AI-driven resource scheduling methods and systems for smart fire protection

Abstract

This application relates to the field of public service data processing technology, and provides an AI-driven resource scheduling method and system for smart fire protection. First data is acquired through intelligent sensors and multi-source data fusion is performed. A first tensor composed of a spatial map is generated through neural network and modeling processing, and node and edge features are extracted from it. A risk propagation matrix for the fire situation is determined by linear transformation of the node and edge features based on a learnable weight matrix and activation function. Based on the risk propagation matrix and resource data, alternative strategies are generated, and the total loss of the alternative strategies is minimized to generate a resource allocation strategy. Based on the resource allocation strategies of adjacent areas and a preset inter-regional cooperation agreement, a scheduling strategy is determined through Nash equilibrium, and the generated scheduling instructions are sent to the target terminal. Through multi-source data fusion and spatiotemporal modeling, accurate risk prediction is achieved, effectively reducing the risk of fire spread and emergency response costs, and improving the efficiency and effectiveness of fire resource scheduling.

Description

Technical Field

[0001] This application relates to the field of public service data processing technology, and more specifically, to an AI-driven resource scheduling method and system for smart fire protection. Background Technology

[0002] With the development of smart fire protection, promoting the informatization and intelligent transformation of fire protection and building an IoT-based fire supervision model has become crucial. Traditional fire protection systems rely heavily on manual inspections and post-incident response, resulting in problems such as delayed response and uneven resource allocation. For example, after a fire breaks out, dispatchers need to manually analyze information such as the location of the fire, the distribution of fire stations, and road traffic, leading to low decision-making efficiency and a risk of missing the best rescue opportunity.

[0003] Existing technologies, through the deep integration of IoT, big data, AI, and 5G, provide a technological foundation for smart fire protection. For example, sensors monitor the status of fire protection facilities and environmental parameters in real time, combined with AI for dynamic risk assessment. However, due to the large number of data types collected by various sensors, it is difficult to extract precise features. Furthermore, the complexity of the external environment and fire protection resources makes it difficult to efficiently and accurately determine fire protection methods and allocate resources, resulting in low efficiency and effectiveness in fire protection resource allocation. Summary of the Invention

[0004] This application provides an AI-driven resource scheduling method and system for smart fire protection, which can at least partially solve the problem of low efficiency and effectiveness of fire resource scheduling.

[0005] Other features and advantages of this application will become apparent from the following detailed description, or may be learned in part from practice of this application.

[0006] According to one aspect of this application, an AI-driven resource scheduling method for smart fire protection is provided, comprising: acquiring first data through intelligent sensors, wherein the first data includes environmental data and satellite data; performing multi-source data fusion on the first data, and generating a first tensor composed of a spatial map through neural network and modeling processing; extracting node features and edge features from the first tensor and the spatial map, and performing linear transformation on the node features and edge features based on a learnable weight matrix and an activation function to determine a fire risk propagation matrix; generating alternative resource allocation strategies based on the risk propagation matrix and the acquired resource data, and minimizing the total loss of the alternative strategies to generate a resource allocation strategy; determining a scheduling strategy through Nash equilibrium based on the resource allocation strategies of adjacent areas and a preset inter-regional cooperation protocol; and determining a target terminal related to the scheduling strategy based on a preset fire protection system, and sending the scheduling instructions generated by the scheduling strategy to the target terminal.

[0007] In this application, based on the aforementioned scheme, the step of performing multi-source data fusion on the first data and generating a first tensor composed of a spatial map through neural network and modeling processing includes: performing three-dimensional convolution processing on the first data based on a preset spatiotemporal convolution kernel to generate second data; generating dynamic weight coefficients based on an attention mechanism, and processing the second data through the dynamic weight coefficients to generate third data; modeling the satellite data in the first data to generate a geographic topology, and processing the geographic topology through a graph neural network to generate a spatial map; and generating a first tensor composed of the spatial map based on the third data and the spatial map.

[0008] In this application, based on the aforementioned scheme, the step of extracting node features and edge features from the first tensor and the spatial graph, and performing a linear transformation on the node features and edge features based on a learnable weight matrix and an activation function to determine the fire risk propagation matrix includes: aggregating information of neighboring nodes in the spatial graph based on a preset graph attention network to generate node features; extracting edge features between neighboring nodes from the first tensor and the spatial graph, and concatenating the node features and the edge features to generate a first feature; performing a linear transformation on the first feature based on a first learnable weight matrix and a first activation function to generate a second feature; and performing a linear transformation on the second feature based on a second learnable weight matrix and a second activation function to obtain the fire risk propagation matrix.

[0009] In this application, based on the aforementioned scheme, the step of generating alternative resource allocation strategies based on the risk propagation matrix and the acquired resource data, and minimizing the total loss of the alternative strategies to generate a resource allocation strategy, includes: generating alternative resource allocation strategies based on the risk propagation matrix and the acquired resource data; determining the risk loss based on the risk parameters of the building at a set time in the risk propagation matrix; determining the resource mobilization cost based on the time-varying cost function of the resources in the alternative strategies at each time; determining the timeout penalty based on the time of resource arrival at the building in the alternative strategies; constructing the total loss based on the risk loss, the resource mobilization cost, and the timeout penalty; minimizing the total loss by adjusting the alternative strategies, and outputting the final resource allocation strategy.

[0010] In this application, based on the aforementioned scheme, determining the risk loss based on the risk parameters of the building at a set time in the risk propagation matrix includes: determining the risk loss based on the risk parameters of the building at a set time in the risk propagation matrix. for:

[0011]

[0012] Where t and T represent time and the total duration of the strategy prediction, respectively. This represents the preset discount factor. Let N represent the risk parameters of building i in the risk propagation matrix at time t, where i and N represent the identifier and total number of buildings in the risk propagation matrix, respectively. This indicates an indicator function.

[0013] In this application, based on the aforementioned scheme, the step of determining the scheduling strategy through Nash equilibrium based on the resource allocation strategy of adjacent regions and the preset inter-regional cooperation protocol includes: obtaining the resource allocation strategy of adjacent regions and the preset inter-regional cooperation protocol; determining the utility function based on the resource allocation strategy of adjacent regions and the preset inter-regional cooperation protocol; and iterating the utility function through the alternating direction multiplier method to determine the scheduling strategy.

[0014] In this application, based on the aforementioned scheme, the step of determining the target terminal related to the dispatch strategy based on a preset fire protection system and sending the dispatch instructions generated by the dispatch strategy to the target terminal includes: determining the target terminal related to the dispatch strategy based on a preset fire protection system; generating a dispatch instruction based on the dispatch strategy and the target terminal; and sending the dispatch instruction to the target terminal.

[0015] According to one aspect of this application, an AI-driven resource scheduling system for smart fire protection is provided, comprising:

[0016] An acquisition unit is configured to acquire first data via a smart sensor, wherein the first data includes environmental data and satellite data;

[0017] The coupling unit is used to perform multi-source data fusion on the first data and generate a first tensor composed of a spatial map through neural network and modeling processing.

