An emergency communication resource dynamic allocation method and system based on artificial intelligence

Abstract

The application relates to an emergency communication resource dynamic allocation method and system based on artificial intelligence, and relates to the technical field of emergency communication and intelligent resource scheduling. The method comprises the following steps: collecting and fusing multi-source heterogeneous data in an emergency scene to generate comprehensive situation awareness information; constructing a multi-target resource allocation optimization model based on the information, and solving the model by using a multi-target evolutionary algorithm to obtain a preliminary resource allocation strategy; using a pre-trained artificial intelligence model to predict the disaster evolution trend and communication demand change; receiving actual performance feedback data, adopting a composite trigger strategy, combining a fusion proportional-integral control mechanism, dynamically adjusting the weight coefficients of each optimization target, and generating a final resource scheduling instruction; converting the final instruction into a control command executable by equipment and issuing the control command to drive the communication resource to complete dynamic reconfiguration. The application realizes global optimization, active prediction and rapid self-adaptation of emergency communication resource scheduling, and improves the intelligent level and robustness of the system.

Description

Technical Field

[0001] This invention relates to the field of emergency communication and intelligent resource scheduling technology, specifically to an artificial intelligence-based method and system for dynamic allocation of emergency communication resources. Background Technology

[0002] Emergency communication systems play a crucial role in information transmission and command and dispatch during sudden events such as natural disasters and accidents. Current emergency communication resource allocation schemes generally suffer from the following shortcomings: resource allocation decisions primarily rely on pre-set static resource configuration strategies or local information from a single network access point, lacking global situational awareness and deep fusion capabilities for multi-source heterogeneous environmental data at disaster sites; algorithms often employ fixed-weight or rule-driven allocation mechanisms, making it difficult to achieve adaptive trade-offs and global optimization among conflicting objectives such as bandwidth, latency, energy consumption, and coverage; the system lacks a closed-loop feedback mechanism for communication resource status, failing to adjust allocation strategies in real time based on actual task execution results. This leads to inefficiencies if initial weights are improperly set or the environment changes abruptly, resulting in low resource utilization or insufficient support for critical services; furthermore, existing architectures generally lack the predictive and reasoning capabilities of artificial intelligence, failing to proactively predict disaster evolution trends and thus hindering advance deployment of communication resources, resulting in delayed responses and resource misallocation. These deficiencies make it difficult for existing technologies to meet the stringent requirements of accuracy, foresight, and adaptability in resource scheduling during large-scale, highly dynamic, and severely damaged extreme emergency scenarios. Summary of the Invention

[0003] The technical problem to be solved by this invention is to provide a dynamic allocation scheme for emergency communication resources with global perception, forward prediction and closed-loop feedback. In particular, it solves the core problems of poor adaptability and insufficient robustness of existing technologies by using an innovative event-performance dual-driven composite triggering strategy and a proportional-integral control mechanism with feedforward compensation.

[0004] The technical solution of this invention to solve the above-mentioned technical problems is as follows: A dynamic allocation method for emergency communication resources based on artificial intelligence, comprising the following steps: S1, collecting multi-source heterogeneous data in emergency scenarios and fusing the multi-source heterogeneous data to generate comprehensive situational awareness information; S2, based on the comprehensive situational awareness information, constructing a resource allocation optimization model including multiple priority objectives, and solving the resource allocation optimization model using a multi-objective evolutionary algorithm to obtain a preliminary resource allocation strategy; S3, using a pre-trained artificial intelligence model to predict the evolution trend of the disaster and changes in communication demand, and performing a forward-looking performance evaluation and strategy correction on the preliminary resource allocation strategy based on the prediction results; S4, collecting key performance indicators of the scheduled communication network through a dedicated monitoring interface. For actual performance feedback data, the deviation between the actual performance feedback data and the preset expected target is calculated. A composite triggering strategy based on event-performance dual drive is adopted, and combined with a proportional-integral control mechanism with fused feedforward compensation, the weight coefficients of each optimization objective in the resource allocation optimization model are dynamically adjusted to generate an adaptively adjusted final resource scheduling instruction; S5, the final resource scheduling instruction is parsed into a specific sequence of device configuration parameters to generate a standard control command containing power value, frequency point number, timestamp, starting frequency point, bandwidth size, authorization period, target coordinates, travel route, deployment time, bandwidth ratio, priority queue and traffic shaping parameters. The standardized control command is sent to the network infrastructure through the corresponding communication protocol interface to drive the communication resources to complete dynamic reconfiguration.

[0005] The beneficial effects of this invention are as follows: By fusing multi-source heterogeneous data, comprehensive and accurate situational awareness is ensured, providing reliable support for decision-making; multi-objective optimization combined with evolutionary algorithms balances core needs and avoids resource misallocation; AI prediction is introduced to anticipate disasters and communication needs, allowing for early strategy correction and reducing response lag; innovative dual-drive triggering and feedforward compensation control dynamically adjusts weights, balancing emergency response and long-term efficiency, and improving allocation accuracy and timeliness; standardized command conversion and multi-protocol adaptation ensure universal and efficient resource reconfiguration, adapting to various network facilities.

[0006] Based on the above technical solution, the present invention can be further improved as follows.

[0007] Furthermore, S1 specifically includes: S101, synchronously acquiring geographical information of the disaster impact range, regional population density heat map, real-time base station load rate, wireless link quality indicators, and emergency service priority list through satellite remote sensing, ground sensor networks, and mobile terminal reporting mechanisms; S102, performing preprocessing operations such as denoising, normalization, and missing value imputation on the collected multi-source heterogeneous data; S103, using a multi-source information fusion unit based on confidence theory to fuse the preprocessed multi-source heterogeneous data to obtain the comprehensive situational awareness information.

[0008] The beneficial effect of adopting the above-mentioned further solution is that by using a multi-source information fusion unit based on confidence theory, it supports the replacement of multiple fusion algorithms, adapts to different data source types and emergency scenarios, and enhances the flexibility and versatility of the solution.

[0009] Further, S2 specifically includes: S201, constructing the resource allocation optimization model, wherein the optimization objectives include at least two of the following: minimizing total system energy consumption, minimizing average service latency, maximizing network coverage, and maximizing high-priority service guarantee rate; the decision variables include at least one of the following: base station transmit power adjustment amount, spectrum block allocation scheme, mobile communication device deployment location, and backhaul link bandwidth allocation ratio; the constraints include: total available spectrum resource limit, maximum transmit power limit of a single base station, minimum service quality requirement, and equipment physical location reachability constraint; S202, adopting a population-based multi-camera system. The standard evolutionary optimization unit solves the resource allocation optimization model and obtains the Pareto optimal solution set within a preset number of iterations. From the Pareto optimal solution set, based on the degree of disaster impact, communication demand level, and resource scarcity status reflected by the integrated situational awareness information, the solution with the highest degree of matching with the current integrated situational awareness information is selected as the preliminary resource allocation strategy. Each solution in the Pareto optimal solution set represents a resource allocation scheme, and the solutions do not dominate each other on multiple optimization objectives. Each solution cannot be a compromise solution that improves any optimization objective without harming other optimization objectives.

[0010] The advantages of adopting the above-mentioned further scheme are that it supports the replacement of multiple multi-objective evolutionary algorithms, adapts to the solution requirements of different emergency scenarios, improves the efficiency and accuracy of model solution, and achieves the balance of multi-dimensional objectives through the selection of Pareto optimal solution set and compromise solution, ensuring that the resource allocation strategy meets the real-time situation requirements and achieves global optimum.

[0011] Further, S3 specifically includes: S301, using a prediction unit based on time-series feature extraction, based on historical disaster data and real-time situation data, predicting the spread trend of the disaster area, changes in the number of affected people, and time-series fluctuations in communication service request volume in different regions within a future preset time window; S302, using a prediction unit based on network topology analysis, modeling the communication network topology, and predicting the impact of network connectivity changes caused by the disaster on the stability of the resource allocation strategy; S303, fusing the outputs of S301 and S302, inputting them into a comprehensive inference network for comprehensive inference, and outputting an effectiveness evaluation score and correction suggestions for the preliminary resource allocation strategy under the future situation; S304, setting a preset effectiveness score as... The current performance score is Correction factor Determined based on forecast uncertainty, The revised strategy score is then expressed as: ,when If the value falls below a preset threshold, a correction to the initial resource allocation strategy will be triggered.

[0012] The beneficial effects of adopting the above-mentioned further scheme are that by using a prediction unit based on network topology analysis, the impact of network connectivity changes on resource allocation strategies can be predicted, thereby improving the stability of the strategies; the feature fusion and comprehensive reasoning of the two prediction units can achieve complementarity of multi-dimensional prediction information, thereby improving the accuracy and comprehensiveness of prediction.

