the risk coupling relationship model and in combination with the spatial structure and connectivity of the target area. The results output module is used to perform time-series tracking of the changes of the risk factors along the path based on the risk propagation path network, analyze the co-evolution state of the risk factors, and output the co-evolution characterization results. The fire risk early warning module is used to perform path-level cumulative calculation of risks on the risk propagation path network based on the collaborative evolution characterization results, obtain the degree of risk accumulation on each propagation path, and establish a fire risk early warning.

Images