[0018] The feature unit is used to extract node features and edge features from the first tensor and spatial graph, and to perform a linear transformation on the node features and edge features based on a learnable weight matrix and activation function to determine the risk propagation matrix of the fire.

[0019] The strategy unit is used to generate alternative resource allocation strategies based on the risk propagation matrix and the acquired resource data, and to minimize the total loss of the alternative strategies, so as to generate a resource allocation strategy.

[0020] The scheduling unit is used to determine the scheduling strategy through Nash equilibrium based on the resource allocation strategy of adjacent regions and the preset inter-regional cooperation protocol.

[0021] The instruction unit is used to determine the target terminal related to the dispatch strategy based on the preset fire protection system, and send the dispatch instructions generated by the dispatch strategy to the target terminal.

[0022] According to one aspect of this application, a computer-readable medium is provided having a computer program stored thereon, which, when executed by a processor, implements the AI-driven resource scheduling method for smart fire protection as described in the above embodiments.

[0023] According to one aspect of this application, an electronic device is provided, comprising: one or more processors; and a storage device for storing one or more programs, which, when executed by the one or more processors, cause the one or more processors to implement the AI-driven resource scheduling method for smart fire protection as described in the above embodiments.

[0024] According to one aspect of this application, a computer program product or computer program is provided, comprising computer instructions stored in a computer-readable storage medium. A processor of a computer device reads the computer instructions from the computer-readable storage medium and executes the computer instructions, causing the computer device to perform the AI-driven resource scheduling method for smart fire protection provided in the various optional implementations described above.

[0025] The technical solution of this application acquires first data through intelligent sensors, performs multi-source data fusion on the first data, and generates a first tensor composed of a spatial map through neural network and modeling processing. Node features and edge features are extracted from the first tensor and the spatial map. Based on a learnable weight matrix and activation function, the node features and edge features are linearly transformed to determine the fire risk propagation matrix. Based on the risk propagation matrix and acquired resource data, alternative resource allocation strategies are generated, and the total loss of the alternative strategies is minimized to generate a resource allocation strategy. Based on the resource allocation strategies of adjacent areas and a preset inter-regional cooperation protocol, a scheduling strategy is determined through Nash equilibrium. Based on a preset fire protection system, target terminals related to the scheduling strategy are identified, and the scheduling instructions generated by the scheduling strategy are sent to the target terminals. The above process achieves accurate risk prediction through multi-source data fusion and spatiotemporal modeling. Combined with dynamic resource allocation and cross-regional collaborative scheduling, it constructs a closed-loop system from data perception to instruction execution, effectively reducing the risk of fire spread and emergency response costs, and improving the efficiency and effectiveness of fire resource scheduling.

[0026] It should be understood that the above general description and the following detailed description are exemplary and explanatory only, and do not limit this application. Attached Figure Description

[0027] The accompanying drawings, which are incorporated in and form part of this specification, illustrate embodiments consistent with this application and, together with the description, serve to explain the principles of this application. It is obvious that the drawings described below are merely some embodiments of this application, and those skilled in the art can obtain other drawings based on these drawings without any inventive effort.

[0028] Figure 1 The flowchart illustrating an AI-driven resource scheduling method for smart fire protection in one embodiment of this application is shown.

[0029] Figure 2 The flowchart illustrating the generation of the first tensor is shown in one embodiment of this application.

[0030] Figure 3 The illustration shows a schematic diagram of an AI-driven resource scheduling system for smart fire protection in one embodiment of this application.

[0031] Figure 4 A schematic diagram of the structure of a computer system suitable for implementing the electronic device of the present application is shown. Detailed Implementation

[0032] Exemplary embodiments will now be described more fully with reference to the accompanying drawings. However, these exemplary embodiments can be implemented in many forms and should not be construed as limited to the examples set forth herein; rather, these embodiments are provided to make this application more comprehensive and complete, and to fully convey the concept of the exemplary embodiments to those skilled in the art.

[0033] Furthermore, the described features, structures, or characteristics can be combined in any suitable manner in one or more embodiments. Numerous specific details are provided in the following description to give a thorough understanding of embodiments of this application. However, those skilled in the art will recognize that the technical solutions of this application can be practiced without one or more of the specific details, or other methods, components, apparatuses, steps, etc., can be employed. In other instances, well-known methods, apparatuses, implementations, or operations are not shown or described in detail to avoid obscuring various aspects of this application.

[0034] The block diagrams shown in the accompanying drawings are merely functional entities and do not necessarily correspond to physically independent entities. That is, these functional entities can be implemented in software, in one or more hardware modules or integrated circuits, or in different network and / or processor devices and / or microcontroller devices.

[0035] The flowcharts shown in the accompanying drawings are merely illustrative and do not necessarily include all content and operations / steps, nor do they necessarily have to be performed in the described order. For example, some operations / steps can be broken down, while others can be combined or partially combined; therefore, the actual execution order may change depending on the specific circumstances.

[0036] The implementation details of the technical solution of this application are described below:

[0037] Figure 1 A flowchart of an AI-driven resource scheduling method for smart fire protection according to an embodiment of this application is shown. (Refer to...) Figure 1 As shown, this AI-driven resource scheduling method for smart fire protection includes at least steps S110 to S160, which are detailed below:

[0038] S110 acquires first data through a smart sensor, wherein the first data includes environmental data and satellite data.

[0039] In one embodiment of this application, smart sensors are deployed in the environment, such as inside buildings, public places, and forest areas, within a fire protection system to comprehensively perceive the environmental conditions. These sensors possess high sensitivity and accuracy, capable of collecting environmental data in real time, such as temperature, humidity, smoke concentration, and gas composition, providing the fire protection system with first-hand environmental information. Simultaneously, some smart sensors also have the ability to receive satellite data, such as satellite remote sensing images and meteorological satellite data, thereby further broadening the dimensions of data sources.

[0040] In addition, the first data in this embodiment also includes: various sensor data, historical fire cases obtained from the database, geographic topology data, and meteorological data.

[0041] In one embodiment of this application, the data collected by the smart sensors, including environmental data and satellite data, needs to be transmitted in real time to the data center of the fire protection system via a wireless or wired network. During the transmission process, the data undergoes preliminary preprocessing, such as data cleaning, noise reduction, and format conversion, to ensure the accuracy and consistency of the data and improve the efficiency and accuracy of subsequent data analysis.

[0042] In one embodiment of this application, distributed storage, cloud computing, and other technologies are employed to classify, archive, and back up real-time data, ensuring data security and accessibility. Simultaneously, data query, analysis, and visualization functions are provided, enabling firefighters to quickly obtain the necessary information and providing strong support for developing dispatch strategies and responding to fires.