[0013] Furthermore, the event-performance dual-driven composite triggering strategy in S4 specifically includes: S401, continuously monitoring key performance indicators of the scheduled communication network, including actual service throughput, end-to-end latency, call drop rate, and resource utilization; S402, receiving major disaster change prediction information from S3 and receiving resource reconfiguration failure event information from the execution layer; S403, when receiving the major disaster change prediction information or the resource reconfiguration failure information, triggering the event-driven weight correction mode; when receiving the major disaster change prediction information and the resource reconfiguration failure information, continuously monitoring the key performance indicators at a preset period, and when the cumulative deviation of the key performance indicators exceeds a preset steady-state threshold, triggering the performance-driven weight correction mode.

[0014] The beneficial effect of adopting the above-mentioned further solution is that by combining the dual triggering modes, it can not only respond quickly to emergencies and avoid policy lag, but also optimize the long-term operating efficiency of the system and achieve dynamic balance of resource allocation.

[0015] Furthermore, the proportional-integral control mechanism with integrated feedforward compensation specifically includes: in the event-driven weight correction mode, temporarily increasing the gain values ​​of the proportional control gain coefficient and the feedforward compensation gain coefficient to a preset multiple of the normal value, and temporarily freezing the cumulative effect of the historical deviation accumulation term; in the performance-driven weight correction mode, under the preset proportional control gain coefficient, integral control gain coefficient and feedforward compensation gain coefficient, continuously making gradual fine adjustments to the weights based on the deviation of key performance indicators.

[0016] The beneficial effect of adopting the above-mentioned further solutions is that by introducing feedforward compensation terms, we can proactively address sudden changes in business priorities and network infrastructure failures, avoid lag in weight adjustments, and improve the system's emergency response capabilities.

[0017] Furthermore, the dynamic adjustment of the weighting coefficients is based on the following formula: ;in, Let be the weight coefficient of the i-th target at time t. The weight coefficient of the i-th target mentioned in the previous time step is... Let be the deviation of the i-th target at time t. This is the proportional control gain coefficient. The integral control gain coefficient. Let be the cumulative sum of historical deviations of the i-th target from the initial time to the current time t. This represents the change in the emergency response priority vector compared to the previous time step. This is a real-time failure indication function for network infrastructure. For feedforward compensation gain coefficient, and the This item is used to strengthen the optimization weight of key business support objectives when the disaster situation escalates or the rescue phase transitions. This item is used to assign extra weight to remaining resources when part of the network infrastructure fails, in order to prioritize maintaining critical coverage and connectivity.

[0018] The beneficial effect of adopting the above-mentioned further solutions is that by ensuring the accuracy of weight adjustment, the resource allocation strategy can adapt to changes in the scenario in real time, thereby improving resource utilization and critical business assurance rate.

[0019] Furthermore, S5 specifically includes: S501, parsing the final resource scheduling instruction into a specific sequence of device configuration parameters, wherein the final resource scheduling instruction includes an abstract instruction for optimizing the target weight adjustment result and resource allocation strategy parameters, and the device configuration parameter sequence is directly executable configuration data corresponding to a specific communication device and communication protocol; S502, through an interface unit supporting multiple standard communication protocols, distributing the device configuration parameter sequence to the corresponding physical device or virtualized network function to complete spectrum reconfiguration, power adjustment, beamforming, and routing strategy update operations.

[0020] The beneficial effect of adopting the above-mentioned further solutions is to ensure the timeliness of resource reconfiguration, thereby ensuring that instruction execution and resource reconfiguration are completed within a preset time, thus meeting the needs of emergency scenarios.

[0021] Furthermore, this application also provides an artificial intelligence-based emergency communication resource dynamic allocation system, comprising: a multi-source heterogeneous data perception and fusion module, used to collect multi-source heterogeneous data in emergency scenarios and fuse the multi-source heterogeneous data to generate comprehensive situational awareness information; a dynamic multi-objective optimization decision-making module, used to construct a resource allocation optimization model including multiple priority objectives based on the comprehensive situational awareness information, and use a multi-objective evolutionary algorithm to solve the resource allocation optimization model to obtain a preliminary resource allocation strategy; an artificial intelligence prediction and reasoning module, used to use a pre-trained artificial intelligence model to predict the evolution trend of disaster and changes in communication demand, and to perform a forward-looking performance evaluation and strategy correction on the preliminary resource allocation strategy based on the prediction results; and a closed-loop feedback and strategy adaptive adjustment module, used to collect key performance indicators of the scheduled communication network as actual performance feedback data through a dedicated monitoring interface, and calculate the actual performance. The deviation between the feedback data and the preset expected target is addressed by employing a composite triggering strategy based on event-performance dual-drive, combined with a proportional-integral control mechanism with fused feedforward compensation. This dynamically adjusts the weight coefficients of each optimization objective in the resource allocation optimization model, generating an adaptively adjusted final resource scheduling instruction. The closed-loop feedback and strategy adaptive adjustment module incorporates the composite triggering strategy based on event-performance dual-drive and the proportional-integral control mechanism with fused feedforward compensation. The resource scheduling execution and interface module parses the final resource scheduling instruction into a specific sequence of device configuration parameters, generating a standard control command containing power value, frequency point number, timestamp, starting frequency point, bandwidth size, authorization period, target coordinates, travel route, deployment time, bandwidth ratio, priority queue, and traffic shaping parameters. This standardized control command is then sent to the network infrastructure through the corresponding communication protocol interface to drive the communication resources to complete dynamic reconfiguration.

[0022] Furthermore, the AI-based emergency communication resource dynamic allocation system adopts a cloud-edge-device collaborative architecture. The multi-source heterogeneous data perception and fusion module is deployed at the data acquisition front end to access satellite remote sensing data interfaces, ground sensor networks, and mobile terminal reporting channels. The dynamic multi-objective optimization decision-making module and the AI ​​prediction and inference module are deployed on edge computing nodes, which are set up in emergency command vehicles or temporary command centers to complete comprehensive situational fusion and optimization calculations locally. The closed-loop feedback and policy adaptive adjustment module is deployed on the edge computing nodes or cloud servers. The resource scheduling execution and interface module is deployed on the network infrastructure side to directly interface with physical communication equipment or virtualized network functions.

[0023] Furthermore, the emergency communication resource dynamic allocation system also includes a federated learning collaborative optimization layer, which is deployed on a cloud server or among multiple edge computing nodes to achieve cross-regional or cross-departmental model collaborative training and strategy optimization by exchanging encrypted model parameter updates.

[0024] Furthermore, this application also provides an electronic device, including: a processor; a memory for storing processor-executable instructions; wherein the processor is configured to execute the instructions to implement the aforementioned artificial intelligence-based emergency communication resource dynamic allocation method.

[0025] Furthermore, this application also provides a computer-readable storage medium storing at least one computer program, which is loaded and executed by a processor to enable a computer to implement the aforementioned artificial intelligence-based emergency communication resource dynamic allocation method. Attached Figure Description

[0026] Figure 1 This is a flowchart illustrating an artificial intelligence-based dynamic allocation method for emergency communication resources according to the present invention. Figure 2 This is a schematic diagram of the structure of an emergency communication resource dynamic allocation system based on artificial intelligence according to the present invention; Figure 3 This is a schematic diagram of the structure of an electronic device according to the present invention. Detailed Implementation

[0027] The principles and features of the present invention are described below. The examples given are only for explaining the present invention and are not intended to limit the scope of the present invention.

[0028] Example 1: The existing technology for emergency communication resource allocation has the following problems: On the one hand, it does not systematically integrate multi-source heterogeneous data, resulting in one-sided decision-making basis and failing to fully reflect the real-time situation of emergency scenarios; on the other hand, the weight coefficients of resource allocation optimization models are mostly fixed or adjusted only through single performance feedback, lacking the ability to respond quickly to events such as sudden disasters and resource reconfiguration failures.