[0043] The above process, through the acquisition of real-time environmental and satellite data by intelligent sensors, achieves comprehensive perception of the fire protection environment. Multi-source data fusion technology integrates this data from different sources and in different formats to generate the first tensor, which consists of a spatial map. This tensor not only contains rich spatiotemporal information but also reflects the complex correlations between the data, providing a comprehensive and accurate data foundation for subsequent risk analysis and resource allocation.

[0044] S120, perform multi-source data fusion on the first data, and generate a first tensor composed of spatial maps through neural network and modeling processing.

[0045] In one embodiment of this application, during the process of multi-source data fusion of the first data and generating a first tensor composed of a spatial map through neural network and modeling processing, real-time first data from smart sensors is first collected. This data covers multiple sources such as environmental data and satellite data, and is characterized by multi-source heterogeneity. In order to make full use of this data, multi-source data fusion technology is adopted to integrate data from different sources and in different formats to form a unified and complete dataset. Next, the fused data is deeply modeled and processed using a neural network. The neural network can automatically learn complex patterns and spatiotemporal correlations in the data, thereby extracting more valuable features and information and generating a spatiotemporally coupled first tensor. This tensor not only contains rich spatiotemporal information, but also reflects the complex correlations between data, providing a solid foundation for subsequent risk propagation modeling and resource scheduling optimization.

[0046] like Figure 2 As shown, in one embodiment of this application, the first data undergoes multi-source data fusion, and through neural network and modeling processing, a first tensor composed of a spatial map is generated, including:

[0047] S210, based on a preset spatiotemporal convolution kernel, perform three-dimensional convolution processing on the first data to generate the second data;

[0048] S220, Based on the attention mechanism, dynamic weight coefficients are generated, and the second data is processed using the dynamic weight coefficients to generate the third data;

[0049] S230, Model the satellite data in the first data to generate a geographic topology, and process the geographic topology through a graph neural network to generate a spatial map;

[0050] S240, Based on the third data and the spatial map, generate a first tensor composed of the spatial map.

[0051] In one embodiment of this application, for the temporal sequence of the first data of the m-th class, a learnable spatiotemporal convolution kernel is used to perform three-dimensional convolution processing to generate the second data. This is to capture the local spatiotemporal features in the first data, that is, the change patterns of the first data in time and space.

[0052] In one embodiment of this application, dynamic weight coefficients are generated through an attention mechanism. This attention mechanism dynamically adjusts the weights of different data sources based on the characteristics of the input data, reflecting the confidence level of the m-th data source and enabling the model to focus on more important information. Based on the dynamically generated weight coefficients, the second data is weighted using these coefficients to generate the third data.

[0053] In one embodiment of this application, satellite data in the first data is modeled to generate a geographic topology. This geographic topology is then processed using a graph neural network to generate a spatial map. The graph neural network embeds the geographic topology into spatial adjacency relationships, generating a three-dimensional spatial knowledge graph. By combining the impact of geospatial information on fire risk, the graph neural network captures the dependencies within the spatial structure. Simultaneously, a balancing coefficient is introduced to balance the influence of different components, ensuring the model maintains stability when fusing multi-source data.

[0054] Specifically, in the processing of graph neural networks, the representation of the current node is updated by aggregating information from neighboring nodes, thereby encoding the structure and feature information of the graph. This process includes three steps: message generation, aggregation, and update. In each layer of the network, nodes receive messages from neighboring nodes, which are generated based on the features of the neighbors and edge weights. Subsequently, nodes aggregate these messages; commonly used aggregation functions include summation, averaging, or max pooling. Finally, nodes combine the aggregated messages with their own features and update their state through a nonlinear transformation.

[0055] In one embodiment of this application, a first tensor composed of the spatial map is generated based on the third data and the spatial map. Summarizing the above process, the generated first tensor... for:

[0056]

[0057] Where m and M represent the category identifier and the total number of categories of the first data, respectively. Indicates the dynamic weighting coefficient. This represents 3D convolution processing. This represents the time series sequence corresponding to the first data in the m-th class. Represents a learnable spatiotemporal convolution kernel. Represents the balance coefficient. The first tensor generated in this embodiment integrates the local spatiotemporal patterns, dynamic weights, and geographic topological constraints of different sensor data to generate a feature representation that comprehensively reflects fire risk.

[0058] The first tensor determined in this embodiment is the dynamic risk feature tensor. This tensor integrates information from multiple data sources, including sensor data, historical fire cases, geographic topology data, and meteorological data, providing a foundation for subsequent risk assessment.

[0059] S130, extract node features and edge features from the first tensor and spatial graph, and perform linear transformation on the node features and edge features based on the learnable weight matrix and activation function to determine the risk propagation matrix of the fire.

[0060] In one embodiment of this application, node features and edge features are extracted from a first tensor and a spatial graph. Node features include the physical attributes of buildings, historical risk data, etc., while edge features include spatial relationships between buildings, risk propagation paths, etc. These features are linearly transformed using a learnable weight matrix and an activation function. The learnable weight matrix can automatically adjust the importance of features according to different tasks and data, while the activation function can introduce nonlinear factors to enhance the expressive power of the model. Finally, the fire risk propagation matrix is ​​determined using the linearly transformed node features and edge features. This matrix reflects the probability and intensity of risk propagation between buildings, providing an important foundation for subsequent risk propagation modeling and resource scheduling optimization.

[0061] In one embodiment of this application, node features and edge features are extracted from the first tensor and spatial graph, and a linear transformation is performed on the node features and edge features based on a learnable weight matrix and activation function to determine the fire risk propagation matrix, including:

[0062] Based on a preset graph attention network, information of neighboring nodes in the spatial graph is aggregated to generate node features;

[0063] The edge features between the neighboring nodes are extracted from the first tensor and the spatial graph, and the node features and the edge features are concatenated to generate the first feature;

[0064] The first feature is linearly transformed based on the first learnable weight matrix and the first activation function to generate the second feature;

[0065] The risk propagation matrix of the fire is obtained by linearly transforming the second feature based on the second learnable weight matrix and the second activation function.

[0066] In one embodiment of this application, based on a preset graph attention network, information of neighboring nodes in the spatial graph is aggregated to generate node features. The graph attention network (GAT) is used on neighboring nodes i and j in the spatial graph, utilizing the graph attention mechanism to enable node features to capture local information of the graph structure, and to generate node features by aggregating the information of their neighboring nodes.

[0067] In one embodiment of this application, edge features between neighboring nodes are extracted from the first tensor and the spatial graph. These edge features may include building spacing, connectivity, and wind direction factors, which are crucial for quantifying risk propagation paths. Subsequently, the node features and the edge features are concatenated to generate a first feature, comprehensively considering the interactions between nodes and the attributes of the edges, thereby more accurately quantifying risk propagation paths.