[0029] Reference Figure 1This embodiment relates to an artificial intelligence-based method for dynamic allocation of emergency communication resources, comprising the following steps: S1, collecting multi-source heterogeneous data in emergency scenarios and fusing the multi-source heterogeneous data to generate comprehensive situational awareness information; S2, based on the comprehensive situational awareness information, constructing a resource allocation optimization model including multiple priority objectives, and solving the resource allocation optimization model using a multi-objective evolutionary algorithm to obtain a preliminary resource allocation strategy; S3, using a pre-trained artificial intelligence model to predict the evolution trend of the disaster situation and changes in communication demand, and performing a forward-looking performance evaluation and strategy correction on the preliminary resource allocation strategy based on the prediction results; S4, collecting key performance indicators of the scheduled communication network through a dedicated monitoring interface as actual performance indicators. Feedback data is used to calculate the deviation between the actual performance feedback data and the preset expected target. A composite triggering strategy based on event-performance dual drive is adopted, combined with a proportional-integral control mechanism with fused feedforward compensation, to dynamically adjust the weight coefficients of each optimization target in the resource allocation optimization model and generate an adaptively adjusted final resource scheduling instruction; S5, the final resource scheduling instruction is parsed into a specific sequence of device configuration parameters to generate a standard control command containing power value, frequency point number, timestamp, starting frequency point, bandwidth size, authorization period, target coordinates, travel route, deployment time, bandwidth ratio, priority queue and traffic shaping parameters. The standardized control command is sent to the network infrastructure through the corresponding communication protocol interface to drive the communication resources to complete dynamic reconfiguration.

[0030] In another possible embodiment, firstly, multi-source heterogeneous data, including geographical information on the disaster impact area, regional population density heat maps, real-time base station load rates, wireless link quality indicators, and emergency service priority lists, are simultaneously collected through various channels such as satellite remote sensing equipment, ground sensor networks, and mobile terminal reporting mechanisms. Then, this data undergoes denoising, normalization, and confidence level fusion to generate comprehensive situational awareness information reflecting the degree of disaster impact, communication demand levels, and resource scarcity. Next, based on this comprehensive situational awareness information, a resource allocation optimization model is constructed with the objectives of minimizing total system energy consumption, minimizing average service latency, maximizing network coverage, and maximizing the guarantee rate of high-priority services. The decision variable... The parameters include base station transmit power adjustment, spectrum block allocation scheme, deployment location of mobile emergency communication equipment, and backhaul link bandwidth allocation ratio. Constraints include the upper limit of total available spectrum resources, maximum transmit power limit for a single base station, minimum service quality requirements, and physical location reachability constraints. Subsequently, a population-based multi-objective evolutionary optimization unit is used to solve the model, obtaining the Pareto optimal solution set within a preset number of iterations. A compromise solution is selected as the initial resource allocation strategy based on the current situation. Then, a pre-trained artificial intelligence model is used to predict the disaster evolution trend and changes in communication demand. Based on historical disaster data and real-time situation data, the spread trend of the disaster area and the impact within a preset time window are predicted. The system considers changes in population size and the temporal fluctuations in communication service requests in different regions. It also conducts a forward-looking performance evaluation of the initial resource allocation strategy, generating correction suggestions based on the evaluation results. If the score of the corrected strategy falls below a preset threshold, strategy correction is triggered. Subsequently, key performance indicators of the communication network after scheduling are collected through a dedicated monitoring interface as actual performance feedback data. Deviations from preset expected targets are calculated, and a composite triggering strategy based on event-performance dual-drive is adopted. When a major disaster change prediction or resource reconfiguration failure event is received, an event-driven weight correction is triggered. When the cumulative deviation of the actual performance feedback data exceeds a steady-state threshold, a performance-driven progressive weight adjustment is triggered, combined with a proportional-integral control mechanism with fused feedforward compensation. The system dynamically adjusts the weight coefficients of each optimization objective in the resource allocation optimization model to generate an adaptively adjusted final resource scheduling instruction. When a major disaster change prediction or resource reconfiguration failure event is received, an event-driven strong weight correction is triggered. When the cumulative deviation of key performance indicators exceeds the steady-state threshold, a performance-driven gradual weight adjustment is triggered. The final resource scheduling instruction is parsed into a specific sequence of device configuration parameters, generating a standard control command containing power value, frequency point number, timestamp, starting frequency point, bandwidth size, authorization period, target coordinates, travel route, deployment time, bandwidth ratio, priority queue, and traffic shaping parameters. This command is then sent to the network infrastructure through the corresponding communication protocol interface to drive the communication resources to complete the dynamic reconfiguration.

[0031] By fusing multi-source heterogeneous data, the comprehensiveness and accuracy of situational awareness information are ensured, providing a reliable basis for resource allocation decisions. A multi-objective optimization model combined with a multi-objective evolutionary algorithm achieves a multi-dimensional balance between energy consumption, latency, coverage, and high-priority service availability, avoiding resource misallocation caused by single-objective optimization. The introduction of an artificial intelligence prediction model enables proactive prediction of disaster evolution and communication needs, allowing for early correction of resource allocation strategies and reducing response delays. An innovative event-performance dual-driven composite triggering strategy, combined with a proportional-integral control mechanism for fusion feedforward compensation, achieves dynamic adaptive adjustment of weight coefficients, enabling rapid response to emergencies and optimizing long-term system performance, significantly improving the accuracy, robustness, and timeliness of resource allocation. Standardized instruction conversion and protocol interface adaptation ensure the universality and efficiency of resource reconfiguration, adapting to different types of network infrastructure.

[0032] Preferably, in this embodiment, S1 specifically includes: S101, synchronously acquiring geographical information of the disaster impact range, regional population density heat map, real-time base station load rate, wireless link quality indicators, and emergency service priority list through satellite remote sensing, ground sensor networks, and mobile terminal reporting mechanisms; S102, performing preprocessing operations such as denoising, normalization, and missing value imputation on the collected multi-source heterogeneous data; S103, using a multi-source information fusion unit based on confidence theory to fuse the preprocessed multi-source heterogeneous data to obtain the comprehensive situational awareness information. In another possible embodiment, firstly, satellite remote sensing equipment is used to continuously monitor the geographical changes of the disaster-affected area to obtain geographical information of the disaster area; ground sensor networks are used to collect regional population density heat map data; through a mobile terminal reporting mechanism, data such as the location information of surviving mobile terminals in the area, real-time base station load rate, and wireless link quality indicators are collected; simultaneously, an emergency service priority list is obtained through a dedicated interface, with life-saving voice communication and command and dispatch data having the highest priority, and ordinary data transmission having the lowest priority. Next, the collected raw data undergoes preprocessing: wavelet transform algorithms are used to denoise the population density data and base station load rate data collected by ground sensors to eliminate signal fluctuations caused by electromagnetic interference; a maximum-minimum normalization method is used to uniformly convert data of different dimensions to a preset range to eliminate data scale differences; for missing data caused by transmission interruptions, a time-series-based linear interpolation algorithm is used to fill in the missing data to ensure data integrity. Finally, a multi-source information fusion unit based on confidence theory is adopted. In this embodiment, the DS evidence theory fusion unit is preferred. The specific operation is as follows: Define the degree of disaster impact, including three levels: high, medium, and low; the level of communication demand, including three levels: high, medium, and low; and the state of resource scarcity, including three levels: tense, sufficient, and moderate. Assign a basic probability assignment function to each data source to characterize the degree of support of the data for core propositions such as the degree of disaster impact, the level of communication demand, and the state of resource scarcity. Then, use the Dempster combination rule to orthogonally synthesize the basic probability assignments of multiple data sources and calculate the confidence interval of each core proposition after synthesis. Finally, select the data features corresponding to the core propositions whose confidence exceeds the preset threshold and integrate them to form comprehensive situational awareness information. If the data source type changes in the scenario, it can be replaced with a fusion unit based on Bayesian networks. By calculating the conditional probability of each data source, data fusion is achieved, ensuring the flexibility of the fusion scheme.

[0033] Standardized preprocessing effectively eliminates data noise, scale differences, and missing values, improving the quality of raw data. A multi-source information fusion unit based on confidence theory supports multiple fusion algorithm replacements, adapting to different data source types and emergency scenarios, enhancing the flexibility and versatility of the solution. The fusion process, based on confidence calculations of multi-dimensional propositions, ensures the spatiotemporal consistency of the generated comprehensive situational awareness information, comprehensively and accurately reflecting the real-time situation of the emergency scenario and providing reliable support for subsequent resource allocation decisions. Clear data collection channels and preprocessing standards ensure the comprehensiveness of data collection and the standardization of processing, avoiding decision-making biases caused by data issues.

[0034] Preferably, in this embodiment, S2 specifically includes: S201, constructing the resource allocation optimization model, wherein the optimization objectives include at least two of the following: minimizing total system energy consumption, minimizing average service latency, maximizing network coverage, and maximizing high-priority service guarantee rate; the decision variables include at least one of the following: base station transmit power adjustment amount, spectrum block allocation scheme, mobile communication device deployment location, and backhaul link bandwidth allocation ratio; the constraints include: total available spectrum resource limit, maximum transmit power limit of a single base station, minimum service quality requirement, and device physical location reachability constraint. S202. The resource allocation optimization model is solved using a population-based multi-objective evolutionary optimization unit. Within a preset number of iterations, a Pareto optimal solution set is obtained. From the Pareto optimal solution set, based on the degree of disaster impact, communication demand level, and resource scarcity status reflected by the integrated situational awareness information, the solution with the highest degree of matching with the current integrated situational awareness information is selected as the preliminary resource allocation strategy. Each solution in the Pareto optimal solution set represents a resource allocation scheme, and the solutions do not dominate each other on multiple optimization objectives. Each solution cannot be a compromise solution that improves any optimization objective without harming other optimization objectives.

[0035] In another possible embodiment, a resource allocation optimization model is first constructed. The optimization objectives include at least two of the following: minimizing total system energy consumption, minimizing average service latency, maximizing network coverage, and maximizing the high-priority service guarantee rate. Total system energy consumption is calculated by summing the power consumption of all base stations, mobile emergency communication equipment, and satellite terminals. Average service latency is calculated by measuring the end-to-end transmission delay of all services. Network coverage is calculated by measuring the area where the signal strength meets the preset requirements. The high-priority service guarantee rate is calculated by measuring the successful transmission ratio of high-priority services. Decision variables include base station transmit power adjustment, spectrum block allocation scheme, deployment location of mobile emergency communication equipment, and backhaul link bandwidth allocation ratio. The constraints are specifically set as follows: total available spectrum resources do not exceed a preset upper limit, the maximum transmit power of a single base station does not exceed a preset limit, all types of services meet the minimum service quality requirements, the deployment location of mobile emergency communication equipment meets the accessibility constraints, and the base station transmit power adjustment does not exceed a preset ratio. Next, a population-based multi-objective evolutionary optimization unit is employed. This embodiment prioritizes the improved NSGA-II algorithm, and the specific solution process is as follows: Initialize a population of a preset size, with each individual corresponding to a complete resource allocation scheme, and the individual encoding length equal to the total number of decision variables; calculate the four-term objective function values ​​for each individual in the population; stratify the population individuals according to the non-dominated ranking, dividing the population into multiple non-dominated layers, where individuals within the same non-dominated layer do not dominate each other; calculate the crowding distance of individuals within the same non-dominated layer, characterizing the dispersion of individuals in the solution set; perform a selection operation based on the non-dominated level and crowding distance, selecting the better-performing individual as the parent generation; perform crossover and mutation operations on the selected parent individuals to generate the offspring population; merge the parent and offspring populations, and again calculate the non-dominated ranking and crowding distance, selecting the best-performing individual from among them. The best individuals form a new generation of population; the above iterative process is repeated until the preset number of iterations is reached, and the final Pareto optimal solution set is output, where each solution represents a resource allocation scheme, and the solutions do not dominate each other on multiple optimization objectives. Each solution is a compromise solution that cannot improve any optimization objective without harming other optimization objectives. Then, based on the degree of disaster impact, communication demand level, and resource scarcity status reflected by the comprehensive situational awareness information, the solution with the highest degree of matching with the current comprehensive situational awareness information is selected from the Pareto optimal solution set as the initial resource allocation strategy to ensure that the strategy is accurately adapted to the current scenario requirements. If the resource constraints in the scenario change, the MOEA / D algorithm or SPEA2 algorithm can be used instead, and the algorithm parameters can be adjusted and the solution can be solved again to ensure the adaptability of the solution.

[0036] By optimizing the target design from multiple dimensions, the system takes into account core requirements such as system energy consumption, service latency, network coverage, and high-priority service assurance, meeting the actual needs of emergency communication scenarios. The rich decision variable settings cover multiple dimensions, including base station power, spectrum allocation, mobile device deployment, and bandwidth allocation, ensuring the flexibility and comprehensiveness of resource allocation. Clear constraints align with actual equipment performance and service requirements, ensuring the executability of the resource allocation strategy. It supports multiple multi-objective evolutionary algorithms for replacement, adapting to the solution requirements of different emergency scenarios and improving the efficiency and accuracy of model solving. Through the selection of Pareto optimal solution sets and compromise solutions, a balance of multi-dimensional objectives is achieved, ensuring that the resource allocation strategy meets both real-time situational requirements and global optimum.

[0037] Preferably, in this embodiment, S3 specifically includes: S301, using a prediction unit based on time-series feature extraction, based on historical disaster data and real-time situation data, predicting the spread trend of the disaster range, changes in the number of affected people, and time-series fluctuations in communication service request volume in different regions within a future preset time window; S302, using a prediction unit based on network topology analysis, modeling the communication network topology, and predicting the impact of network connectivity changes caused by the disaster on the stability of the resource allocation strategy; S303, fusing the outputs of S301 and S302, inputting them into a comprehensive inference network for comprehensive inference, and outputting an effectiveness evaluation score and correction suggestions for the preliminary resource allocation strategy under the future situation; S304, setting a preset effectiveness score as... The current performance score is Correction factor Determined based on forecast uncertainty, The revised strategy score is then expressed as: ,when If the value falls below a preset threshold, a correction to the initial resource allocation strategy will be triggered.

[0038] In another possible embodiment, a prediction unit based on temporal feature extraction is first utilized. This embodiment preferentially employs a Long Short-Term Memory (LSTM) network model, which is pre-trained using historical disaster data. The input data includes historical disaster data and real-time situational data, and the output is the temporal prediction result within a preset future time window, specifically the spread trend of the disaster area, changes in the number of affected people, and temporal fluctuations in communication service request volume in different regions. If the complexity of the temporal data changes in the scenario, it can be replaced with a gated recurrent unit (GRU) model or a Transformer model, adjusting the model parameters to ensure prediction accuracy. Next, a prediction unit based on network topology analysis is utilized. This embodiment preferentially employs a graph convolutional neural network (GNN) model. This model models the current communication network topology, using base stations, mobile emergency communication equipment, and satellite terminals as network nodes, and wireless links and backhaul links as connecting edges between nodes. By calculating the importance of network nodes and the vulnerability of links, it predicts changes in network connectivity caused by disasters and outputs the impact of link interruption on policy stability. If the network topology is more complex, it can be replaced with a graph attention network model or a graph isomorphic network model to improve the accuracy of topology analysis. Then, the time-series prediction results and network topology prediction results are concatenated to form a fused feature vector, which is then input into the integrated inference network. The integrated inference network processes the fused features through a fully connected layer and outputs an effectiveness evaluation score for the initial resource allocation strategy under future conditions. And suggested revisions; simultaneously calculate the effectiveness score of the initial resource allocation strategy under the current situation. Finally, the corrected strategy score is calculated according to the preset formula. Among them, the correction factor Based on the predicted uncertainty, the calculation was performed by substituting the data. ;like If the value falls below a preset threshold, the initial resource allocation strategy will be immediately modified, and the relevant resource allocation scheme will be adjusted.

[0039] By employing a prediction unit based on temporal feature extraction, the system accurately predicts the temporal fluctuations in disaster evolution and communication demand, providing advance notice for strategy adjustments. A prediction unit based on network topology analysis anticipates the impact of changes in network connectivity on resource allocation strategies, enhancing strategy stability. Feature fusion and comprehensive reasoning between the two prediction units achieve complementarity of multi-dimensional prediction information, improving prediction accuracy and comprehensiveness. Clear performance evaluation formulas and trigger correction thresholds ensure the rationality of resource allocation strategies, correcting deficiencies in advance and avoiding resource misallocation. The system supports multiple prediction model replacements to adapt to the prediction needs of different emergency scenarios, enhancing the flexibility and versatility of the solution.

[0040] Preferably, the event-performance dual-driven composite triggering strategy in S4 of this embodiment specifically includes: S401, continuously monitoring key performance indicators of the scheduled communication network, including actual service throughput, end-to-end latency, dropout rate, and resource utilization; S402, receiving major disaster change prediction information from S3 and receiving resource reconfiguration failure event information from the execution layer; S403, triggering an event-driven weight correction mode when receiving the major disaster change prediction information or the resource reconfiguration failure information; when receiving the major disaster change prediction information and the resource reconfiguration failure information, continuously monitoring the key performance indicators at a preset period, and triggering a performance-driven weight correction mode when the cumulative deviation of the key performance indicators exceeds a preset steady-state threshold.