[0068] In one embodiment of this application, a second feature is generated by linearly transforming the first feature based on a first learnable weight matrix and a first activation function. Specifically, the concatenated features are linearly transformed using the first learnable weight matrix, a bias is added, and a ReLU activation function is applied. By introducing nonlinearity through multiple linear transformations and activation functions, the model is able to learn complex nonlinear relationships.

[0069] In one embodiment of this application, the second feature is linearly transformed based on a second learnable weight matrix and a second activation function to obtain the fire risk propagation matrix. Specifically, the transformed feature is linearly transformed using the second learnable weight matrix, a bias is added, and a sigmoid function is applied to obtain the fire risk propagation matrix.

[0070] Based on the above process, a risk propagation matrix for the fire is generated. for:

[0071]

[0072] in, These represent the information of neighbor nodes i and j, respectively. Represents the edge characteristics between neighboring nodes. Let these represent the first learnable weight matrix and the second learnable weight matrix, respectively. These represent the bias terms, This represents the Sigmoid function. By calculating the risk propagation matrix of a fire, we can intuitively understand the spatiotemporal spread path of a fire within a building complex, providing important information for the dispatch of fire-fighting resources.

[0073] The above process extracts node and edge features from the first tensor and spatial graph. These features reflect the spatial relationships between buildings and the risk propagation paths. By applying a linear transformation to these features using a learnable weight matrix and activation function, the complex relationships between features can be automatically learned, thereby accurately determining the fire risk propagation matrix and improving the accuracy of risk propagation probability calculations. Based on the risk propagation probability, priority can be provided for fire resource dispatch, prioritizing higher-risk areas and thus improving firefighting efficiency.

[0074] S140, Based on the risk propagation matrix and the acquired resource data, generate alternative resource allocation strategies and minimize the total loss of the alternative strategies to generate resource allocation strategies.

[0075] This embodiment uses a risk propagation matrix to assess the probability and intensity of risk propagation in different buildings or areas, thereby determining which areas require more resources to address potential risks. Combining acquired resource data, such as the quantity and status of fire trucks, drones, and rescue teams, multiple alternative resource allocation strategies are generated. These strategies consider different resource allocation schemes to ensure a rapid and effective response when a risk occurs. These alternative strategies are evaluated, and the total loss for each strategy is calculated, including aspects such as response time or resource consumption. By minimizing the total loss, the optimal resource allocation strategy is selected. This optimal strategy is used as the final resource allocation strategy to guide actual resource scheduling and allocation, ensuring a rapid and effective response when a risk occurs and minimizing losses.

[0076] In one embodiment of this application, based on the risk propagation matrix and the acquired resource data, alternative resource allocation strategies are generated, and the total loss of the alternative strategies is minimized to generate a resource allocation strategy, including:

[0077] Based on the risk propagation matrix and the acquired resource data, alternative resource allocation strategies are generated.

[0078] Based on the risk parameters of the building at a set time in the risk propagation matrix, the risk loss is determined;

[0079] Based on the time-varying cost function of the resources in the alternative strategies at each time, the resource mobilization cost is determined.

[0080] Based on the time it takes for resources to reach the building in the alternative strategies, determine the timeout penalty;

[0081] The total loss is calculated based on the aforementioned risk losses, the aforementioned resource mobilization costs, and the aforementioned timeout penalties.

[0082] By adjusting the alternative strategies and minimizing the total loss, the final resource allocation strategy is output.

[0083] In one embodiment of this application, alternative resource allocation strategies are generated based on the risk propagation matrix and the acquired resource data. Optionally, a deep reinforcement learning algorithm is used to generate alternative strategies, which are mappings from time to a set of resources, determining when and where to deploy which resources.

[0084] For example, in the process of generating alternative policies based on reinforcement learning, the current resource allocation state and the environmental state (such as the risk propagation state) are represented as state vectors. Based on the current state and the action value function table, the action with the maximum action value function is selected, i.e., the resource allocation policy. A reward function is designed to reflect the performance of the resource allocation policy, such as minimizing the total loss. The action value function table is updated based on the reward after executing the action and the new state. The above process is repeated until the action value function table converges or the maximum number of iterations is reached, thus obtaining the alternative policies.

[0085] In one embodiment of this application, based on the alternative strategies, the total loss consisting of risk loss, resource mobilization cost, and timeout penalty is determined. Specifically, in this process, the risk loss is determined based on the risk parameters built at a set time in the risk propagation matrix. for:

[0086]

[0087] Where t and T represent time and the total duration of the strategy prediction, respectively. This represents the preset discount factor. Let N represent the risk parameters of building i in the risk propagation matrix at time t, where i and N represent the identifier and total number of buildings in the risk propagation matrix, respectively. This indicates the indicator function, representing whether building i is not covered at time t. If building i is not protected by any resource at time t, then... ;otherwise, .

[0088] In one embodiment of this application, the resource mobilization cost is determined based on the time-varying cost function of the resources in the alternative strategies at each time. for:

[0089]

[0090] Where k and K represent the identifier and total number of fire-fighting resources, respectively. Representing resources The time-varying cost function at time t; It is an indicator function, representing the resource at time t. Is it active? If the resource is active... If time t is being used (e.g., a fire truck is on its way to a fire), then ;otherwise .

[0091] In one embodiment of this application, a timeout penalty is determined based on the time it takes for resources in the alternative strategies to reach the building. for:

[0092]

[0093] in, This represents the time it takes for resources to arrive at building i, that is, the time interval from the occurrence of the fire to the first arrival of resources at building i; This represents the response time threshold, indicating the maximum permissible time interval from the occurrence of a fire to the first arrival of resources at building i. If the time for resources to arrive at building i exceeds [a certain threshold], [the response time threshold will be lowered]. If this is done, the building will be punished.

[0094] In one embodiment of this application, based on the alternative strategies, the total loss is determined by summing the risk loss, resource mobilization cost, and timeout penalty. By adjusting the alternative strategies, the total loss is minimized, and the final resource allocation strategy is output. Here, risk loss considers the risk loss caused by uncovered buildings at different time points; resource mobilization cost calculates the activity cost of all resources at different time points; and the timeout penalty penalizes situations where the time for resources to reach a building exceeds a response time threshold.

[0095] The above process, based on the risk propagation matrix and acquired resource data, generates multiple alternative resource allocation strategies and selects the optimal strategy by minimizing the total loss. Reinforcement learning algorithms can adapt to dynamically changing environments, such as fire propagation and resource consumption, thereby generating more robust and flexible scheduling schemes. This effectively reduces accumulated risk losses, resource mobilization costs, and timeout penalties, improving the efficiency and effectiveness of resource scheduling. It can generate the optimal resource allocation strategy while meeting scheduling cost constraints and response time thresholds, ensuring the efficiency and effectiveness of resource allocation and enabling resources to achieve maximum utility in the shortest possible time.

[0096] S150 determines the scheduling strategy through Nash equilibrium based on the resource allocation strategy of adjacent regions and the preset inter-regional cooperation protocol.