[0041] In another possible embodiment, after the closed-loop feedback and policy adaptive adjustment module is activated, it continuously monitors the key performance indicators (KPIs) of the scheduled communication network. These KPIs include actual service throughput, end-to-end latency, call drop rate, and resource utilization, and data for each KPI is collected in real time. Simultaneously, the module continuously receives major disaster change prediction information from the artificial intelligence prediction and inference engine, as well as resource reconfiguration failure event information from the resource scheduling execution and interface module. Then, the triggering conditions are determined: when the aforementioned major disaster change prediction information or resource reconfiguration failure event information is received, an event-driven weight correction mode is immediately triggered to prioritize responses to emergencies and ensure the smooth operation of critical services; when neither of these two types of event information is received, the cumulative deviation of KPIs is continuously calculated in preset periods. When the cumulative deviation of KPIs exceeds a preset steady-state threshold, a performance-driven progressive weight adjustment mode is triggered to gradually optimize the long-term operating performance of the system; if the cumulative deviation of each KPI within the preset period does not exceed the preset steady-state threshold, the current weight coefficient remains unchanged, and monitoring continues. Among them, information on major disaster changes and information on failed resource reallocation events are identified according to preset judgment criteria.

[0042] By combining dual-trigger modes, the system can quickly respond to emergencies, avoid policy lag, and optimize long-term system performance, achieving dynamic balance in resource allocation. It clearly defines the scope and specific indicators of KPI monitoring, ensuring comprehensive and accurate performance feedback; it clearly defines event triggering conditions, accurately identifying emergency events such as major disaster changes and resource reconfiguration failures, avoiding false triggers; performance triggering uses cumulative deviation judgment to avoid erroneous adjustments caused by fluctuations in indicators at a single moment, improving policy stability; and the adaptive switching of dual-trigger modes ensures that resource allocation strategies can adapt to dynamic changes in emergency scenarios, enhancing system robustness.

[0043] If only a single proportional-integral control is used without introducing a feedforward compensation mechanism, it will be impossible to respond in advance to sudden situations such as sudden changes in emergency service priority or network infrastructure failure, which will lead to the problem of delayed weight adjustment.

[0044] Preferably, the proportional-integral control mechanism with feedforward compensation in this embodiment specifically includes: in the event-driven weight correction mode, temporarily increasing the gain values ​​of the proportional control gain coefficient and the feedforward compensation gain coefficient to a preset multiple of the normal value, and temporarily freezing the cumulative effect of the historical deviation accumulation term; in the performance-driven weight correction mode, under the preset proportional control gain coefficient, integral control gain coefficient and feedforward compensation gain coefficient, continuously making gradual fine adjustments to the weights based on the deviation of key performance indicators.

[0045] In another possible embodiment, firstly, conventional control parameters, including the proportional control gain coefficient, are preset. Integral control gain coefficient Feedforward compensation gain coefficient These parameters have been optimized and determined to meet the needs of emergency scenarios. When the event-driven weight correction mode is triggered, in order to achieve rapid weight adjustment and cope with sudden changes in the situation under emergency scenarios, the proportional control gain coefficient will be adjusted. and feedforward compensation gain coefficient The gain value is temporarily increased to a preset multiple of the normal value, while the cumulative effect of historical deviation accumulation and term accumulation is temporarily frozen, and the integral control gain coefficient is adjusted. When temporarily set to 0, the weight coefficients are rapidly adjusted based solely on the current deviation and feedforward compensation, ensuring priority for critical business operations and rapid adaptation to unforeseen circumstances. When the performance-driven progressive weight adjustment mode is triggered, the system operates with the aforementioned preset control parameters, continuously fine-tuning the weights based on the deviation of key performance indicators. During weight coefficient adjustment, both the current deviation and the cumulative sum of historical deviations are considered, along with feedforward compensation, to gradually adjust the weight coefficients corresponding to the optimization targets. This ensures long-term system performance optimization and avoids strategy fluctuations caused by excessively rapid weight adjustments. The preset multiplier can be adjusted according to the urgency of the event, and the adjustment step size during performance triggering is set according to the preset value to ensure the gradual and stable nature of the adjustment.

[0046] By introducing feedforward compensation, we can proactively address sudden changes in business priority and network infrastructure failures, avoiding delayed weight adjustments and enhancing the system's emergency response capabilities. Control parameters adaptively adjust under different triggering modes, enabling rapid weight correction upon event triggering and gradual fine-tuning upon performance triggering, balancing response speed and long-term performance. Pre-set control parameters, tailored to the actual needs of emergency scenarios, ensure the rationality and stability of weight adjustments. The integration of proportional-integral control and feedforward compensation eliminates current deviations and compensates for potential future deviations, improving the accuracy of weight adjustments. Preferably, in this embodiment, the dynamic adjustment of the weighting coefficients is based on the following formula: ;in, Let be the weight coefficient of the i-th target at time t. The weight coefficient of the i-th target mentioned in the previous time step is... Let be the deviation of the i-th target at time t. This is the proportional control gain coefficient. The integral control gain coefficient. Let be the cumulative sum of historical deviations of the i-th target from the initial time to the current time t. This represents the change in the emergency response priority vector compared to the previous time step. This is a real-time failure indication function for network infrastructure. For feedforward compensation gain coefficient, and the This item is used to strengthen the optimization weight of key business support objectives when the disaster situation escalates or the rescue phase transitions. This item is used to assign extra weight to remaining resources when part of the network infrastructure fails, in order to prioritize maintaining critical coverage and connectivity.

[0047] In another possible embodiment, firstly, the formula for updating the weight coefficients is defined as follows: The specific meanings of each parameter are as follows: Let be the weight coefficient of the i-th target at time t. Let be the weight coefficient of the i-th objective at the previous time step. This embodiment takes the optimization objective of "high-priority business guarantee rate" as an example for explanation. Let be the deviation of the i-th target at time t, that is, the difference between the actual value and the target value; The gain coefficient is a proportional control factor. The normal value is a preset normal value, and it can be temporarily adjusted in event-driven mode. The integral control gain coefficient is normally set to a preset value, and is temporarily set to 0 in event-driven mode. This is the cumulative sum of historical deviations for the i-th target from the initial time to the current time t. This is the feedforward compensation gain coefficient, which is usually set to a preset value and temporarily adjusted in event-driven mode. This represents the change in the emergency response priority vector compared to the previous time step. This is a real-time failure indication function for network infrastructure, where failure is represented by 1 and normal operation by 0. Substituting the above parameters into the formula, the weighting coefficient for the current time t is calculated. If the event-driven mode is triggered, the corresponding control parameters are adjusted and recalculated using the formula to quickly adjust the weight coefficients, prioritizing high-priority business operations. This item is used to strengthen the optimization weight of key business support objectives when the disaster situation escalates or the rescue phase transitions. This item is used to assign extra weight to remaining resources when part of the network infrastructure fails, in order to prioritize maintaining critical coverage and connectivity and avoid overall performance collapse due to absolute reduction in resources.

[0048] Through feedforward compensation term ( The introduction of this approach enables rapid response to sudden changes in business priorities and network infrastructure failures, prioritizing critical business and core coverage; the formula is highly versatile, adaptable to different emergency scenarios and optimization goals, enhancing the applicability of the solution; the precision of weight adjustments ensures that resource allocation strategies can adapt to scenario changes in real time, improving resource utilization and critical business assurance rates.

[0049] Preferably, in this embodiment, S5 specifically includes: S501, parsing the final resource scheduling instruction into a specific sequence of device configuration parameters, wherein the final resource scheduling instruction includes an abstract instruction for optimizing the target weight adjustment result and resource allocation strategy parameters, and the device configuration parameter sequence is configuration data that is directly executable and corresponds to a specific communication device and communication protocol; S502, through an interface unit that supports multiple standard communication protocols, sending the device configuration parameter sequence to the corresponding physical device or virtualized network function to complete spectrum reconfiguration, power adjustment, beamforming, and routing strategy update operations.

[0050] In another possible embodiment, the generated final resource scheduling instructions are first parsed and converted into a specific sequence of device configuration parameters. These final resource scheduling instructions include four core instructions: base station power adjustment, spectrum block allocation, mobile emergency communication equipment deployment, and backhaul link bandwidth allocation. The parsed configuration parameter sequence corresponds to the specific execution requirements of each scheduling instruction. Next, through an interface unit supporting multiple standard communication protocols, the above device configuration parameter sequence is distributed to the corresponding physical devices or virtualized network functions. In this embodiment, the interface unit supports protocols including 4G / 5G base station management interfaces, satellite communication terminal control interfaces, and software-defined network controller southbound interfaces, each corresponding to different types of configuration parameters. If other types of devices are added to the scenario, SNMP interfaces, NETCONF / YANG interfaces, or gRPC interfaces can be adapted to achieve accurate distribution of the parameter sequence. After the parameter sequence is distributed, the interface unit receives execution feedback from the devices in real time, ensuring that each device can accurately execute the configuration parameters. After all devices have completed their execution, spectrum reconfiguration, power adjustment, beamforming, and routing policy update operations are completed, ensuring the timeliness of dynamic resource reconfiguration and meeting emergency communication needs.