[0097] In this embodiment, resource allocation strategies formulated by neighboring regions based on their own risk profiles and resources are obtained. These strategies reflect the resource deployment tendencies of each region under independent decision-making. Simultaneously, a pre-defined inter-regional cooperation agreement is incorporated, which clarifies rules such as resource sharing, mutual assistance mechanisms, and conflict resolution principles, providing a framework for cross-regional collaboration. The multi-regional resource allocation problem is modeled as a game scenario, where each region is considered a rational participant aiming to maximize its own utility. By defining a utility function that comprehensively considers local losses, cross-regional cooperation benefits, and resource conflict penalties, the gains of each region under different strategy combinations are quantitatively evaluated. Using the concept of Nash equilibrium in game theory, a stable scheduling strategy is iteratively solved. This strategy ensures that, given the strategies of other regions, no region can obtain higher gains by unilaterally adjusting its strategy, thus achieving a balance between global optimality and individual regional rationality.

[0098] In one embodiment of this application, a scheduling strategy is determined through Nash equilibrium based on the resource allocation strategy of adjacent regions and a preset inter-regional cooperation protocol, including:

[0099] Obtain resource allocation strategies and preset inter-regional cooperation protocols from adjacent regions;

[0100] The utility function is determined based on the resource allocation strategy of adjacent regions and the pre-set inter-regional cooperation agreement.

[0101] The scheduling strategy is determined by iterating the utility function using the alternating direction multiplier method.

[0102] In one embodiment of this application, resource allocation strategies for adjacent regions are obtained; these strategies describe how other regions allocate their resources. Simultaneously, an inter-regional cooperation protocol A is obtained, which defines how regions cooperate and share resources.

[0103] In one embodiment of this application, a utility function is determined based on the resource allocation strategies of adjacent regions and a preset inter-regional cooperation protocol. In this process, for each region i, its utility function is calculated according to a given strategy (the resource allocation strategy of this region) and other region strategies (the set of resource allocation strategies for other regions besides region i). for:

[0104]

[0105] in, Let represent the utility function of region i, where This refers to the resource allocation strategy for region i. It is the set of resource allocation strategies for regions other than region i; The weighting coefficients representing the local loss; Represents the local loss of region i, and represents the loss suffered by region i itself during resource scheduling. The weighting coefficients representing the benefits of cross-regional collaboration. This indicates the benefits of cross-regional collaboration. This represents the weighting coefficient of the resource conflict penalty term. This represents the conflict penalty for resources being accessed by multiple regions. The utility function measures the benefits for each region under different strategy combinations. It comprehensively considers the local loss of region i, the benefits of cross-region collaboration, and the conflict penalty for resources being accessed by multiple regions.

[0106] In one embodiment of this application, the utility function is iterated using the alternating direction multiplier method to determine the scheduling strategy. The Nash equilibrium strategy is obtained by iteratively solving the utility function using the alternating direction multiplier method. The Nash equilibrium strategy satisfies the resource allocation strategy for each region. And its corresponding other regional strategies are fixed as When, make the utility function The strategy that reaches the maximum value ultimately achieves strategy optimization.

[0107] A Nash equilibrium strategy is a strategy combination in which no participant can improve their own payoff by unilaterally changing their strategy, given that the strategies of other participants are fixed. This strategy achieves globally optimal resource allocation because no region has an incentive to deviate from it. By solving for the Nash equilibrium strategy, optimal global resource allocation can be achieved, improving overall resource utilization efficiency and system performance. This helps in more effectively scheduling and allocating resources and reducing losses in emergencies such as fires.

[0108] S160, based on the preset fire protection system, determine the target terminal related to the dispatch strategy, and send the dispatch instructions generated by the dispatch strategy to the target terminal.

[0109] In this embodiment, based on the generated scheduling strategy, the specific details of the fire situation are comprehensively analyzed, such as the size of the fire, the direction of spread, and the affected area. Simultaneously, the target terminal's attribute information is combined, such as the type of fire-fighting equipment (fire truck, drone, fire-fighting robot, etc.), functional characteristics (firefighting, rescue, reconnaissance, etc.), current status (location, resource reserves, availability, etc.), and historical task execution status, to accurately determine the most suitable target terminal for performing the current scheduling task. After determining the target terminal, its stable and reliable internal communication network is used to send detailed scheduling instructions generated by the scheduling strategy, including the specific content of the task (firefighting, rescue, alert, etc.), action route, required resources, expected completion time, and other key information, to the target terminal in real time and accurately.

[0110] During transmission, technologies such as encrypted transmission and data verification can be employed to ensure the security and integrity of dispatch instructions, preventing tampering or loss. Once the target terminal receives the dispatch instruction, it will immediately initiate the corresponding task execution process to respond quickly to the fire, thereby ensuring that the fire protection system can respond to the fire rapidly and effectively, minimizing the losses caused by the fire.

[0111] In one embodiment of this application, based on a preset fire protection system, a target terminal related to the dispatch strategy is determined, and the dispatch instructions generated by the dispatch strategy are sent to the target terminal, including:

[0112] Based on the preset fire protection system, the target terminals related to the dispatch strategy are determined;

[0113] Based on the scheduling strategy and the target terminal, a scheduling instruction is generated;

[0114] The scheduling instruction is sent to the target terminal.

[0115] In one embodiment of this application, target terminals that need to perform specific tasks are identified based on the generated scheduling strategy. These target terminals may be fire trucks, drones, rescue teams, or other firefighting equipment. By comprehensively considering the location, scale, and development trend of the fire, as well as the current status of each terminal (such as location, resource reserves, and response time), the most suitable target terminal for performing the current task is accurately matched.

[0116] After identifying the target terminal, detailed dispatch instructions are generated based on the dispatch strategy and the characteristics of the target terminal. These instructions include, but are not limited to: the specific content of the task (such as firefighting, rescue, and alert), the route of action, the required resources (such as water, foam, and rescue equipment), and the expected completion time. The generation of dispatch instructions aims to ensure that the target terminal can clearly and accurately understand the task requirements and respond rapidly accordingly.

[0117] The generated dispatch instructions are sent to the target terminal in real time via the communication network. This ensures secure and reliable instruction transmission and minimizes transmission delays. Upon receiving the dispatch instructions, the target terminal initiates the corresponding task execution process, such as fire trucks heading to the fire scene, drones taking off for reconnaissance, or rescue teams assembling and departing. This ensures the fire protection system can respond to the fire quickly and effectively, minimizing fire-related losses.

[0118] The above process, based on a pre-set fire protection system, identifies target terminals related to the dispatch strategy and sends dispatch instructions to these terminals. This ensures the timeliness and accuracy of resource dispatch. The identification of target terminals takes into account their current status and task execution capabilities, while the sending of dispatch instructions ensures that the terminals can respond and execute dispatch tasks in a timely manner. This achieves rapid and effective dispatch of fire protection resources, ultimately optimizing the strategy.

[0119] The technical solution of this application acquires first data through intelligent sensors, performs multi-source data fusion on the first data, and generates a first tensor composed of a spatial map through neural network and modeling processing. Node features and edge features are extracted from the first tensor and the spatial map. Based on a learnable weight matrix and activation function, the node features and edge features are linearly transformed to determine the fire risk propagation matrix. Based on the risk propagation matrix and acquired resource data, alternative resource allocation strategies are generated, and the total loss of the alternative strategies is minimized to generate a resource allocation strategy. Based on the resource allocation strategies of adjacent areas and a preset inter-regional cooperation protocol, a scheduling strategy is determined through Nash equilibrium. Based on a preset fire protection system, target terminals related to the scheduling strategy are identified, and the scheduling instructions generated by the scheduling strategy are sent to the target terminals. The above process achieves accurate risk prediction through multi-source data fusion and spatiotemporal modeling. Combined with dynamic resource allocation and cross-regional collaborative scheduling, it constructs a closed-loop system from data perception to instruction execution, effectively reducing the risk of fire spread and emergency response costs, and improving the efficiency and effectiveness of fire resource scheduling.

[0120] The following describes embodiments of the AI-driven resource scheduling system for smart fire protection according to this application, which can be used to execute the AI-driven resource scheduling method for smart fire protection in the above embodiments of this application. It is understood that the AI-driven resource scheduling system for smart fire protection can be a computer program (including program code) running on a computer device, for example, the AI-driven resource scheduling system for smart fire protection is an application software; the AI-driven resource scheduling system for smart fire protection can be used to execute the corresponding steps in the method provided in the embodiments of this application. For details not disclosed in the embodiments of the AI-driven resource scheduling system for smart fire protection of this application, please refer to the embodiments of the AI-driven resource scheduling method for smart fire protection described above in this application.

[0121] Figure 3 A block diagram of an AI-driven resource scheduling system for smart fire protection according to an embodiment of this application is shown.

[0122] Reference Figure 3 As shown, an AI-driven resource scheduling system for smart fire protection according to an embodiment of this application includes:

[0123] The acquisition unit 310 is used to acquire first data through a smart sensor, wherein the first data includes environmental data and satellite data;

[0124] The coupling unit 320 is used to perform multi-source data fusion on the first data and generate a first tensor composed of a spatial map through neural network and modeling processing.

[0125] Feature unit 330 is used to extract node features and edge features from the first tensor and spatial map, and to perform linear transformation on the node features and edge features based on a learnable weight matrix and activation function to determine the risk propagation matrix of the fire.

[0126] Strategy unit 340 is used to generate alternative resource allocation strategies based on the risk propagation matrix and the acquired resource data, and to minimize the total loss of the alternative strategies, so as to generate a resource allocation strategy.

[0127] The scheduling unit 350 is used to determine the scheduling strategy through Nash equilibrium based on the resource allocation strategy of adjacent areas and the preset inter-regional cooperation protocol.

[0128] The instruction unit 360 is used to determine the target terminal related to the dispatch strategy based on the preset fire protection system, and send the dispatch instructions generated by the dispatch strategy to the target terminal.

[0129] In this application, based on the aforementioned scheme, the step of performing multi-source data fusion on the first data and generating a first tensor composed of a spatial map through neural network and modeling processing includes: performing three-dimensional convolution processing on the first data based on a preset spatiotemporal convolution kernel to generate second data; generating dynamic weight coefficients based on an attention mechanism, and processing the second data through the dynamic weight coefficients to generate third data; modeling the satellite data in the first data to generate a geographic topology, and processing the geographic topology through a graph neural network to generate a spatial map; and generating a first tensor composed of the spatial map based on the third data and the spatial map.

[0130] In this application, based on the aforementioned scheme, the step of extracting node features and edge features from the first tensor and the spatial graph, and performing a linear transformation on the node features and edge features based on a learnable weight matrix and an activation function to determine the fire risk propagation matrix includes: aggregating information of neighboring nodes in the spatial graph based on a preset graph attention network to generate node features; extracting edge features between neighboring nodes from the first tensor and the spatial graph, and concatenating the node features and the edge features to generate a first feature; performing a linear transformation on the first feature based on a first learnable weight matrix and a first activation function to generate a second feature; and performing a linear transformation on the second feature based on a second learnable weight matrix and a second activation function to obtain the fire risk propagation matrix.

[0131] In this application, based on the aforementioned scheme, the step of generating alternative resource allocation strategies based on the risk propagation matrix and the acquired resource data, and minimizing the total loss of the alternative strategies to generate a resource allocation strategy, includes: generating alternative resource allocation strategies based on the risk propagation matrix and the acquired resource data; determining the risk loss based on the risk parameters of the building at a set time in the risk propagation matrix; determining the resource mobilization cost based on the time-varying cost function of the resources in the alternative strategies at each time; determining the timeout penalty based on the time of resource arrival at the building in the alternative strategies; constructing the total loss based on the risk loss, the resource mobilization cost, and the timeout penalty; minimizing the total loss by adjusting the alternative strategies, and outputting the final resource allocation strategy.

[0132] In this application, based on the aforementioned scheme, determining the risk loss based on the risk parameters of the building at a set time in the risk propagation matrix includes: determining the risk loss based on the risk parameters of the building at a set time in the risk propagation matrix. for:

[0133]

[0134] Where t and T represent time and the total duration of the strategy prediction, respectively. This represents the preset discount factor. Let N represent the risk parameters of building i in the risk propagation matrix at time t, where i and N represent the identifier and total number of buildings in the risk propagation matrix, respectively. This indicates an indicator function.

[0135] In this application, based on the aforementioned scheme, the step of determining the scheduling strategy through Nash equilibrium based on the resource allocation strategy of adjacent regions and the preset inter-regional cooperation protocol includes: obtaining the resource allocation strategy of adjacent regions and the preset inter-regional cooperation protocol; determining the utility function based on the resource allocation strategy of adjacent regions and the preset inter-regional cooperation protocol; and iterating the utility function through the alternating direction multiplier method to determine the scheduling strategy.

[0136] In this application, based on the aforementioned scheme, the step of determining the target terminal related to the dispatch strategy based on a preset fire protection system and sending the dispatch instructions generated by the dispatch strategy to the target terminal includes: determining the target terminal related to the dispatch strategy based on a preset fire protection system; generating a dispatch instruction based on the dispatch strategy and the target terminal; and sending the dispatch instruction to the target terminal.