[0051] By supporting multiple standard protocol interfaces, the solution adapts to different types of network infrastructure, enhancing its versatility and compatibility; a clear command issuance process ensures the security and efficiency of command transmission, improving the efficiency of resource reconfiguration; support for multiple interface protocol replacements adapts to different equipment types in different emergency scenarios, enhancing the flexibility of the solution; and the timeliness of resource reconfiguration is guaranteed, ensuring that command execution and resource reconfiguration are completed within a preset time, meeting the needs of emergency scenarios.

[0052] Example 2: Reference Figure 2This embodiment also relates to an artificial intelligence-based emergency communication resource dynamic allocation system, comprising: a multi-source heterogeneous data perception and fusion module, used to collect multi-source heterogeneous data in emergency scenarios and fuse the multi-source heterogeneous data to generate comprehensive situational awareness information; a dynamic multi-objective optimization decision-making module, used to construct a resource allocation optimization model including multiple priority objectives based on the comprehensive situational awareness information, and use a multi-objective evolutionary algorithm to solve the resource allocation optimization model to obtain a preliminary resource allocation strategy; an artificial intelligence prediction and reasoning module, used to use a pre-trained artificial intelligence model to predict the evolution trend of disaster and changes in communication demand, and to perform a forward-looking performance evaluation and strategy correction on the preliminary resource allocation strategy based on the prediction results; and a closed-loop feedback and strategy adaptive adjustment module, used to collect key performance indicators of the scheduled communication network as actual performance feedback data through a dedicated monitoring interface, and calculate the actual performance feedback... The deviation between the feedback data and the preset expected target is addressed by employing a composite triggering strategy based on event-performance dual-drive, combined with a proportional-integral control mechanism for fused feedforward compensation. This dynamically adjusts the weight coefficients of each optimization objective in the resource allocation optimization model, generating an adaptively adjusted final resource scheduling instruction. The closed-loop feedback and strategy adaptive adjustment module incorporates the composite triggering strategy based on event-performance dual-drive and the proportional-integral control mechanism for fused feedforward compensation. The resource scheduling execution and interface module parses the final resource scheduling instruction into a specific sequence of device configuration parameters, generating a standard control command containing power value, frequency point number, timestamp, starting frequency point, bandwidth size, authorization period, target coordinates, travel route, deployment time, bandwidth ratio, priority queue, and traffic shaping parameters. This standardized control command is then sent to the network infrastructure through the corresponding communication protocol interface to drive the communication resources to complete dynamic reconfiguration.

[0053] In another possible embodiment, the multi-source heterogeneous data perception and fusion module first collects multi-source heterogeneous data in the emergency scenario and fuses the data to generate comprehensive situational awareness information. This information is then synchronously transmitted to the dynamic multi-objective optimization decision-making module and the artificial intelligence prediction and inference module. Next, based on the comprehensive situational awareness information, the dynamic multi-objective optimization decision-making module constructs a resource allocation optimization model including multiple priority objectives and solves the model using a multi-objective evolutionary algorithm to obtain a preliminary resource allocation strategy. This strategy is then transmitted to the artificial intelligence prediction and inference module. The artificial intelligence prediction and inference module then uses a pre-trained artificial intelligence model to predict the evolution trend of the disaster and changes in communication demand. Based on the prediction results, it performs a forward-looking performance evaluation and strategy correction on the preliminary resource allocation strategy. The corrected preliminary resource allocation strategy and performance evaluation results are transmitted to the closed-loop feedback and strategy adaptive adjustment module. Finally, the closed-loop feedback and strategy adaptive adjustment module collects key performance indicators of the scheduled communication network through a dedicated monitoring interface as actual performance feedback data and calculates the actual performance... The deviation between performance feedback data and preset expected targets is addressed using a composite triggering strategy based on event-performance dual-drive, combined with a proportional-integral control mechanism that integrates feedforward compensation. This dynamically adjusts the weight coefficients of each optimization objective in the resource allocation optimization model, generating an adaptively adjusted final resource scheduling command. The closed-loop feedback and policy adaptive adjustment module, which incorporates the event-performance dual-drive composite triggering strategy and the proportional-integral control mechanism with feedforward compensation, transmits the final resource scheduling command to the resource scheduling execution and interface module. Finally, the resource scheduling execution and interface module parses the final resource scheduling command into a specific sequence of device configuration parameters, generating a standard control command containing power value, frequency point number, timestamp, starting frequency point, bandwidth size, authorization period, target coordinates, travel route, deployment time, bandwidth ratio, priority queue, and traffic shaping parameters. This standardized control command is then distributed to the network infrastructure through the corresponding communication protocol interface to drive dynamic reconfiguration of communication resources. Simultaneously, the device execution feedback data is sent back to the closed-loop feedback and policy adaptive adjustment module, forming a complete closed-loop process.

[0054] Preferably, the multi-source heterogeneous data sensing and fusion module is deployed at the data acquisition front end for accessing satellite remote sensing data interfaces, ground sensor networks, and mobile terminal reporting channels; the dynamic multi-objective optimization decision-making module and the artificial intelligence prediction and inference module are deployed on edge computing nodes, which are set up in emergency command vehicles or temporary command centers for completing comprehensive situational fusion and optimization calculations locally; the closed-loop feedback and policy adaptive adjustment module is deployed on the edge computing nodes or cloud servers; and the resource scheduling execution and interface module is deployed on the network infrastructure side for direct interface with physical communication equipment or virtualized network functions.

[0055] In another possible embodiment, firstly, the multi-source heterogeneous data perception and fusion module is deployed at the data acquisition front end to access satellite remote sensing data interfaces, ground sensor networks, and mobile terminal reporting channels, directly collecting multi-source heterogeneous data in emergency scenarios. Next, the dynamic multi-objective optimization decision-making module and the artificial intelligence prediction and inference module are deployed at edge computing nodes, located in emergency command vehicles or temporary command centers, to perform comprehensive situational fusion and optimization calculations locally, generating preliminary resource allocation strategies and forward-looking correction suggestions without relying on the cloud. Then, the closed-loop feedback and policy adaptive adjustment module is deployed at edge computing nodes or cloud servers, with the deployment location selected according to scenario requirements; it is deployed at edge nodes when rapid local response is needed, and at cloud servers when global optimization is required. Finally, the resource scheduling execution and interface module is deployed on the network infrastructure side to directly interface with physical communication devices or virtualized network functions, ensuring the real-time and accuracy of command issuance. Through layered deployment, the entire process of front-end acquisition, edge computing, cloud / edge control, and device-side execution is coordinated, improving system response speed and reliability.

[0056] Prior to this, the emergency communication resource dynamic allocation system further includes a federated learning collaborative optimization layer, which is deployed on a cloud server or among multiple edge computing nodes to achieve cross-regional or cross-departmental model collaborative training and strategy optimization by exchanging encrypted model parameter updates.

[0057] In another possible embodiment, the federated learning collaborative optimization layer is first deployed on a cloud server or among multiple edge computing nodes. Edge nodes in each region train artificial intelligence prediction models and multi-objective optimization algorithms locally, using only local data and not transmitting raw data. After training, each edge node only uploads the encrypted model parameters to the cloud server or exchanges them with other edge nodes. The cloud server merges the encrypted model parameters from all regions to generate a global optimization model, and then distributes the optimized model parameters to each edge node. Each edge node updates its local model based on the global optimization model, realizing cross-regional or cross-departmental model collaborative training and strategy optimization. Under the premise of protecting data privacy, it enables experience sharing and model evolution across multiple regions, improving the system's ability to respond to new or complex disasters.

[0058] In some embodiments, the AI-based dynamic allocation system for emergency communication resources of the present invention can be implemented using a combination of hardware and software. As an example, the AI-based dynamic allocation system for emergency communication resources of the present invention can be a processor in the form of a hardware decoding processor, which is programmed to execute the AI-based dynamic allocation method for emergency communication resources of the present invention. For example, the processor in the form of a hardware decoding processor can be one or more application-specific integrated circuits (ASICs), DSPs, programmable logic devices (PLDs), complex programmable logic devices (CPLDs), field-programmable gate arrays (FPGAs), or other electronic components.

[0059] The modules described in the embodiments of this invention can be implemented in software or hardware. The names of the modules are not, in some cases, limiting the scope of the module itself.