[0137] The technical solution of this application acquires first data through intelligent sensors, performs multi-source data fusion on the first data, and generates a first tensor composed of a spatial map through neural network and modeling processing. Node features and edge features are extracted from the first tensor and the spatial map. Based on a learnable weight matrix and activation function, the node features and edge features are linearly transformed to determine the fire risk propagation matrix. Based on the risk propagation matrix and acquired resource data, alternative resource allocation strategies are generated, and the total loss of the alternative strategies is minimized to generate a resource allocation strategy. Based on the resource allocation strategies of adjacent areas and a preset inter-regional cooperation protocol, a scheduling strategy is determined through Nash equilibrium. Based on a preset fire protection system, target terminals related to the scheduling strategy are identified, and the scheduling instructions generated by the scheduling strategy are sent to the target terminals. The above process achieves accurate risk prediction through multi-source data fusion and spatiotemporal modeling. Combined with dynamic resource allocation and cross-regional collaborative scheduling, it constructs a closed-loop system from data perception to instruction execution, effectively reducing the risk of fire spread and emergency response costs, and improving the efficiency and effectiveness of fire resource scheduling.

[0138] Figure 4 A schematic diagram of the structure of a computer system suitable for implementing the electronic device of the present application is shown.

[0139] It should be noted that the computer system of the electronic device in this embodiment is only an example and should not impose any limitations on the function and scope of use of the embodiments of this application.

[0140] In this embodiment, the computer system includes a central processing unit 401, which can perform various appropriate actions and processes based on programs stored in read-only memory 402 or programs loaded from storage section 408 into random access memory 403, such as executing the AI-driven resource scheduling method for smart fire protection described in the above embodiment. The random access memory 403 also stores various programs and data required for system operation. The central processing unit 401, read-only memory 402, and random access memory 403 are interconnected via bus 404. Input / output interface 405 is also connected to bus 404.

[0141] The following components are connected to the input / output interface 405: an input section 406 including a keyboard, mouse, etc.; an output section 407 including a cathode ray tube (CRT), liquid crystal display (LCD), etc., and speakers, etc.; a storage section 408 including a hard disk, etc.; and a communication section 409 including a network interface card such as a LAN (Local Area Network) card, modem, etc. The communication section 409 performs communication processing via a network such as the Internet. A drive 410 is also connected to the input / output interface 405 as needed. A removable medium 411, such as a disk, optical disk, magneto-optical disk, semiconductor memory, etc., is installed on the drive 410 as needed so that computer programs read from it can be installed into the storage section 408 as needed.

[0142] Specifically, according to embodiments of this application, the processes described above with reference to the flowcharts can be implemented as computer software programs. For example, embodiments of this application include a computer program product comprising a computer program carried on a computer-readable medium, the computer program including a computer program for performing the methods shown in the flowcharts. In such embodiments, the computer program can be downloaded and installed from a network via communication section 409, and / or installed from removable medium 411. When the computer program is executed by central processing unit 401, it performs various functions defined in the system of this application.

[0143] It should be noted that the computer-readable medium shown in the embodiments of this application can be a computer-readable signal medium or a computer-readable storage medium, or any combination of the two. A computer-readable storage medium can be, for example, but not limited to, an electrical, magnetic, optical, electromagnetic, infrared, or semiconductor system, apparatus, or device, or any combination thereof. More specific examples of a computer-readable storage medium may include, but are not limited to: an electrical connection having one or more wires, a portable computer disk, a hard disk, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM), flash memory, optical fiber, portable compact disc read-only memory (CD-ROM), optical storage device, magnetic storage device, or any suitable combination thereof. In this application, a computer-readable storage medium can be any tangible medium containing or storing a program that can be used by or in conjunction with an instruction execution system, apparatus, or device. In this application, a computer-readable signal medium can include a data signal propagated in baseband or as part of a carrier wave, carrying a computer-readable computer program. The transmitted data signal can take various forms, including but not limited to electromagnetic signals, optical signals, or any suitable combination thereof. The computer-readable signal medium can also be any computer-readable medium other than a computer-readable storage medium, which can send, propagate, or transmit a program for use by or in connection with an instruction execution system, apparatus, or device. The computer program contained on the computer-readable medium can be transmitted using any suitable medium, including but not limited to wireless, wired, etc., or any suitable combination thereof.

[0144] The flowcharts and block diagrams in the accompanying drawings illustrate the architecture, functionality, and operation of possible implementations of systems, methods, and computer program products according to various embodiments of this application. Each block in a flowchart or block diagram may represent a module, segment, or portion of code, which contains one or more executable instructions for implementing a specified logical function. It should also be noted that in some alternative implementations, the functions indicated in the blocks may occur in a different order than those indicated in the drawings. For example, two consecutively indicated blocks may actually be executed substantially in parallel, and they may sometimes be executed in reverse order, depending on the functions involved. It should also be noted that each block in a block diagram or flowchart, and combinations of blocks in a block diagram or flowchart, can be implemented using a dedicated hardware-based system that performs the specified function or operation, or using a combination of dedicated hardware and computer instructions.

[0145] The units described in the embodiments of this application can be implemented in software or hardware, and the described units can also be located in a processor. The names of these units do not necessarily limit the specific unit itself.

[0146] According to one aspect of this application, a computer program product or computer program is provided, comprising computer instructions stored in a computer-readable storage medium. A processor of a computer device reads the computer instructions from the computer-readable storage medium and executes the computer instructions, causing the computer device to perform the methods provided in the various alternative implementations described above.

[0147] In another aspect, this application also provides a computer-readable medium, which may be included in the electronic device described in the above embodiments; or it may exist independently and not assembled into the electronic device. The computer-readable medium carries one or more programs, which, when executed by the electronic device, cause the electronic device to implement the AI-driven resource scheduling method for smart fire protection described in the above embodiments.

[0148] It should be noted that although several modules or units for the device used to perform actions have been mentioned in the detailed description above, this division is not mandatory. In fact, according to the embodiments of this application, the features and functions of two or more modules or units described above can be embodied in one module or unit. Conversely, the features and functions of one module or unit described above can be further divided and embodied by multiple modules or units.

[0149] Through the above description of the embodiments, those skilled in the art will readily understand that the exemplary embodiments described herein can be implemented by software or by combining software with necessary hardware. Therefore, the technical solutions according to the embodiments of this application can be embodied in the form of a software product, which can be stored in a non-volatile storage medium (such as a CD-ROM, USB flash drive, external hard drive, etc.) or on a network, including several instructions to cause a computing device (such as a personal computer, server, touch terminal, or network device, etc.) to execute the methods according to the embodiments of this application.

[0150] Other embodiments of this application will readily occur to those skilled in the art upon consideration of the specification and practice of the embodiments disclosed herein. This application is intended to cover any variations, uses, or adaptations of this application that follow the general principles of this application and include common knowledge or customary techniques in the art not disclosed herein.

[0151] It should be understood that this application is not limited to the precise structure described above and shown in the accompanying drawings, and various modifications and changes can be made without departing from its scope. The scope of this application is limited only by the appended claims.

Claims

1. An AI-driven resource scheduling method for smart fire protection, characterized in that, include: First data is acquired through smart sensors, wherein the first data includes environmental data and satellite data; The first data is fused from multiple sources, and through neural networks and modeling, a first tensor consisting of a spatial map is generated. Node features and edge features are extracted from the first tensor and spatial graph. The node features and edge features are then linearly transformed based on the learnable weight matrix and activation function to determine the risk propagation matrix of the fire. Based on the risk propagation matrix and the acquired resource data, alternative resource allocation strategies are generated, and the total loss of the alternative strategies is minimized to generate a resource allocation strategy. Based on the resource allocation strategy of adjacent regions and the preset inter-regional cooperation protocol, the scheduling strategy is determined through Nash equilibrium. Based on the preset fire protection system, a target terminal related to the dispatch strategy is identified, and the dispatch instructions generated by the dispatch strategy are sent to the target terminal.

2. The AI-driven resource scheduling method for smart fire protection according to claim 1, characterized in that, The first data is subjected to multi-source data fusion, and through neural network and modeling processing, a first tensor composed of spatial maps is generated, including: Based on a preset spatiotemporal convolution kernel, the first data is subjected to three-dimensional convolution processing to generate the second data; Dynamic weight coefficients are generated based on an attention mechanism, and the second data is processed using these dynamic weight coefficients to generate the third data. The satellite data in the first data is modeled to generate a geographic topology, and the geographic topology is processed through a graph neural network to generate a spatial map; Based on the third data and the spatial map, a first tensor composed of the spatial map is generated.

3. The AI-driven resource scheduling method for smart fire protection according to claim 1, characterized in that, Node features and edge features are extracted from the first tensor and spatial graph. A linear transformation is then performed on the node features and edge features based on a learnable weight matrix and activation function to determine the fire risk propagation matrix, including: Based on a preset graph attention network, information of neighboring nodes in the spatial graph is aggregated to generate node features; The edge features between the neighboring nodes are extracted from the first tensor and the spatial graph, and the node features and the edge features are concatenated to generate the first feature; The first feature is linearly transformed based on the first learnable weight matrix and the first activation function to generate the second feature; The risk propagation matrix of the fire is obtained by linearly transforming the second feature based on the second learnable weight matrix and the second activation function.

4. The AI-driven resource scheduling method for smart fire protection according to claim 1, characterized in that, Based on the risk propagation matrix and the acquired resource data, alternative resource allocation strategies are generated, and the total loss of the alternative strategies is minimized to generate a resource allocation strategy, including: Based on the risk propagation matrix and the acquired resource data, alternative resource allocation strategies are generated. Based on the risk parameters of the building at a set time in the risk propagation matrix, the risk loss is determined; Based on the time-varying cost function of the resources in the alternative strategies at each time, the resource mobilization cost is determined. Based on the time it takes for resources to reach the building in the alternative strategies, determine the timeout penalty; The total loss is calculated based on the aforementioned risk losses, the aforementioned resource mobilization costs, and the aforementioned timeout penalties. By adjusting the alternative strategies and minimizing the total loss, the final resource allocation strategy is output.

5. The AI-driven resource scheduling method for smart fire protection according to claim 4, characterized in that, Based on the risk parameters of the building at a set time in the risk propagation matrix, the risk loss is determined, including: Based on the risk parameters of the building at a given time in the risk propagation matrix, determine the risk loss. for: ； Where t and T represent time and the total duration of the strategy prediction, respectively. This represents the preset discount factor. Let N represent the risk parameters of building i in the risk propagation matrix at time t, where i and N represent the identifier and total number of buildings in the risk propagation matrix, respectively. Indicates an indicator function.

6. The AI-driven resource scheduling method for smart fire protection according to claim 1, characterized in that, Based on the resource allocation strategy of adjacent regions and the preset inter-regional cooperation protocol, the scheduling strategy is determined through Nash equilibrium, including: Obtain resource allocation strategies and preset inter-regional cooperation protocols from adjacent regions; The utility function is determined based on the resource allocation strategy of adjacent regions and the preset inter-regional cooperation agreement. The scheduling strategy is determined by iterating the utility function using the alternating direction multiplier method.

7. The AI-driven resource scheduling method for smart fire protection according to claim 1, characterized in that, Based on a preset fire protection system, a target terminal related to the dispatch strategy is identified, and the dispatch instructions generated by the dispatch strategy are sent to the target terminal, including: Based on the preset fire protection system, the target terminals related to the dispatch strategy are determined; Based on the scheduling strategy and the target terminal, a scheduling instruction is generated; The scheduling instruction is sent to the target terminal.

8. An AI-driven resource scheduling system for smart fire protection, characterized in that, include: An acquisition unit is configured to acquire first data via a smart sensor, wherein the first data includes environmental data and satellite data; The coupling unit is used to perform multi-source data fusion on the first data and generate a first tensor composed of spatial maps through neural network and modeling processing. The feature unit is used to extract node features and edge features from the first tensor and spatial graph, and to perform a linear transformation on the node features and edge features based on a learnable weight matrix and activation function to determine the risk propagation matrix of the fire. The strategy unit is used to generate alternative resource allocation strategies based on the risk propagation matrix and the acquired resource data, and to minimize the total loss of the alternative strategies, so as to generate a resource allocation strategy. The scheduling unit is used to determine the scheduling strategy through Nash equilibrium based on the resource allocation strategy of adjacent regions and the preset inter-regional cooperation protocol. The instruction unit is used to determine the target terminal related to the dispatch strategy based on the preset fire protection system, and send the dispatch instructions generated by the dispatch strategy to the target terminal.

9. The AI-driven resource scheduling system for smart fire protection according to claim 7, characterized in that, The first data is subjected to multi-source data fusion, and through neural network and modeling processing, a first tensor composed of spatial maps is generated, including: Based on a preset spatiotemporal convolution kernel, the first data is subjected to three-dimensional convolution processing to generate the second data; Dynamic weight coefficients are generated based on an attention mechanism, and the second data is processed using these dynamic weight coefficients to generate the third data. The satellite data in the first data is modeled to generate a geographic topology, and the geographic topology is processed through a graph neural network to generate a spatial map; Based on the third data and the spatial map, a first tensor composed of the spatial map is generated.

10. The AI-driven resource scheduling system for smart fire protection according to claim 7, characterized in that, Node features and edge features are extracted from the first tensor and spatial graph. A linear transformation is then performed on the node features and edge features based on a learnable weight matrix and activation function to determine the fire risk propagation matrix, including: Based on a preset graph attention network, information of neighboring nodes in the spatial graph is aggregated to generate node features; The edge features between the neighboring nodes are extracted from the first tensor and the spatial graph, and the node features and the edge features are concatenated to generate the first feature; The first feature is linearly transformed based on the first learnable weight matrix and the first activation function to generate the second feature; The risk propagation matrix of the fire is obtained by linearly transforming the second feature based on the second learnable weight matrix and the second activation function.

Images