[0060] An electronic device according to an embodiment of the present invention includes a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, it implements any of the above-mentioned artificial intelligence-based dynamic allocation methods for emergency communication resources. That is, an electronic device according to an embodiment of the present invention may include, but is not limited to: a processor and a memory; the memory is used to store the computer program; the processor is used to execute the artificial intelligence-based dynamic allocation method for emergency communication resources shown in any embodiment of the present invention by calling the computer program.

[0061] In one alternative embodiment, an electronic device is provided, such as Figure 3 As shown, Figure 3 The illustrated electronic device 4000 includes a processor 4001 and a memory 4003. The processor 4001 and the memory 4003 are connected, for example, via a bus 4002. Optionally, the electronic device 4000 may further include a transceiver 4004, which can be used for data interaction between the electronic device and other electronic devices, such as sending and / or receiving data. It should be noted that in practical applications, the transceiver 4004 is not limited to one type, and the structure of the electronic device 4000 does not constitute a limitation on the embodiments of the present invention.

[0062] Processor 4001 may be a CPU (Central Processing Unit), a general-purpose processor, a DSP (Digital Signal Processor), an ASIC (Application Specific Integrated Circuit), an FPGA (Field Programmable Gate Array), or other programmable logic devices, transistor logic devices, hardware components, or any combination thereof. It can implement or execute the various exemplary logic blocks, modules, and circuits described in conjunction with the disclosure of this invention. Processor 4001 may also be a combination that implements computational functions, such as including one or more microprocessor combinations, a combination of a DSP and a microprocessor, etc.

[0063] Bus 4002 may include a path for transmitting information between the aforementioned components. Bus 4002 may be a PCI (Peripheral Component Interconnect) bus or an EISA (Extended Industry Standard Architecture) bus, etc. Bus 4002 can be divided into address bus, data bus, control bus, etc. For ease of representation, Figure 3 The bus 4002 is represented by only one thick line, but this does not mean that there is only one bus or one type of bus.

[0064] The memory 4003 may be ROM (Read Only Memory) or other types of static storage devices capable of storing static information and instructions, RAM (Random Access Memory) or other types of dynamic storage devices capable of storing information and instructions, or EEPROM (Electrically Erasable Programmable Read Only Memory), CD-ROM (Compact Disc Read Only Memory) or other optical disc storage, optical disc storage (including compressed optical discs, laser discs, optical discs, digital universal optical discs, Blu-ray discs, etc.), magnetic disk storage media or other magnetic storage devices, or any other medium capable of carrying or storing desired program code in the form of instructions or data structures and accessible by a computer, but not limited thereto.

[0065] The memory 4003 stores application code (computer program) for executing the present invention, and its execution is controlled by the processor 4001. The processor 4001 executes the application code stored in the memory 4003 to implement the content shown in the foregoing method embodiments.

[0066] Among them, electronic devices can also be terminal devices, which can be any device that can install applications, including at least one of smartphones, tablets, laptops, desktop computers, smart speakers, smartwatches, smart TVs, and smart in-vehicle devices.

[0067] It should be noted that, Figure 3 The electronic device shown is merely an example and should not be construed as limiting the functionality and scope of use of the embodiments of the present invention.

[0068] An embodiment of the present invention provides a computer-readable storage medium storing a computer program, which, when executed by a processor, implements any of the above-mentioned methods for dynamic allocation of emergency communication resources based on artificial intelligence.

[0069] Alternatively, the computer-readable storage medium may be a read-only memory (ROM), a random access memory (RAM), a compact disc read-only memory (CD-ROM), magnetic tape, a floppy disk, and an optical data storage device, etc.

[0070] In an exemplary embodiment, a computer program product or computer program is also provided, which includes computer instructions stored in a computer-readable storage medium. The processor of an electronic device reads the computer instructions from the computer-readable storage medium and executes the computer instructions, causing the electronic device to perform the aforementioned artificial intelligence-based dynamic allocation method for emergency communication resources.

[0071] Computer program code for performing the operations of this invention can be written in one or more programming languages ​​or a combination thereof, including object-oriented programming languages ​​such as Java, Smalltalk, and C++, and conventional procedural programming languages ​​such as C or similar languages. The program code can be executed entirely on the user's computer, partially on the user's computer, as a standalone software package, partially on the user's computer and partially on a remote computer, or entirely on a remote computer or server. In cases involving remote computers, the remote computer can be connected to the user's computer via any type of network—including a local area network (LAN) or a wide area network (WAN)—or can be connected to an external computer (e.g., via the Internet using an Internet service provider).

[0072] It should be understood that the flowcharts and block diagrams in the accompanying drawings illustrate the architecture, functionality, and operation of possible implementations of methods and computer program products according to various embodiments of the present invention. In this regard, each block in a flowchart or block diagram may represent a module, segment, or portion of code containing one or more executable instructions for implementing the 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 the block diagrams and / or flowcharts, and combinations of blocks in the block diagrams and / or flowcharts, may 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.

[0073] The computer-readable storage medium provided in this invention can be, but is 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 (EEPROM or flash memory), optical fiber, portable compact disk read-only memory (CD-ROM), optical storage device, magnetic storage device, or any suitable combination thereof. In this invention, 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.

[0074] The aforementioned computer-readable storage medium carries one or more programs, which, when executed by the electronic device, cause the electronic device to perform the method shown in the above embodiments.

[0075] The above description is merely a preferred embodiment of the present invention and an explanation of the technical principles employed. Those skilled in the art should understand that the scope of disclosure in this invention is not limited to technical solutions formed by specific combinations of the above-described technical features, but should also cover other technical solutions formed by arbitrary combinations of the above-described technical features or their equivalents without departing from the above-disclosed concept. For example, technical solutions formed by substituting the above features with (but not limited to) technical features with similar functions disclosed in this invention.

[0076] It should be noted that the terms "first," "second," etc., used in the specification and claims of this application are used to distinguish similar objects and represent a limitation on a specific order or sequence. Where appropriate, the order of use for similar objects can be interchanged so that the embodiments of this application described herein can be implemented in an order other than that shown or described.

[0077] Those skilled in the art will recognize that this invention can be implemented as a system, method, or computer program product. Therefore, this invention can be specifically implemented in the following forms: it can be entirely hardware, entirely software (including firmware, resident software, microcode, etc.), or a combination of hardware and software, generally referred to herein as a "circuit," "module," or "system." Furthermore, in some embodiments, this invention can also be implemented as a computer program product contained in one or more computer-readable media, which includes computer-readable program code.

[0078] Although embodiments of the present invention have been shown and described above, it is understood that the above embodiments are exemplary and should not be construed as limiting the present invention. Those skilled in the art can make changes, modifications, substitutions and variations to the above embodiments within the scope of the present invention.

Claims

1. A method for dynamic allocation of emergency communication resources based on artificial intelligence, characterized in that, Includes the following steps: S1. Collect multi-source heterogeneous data in emergency scenarios, and fuse the multi-source heterogeneous data to generate comprehensive situational awareness information; S2. Based on the comprehensive situational awareness information, a resource allocation optimization model including multiple priority objectives is constructed, and a multi-objective evolutionary algorithm is used to solve the resource allocation optimization model to obtain a preliminary resource allocation strategy. S3. Utilize a pre-trained artificial intelligence model to predict the evolution trend of the disaster and changes in communication needs, and conduct a forward-looking performance evaluation and strategy correction of the preliminary resource allocation strategy based on the prediction results. S4. Collect key performance indicators of the scheduled communication network through a dedicated monitoring interface as actual performance feedback data, calculate the deviation between the actual performance feedback data and the preset expected target, adopt a composite triggering strategy based on event-performance dual drive, and combine it with a proportional-integral control mechanism with fusion feedforward compensation to dynamically adjust the weight coefficients of each optimization target in the resource allocation optimization model, and generate the final resource scheduling instruction after adaptive adjustment. S5. Parse the final resource scheduling instruction into a specific sequence of device configuration parameters, generate a standard control command containing power value, frequency point number, timestamp, starting frequency point, bandwidth size, authorization period, target coordinates, travel route, deployment time, bandwidth ratio, priority queue and traffic shaping parameters, and send the standardized control command to the network infrastructure through the corresponding communication protocol interface to drive the communication resources to complete dynamic reconfiguration.

2. The method for dynamic allocation of emergency communication resources based on artificial intelligence according to claim 1, characterized in that, S1 specifically includes: S101. Through satellite remote sensing, ground sensor networks and mobile terminal reporting mechanisms, simultaneously acquire geographical information of the disaster impact range, regional population density heat map, real-time base station load rate, wireless link quality indicators and emergency service priority list. S102. Perform preprocessing operations such as denoising, normalization, and missing value imputation on the collected multi-source heterogeneous data. S103. A multi-source information fusion unit based on confidence theory is used to fuse the preprocessed multi-source heterogeneous data to obtain the comprehensive situational awareness information.

3. The method for dynamic allocation of emergency communication resources based on artificial intelligence according to claim 1, characterized in that, S2 specifically includes: S201. Construct the resource allocation optimization model, wherein the optimization objectives include at least two of the following: minimizing total system energy consumption, minimizing average service latency, maximizing network coverage, and maximizing high-priority service guarantee rate; the decision variables include at least one of the following: base station transmit power adjustment amount, spectrum block allocation scheme, mobile communication equipment deployment location, and backhaul link bandwidth allocation ratio; the constraints include: total available spectrum resource limit, maximum transmit power limit of a single base station, minimum service quality requirement, and equipment physical location reachability constraint. S202. The resource allocation optimization model is solved using a population-based multi-objective evolutionary optimization unit. Within a preset number of iterations, a Pareto optimal solution set is obtained. From the Pareto optimal solution set, based on the degree of disaster impact, communication demand level, and resource scarcity status reflected by the integrated situational awareness information, the solution with the highest degree of matching with the current integrated situational awareness information is selected as the preliminary resource allocation strategy. Each solution in the Pareto optimal solution set represents a resource allocation scheme, and the solutions do not dominate each other on multiple optimization objectives. Each solution cannot be a compromise solution that improves any optimization objective without harming other optimization objectives.

4. The method for dynamic allocation of emergency communication resources based on artificial intelligence according to claim 1, characterized in that, Specifically, S3 includes: S301. Using a prediction unit based on time-series feature extraction, based on historical disaster data and real-time situation data, predict the spread trend of the disaster range, the change in the number of affected people, and the time-series fluctuation of communication service request volume in different regions within a future preset time window. S302. Using a prediction unit based on network topology analysis, model the communication network topology and predict the impact of network connectivity changes caused by disasters on the stability of resource allocation strategies. S303. The outputs of S301 and S302 are feature-fused and input into the comprehensive reasoning network for comprehensive reasoning. The output is an effectiveness evaluation score and correction suggestions for the preliminary resource allocation strategy under the future situation. S304, Set the preset performance score as follows The current performance score is Correction factor Determined based on forecast uncertainty, The revised strategy score is then expressed as: ,when If the value falls below a preset threshold, a correction to the initial resource allocation strategy will be triggered.

5. The method for dynamic allocation of emergency communication resources based on artificial intelligence according to claim 1, characterized in that, The event-performance dual-driven composite triggering strategy in S4 specifically includes: S401. Continuously monitor the key performance indicators of the communication network after scheduling, including actual service throughput, end-to-end latency, disconnection rate and resource utilization. S402, Receive major disaster change prediction information from S3 and receive resource reconfiguration failure event information from the execution layer; S403. When the major disaster change prediction information or the resource configuration failure information is received, the event-driven weight correction mode is triggered. When the major disaster change prediction information and the resource configuration failure information are received, the key performance indicators are continuously monitored at a preset period. When the cumulative deviation of the key performance indicators exceeds the preset steady-state threshold, the performance-driven weight correction mode is triggered.

6. The method for dynamic allocation of emergency communication resources based on artificial intelligence according to claim 5, characterized in that, The proportional-integral control mechanism for fusion feedforward compensation specifically includes: In the event-driven weight correction mode, the gain values ​​of the proportional control gain coefficient and the feedforward compensation gain coefficient are temporarily increased to a preset multiple of the normal value, and the cumulative effect of the historical deviation accumulation term is temporarily frozen. In the performance-driven weight correction mode, the weights are continuously fine-tuned based on the deviation of key performance indicators while operating with the preset proportional control gain coefficient, integral control gain coefficient and feedforward compensation gain coefficient.

7. The method for dynamic allocation of emergency communication resources based on artificial intelligence according to claim 6, characterized in that, The dynamic adjustment of the weighting coefficients is based on the following formula: ； in, Let be the weight coefficient of the i-th target at time t. The weight coefficient of the i-th target mentioned in the previous time step is... Let be the deviation of the i-th target at time t. This is the proportional control gain coefficient. The integral control gain coefficient. Let be the cumulative sum of historical deviations of the i-th target from the initial time to the current time t. This represents the change in the emergency response priority vector compared to the previous time step. This is a real-time failure indication function for network infrastructure. For feedforward compensation gain coefficient, and the This item is used to strengthen the optimization weight of key business support objectives when the disaster situation escalates or the rescue phase transitions. This item is used to assign extra weight to remaining resources when part of the network infrastructure fails, in order to prioritize maintaining critical coverage and connectivity.

8. The method for dynamic allocation of emergency communication resources based on artificial intelligence according to claim 1, characterized in that, S5 specifically includes: S501. The final resource scheduling instruction is parsed into a specific sequence of device configuration parameters, wherein the final resource scheduling instruction includes an abstract instruction for optimizing the target weight adjustment result and resource allocation strategy parameters, and the sequence of device configuration parameters is directly executable configuration data corresponding to specific communication devices and communication protocols. S502. Through an interface unit that supports multiple standard communication protocols, the device configuration parameter sequence is sent to the corresponding physical device or virtualized network function to complete spectrum reconfiguration, power adjustment, beamforming, and routing policy update operations.

9. An emergency communication resource dynamic allocation system based on artificial intelligence, characterized in that, The artificial intelligence-based emergency communication resource dynamic allocation system, applicable to the method described in any one of claims 1-8, comprises: The multi-source heterogeneous data perception and fusion module is used to collect multi-source heterogeneous data in emergency scenarios and perform fusion processing on the multi-source heterogeneous data to generate comprehensive situational awareness information. The dynamic multi-objective optimization decision module is used to construct a resource allocation optimization model including multiple priority objectives based on the comprehensive situational awareness information, and to solve the resource allocation optimization model using a multi-objective evolutionary algorithm to obtain a preliminary resource allocation strategy. The artificial intelligence prediction and reasoning module is used to predict the evolution trend of disaster and changes in communication demand using a pre-trained artificial intelligence model, and to conduct a forward-looking performance evaluation and strategy correction of the preliminary resource allocation strategy based on the prediction results. The closed-loop feedback and policy adaptive adjustment module is used to collect key performance indicators of the scheduled communication network as actual performance feedback data through a dedicated monitoring interface, calculate the deviation between the actual performance feedback data and the preset expected target, adopt a composite triggering strategy based on event-performance dual drive, and combine it with a proportional-integral control mechanism with fusion feedforward compensation to dynamically adjust the weight coefficients of each optimization target in the resource allocation optimization model, and generate the final resource scheduling instruction after adaptive adjustment. The closed-loop feedback and policy adaptive adjustment module has the composite triggering strategy based on event-performance dual drive and the proportional-integral control mechanism with fusion feedforward compensation built in. The resource scheduling execution and interface module is used to parse the final resource scheduling instruction into a specific sequence of device configuration parameters, generate standard control commands containing power value, frequency point number, timestamp, starting frequency point, bandwidth size, authorization period, target coordinates, travel route, deployment time, bandwidth ratio, priority queue and traffic shaping parameters, and send the standardized control commands to the network infrastructure through the corresponding communication protocol interface to drive the communication resources to complete dynamic reconfiguration.

10. The emergency communication resource dynamic allocation system based on artificial intelligence according to claim 9, characterized in that, The multi-source heterogeneous data sensing and fusion module is deployed at the data acquisition front end and is used to access the satellite remote sensing data interface, the ground sensor network and the mobile terminal reporting channel. The dynamic multi-objective optimization decision-making module and the artificial intelligence prediction and reasoning module are deployed on edge computing nodes, which are set up in emergency command vehicles or temporary command posts to complete comprehensive situation fusion and optimization calculations locally. The closed-loop feedback and strategy adaptive adjustment module is deployed on the edge computing node or cloud server. The resource scheduling execution and interface module is deployed on the network infrastructure side and is used to directly interface with physical communication equipment or virtualized network functions.

11. The emergency communication resource dynamic allocation system based on artificial intelligence according to claim 9, characterized in that, The emergency communication resource dynamic allocation system also includes: The federated learning collaborative optimization layer is deployed on a cloud server or among multiple edge computing nodes. It is used to achieve cross-regional or cross-departmental collaborative model training and strategy optimization by exchanging encrypted model parameter updates.

12. An electronic device, characterized in that, include: processor; Memory used to store the processor's executable instructions; The processor is configured to execute the instructions to implement the AI-based dynamic allocation method for emergency communication resources as described in any one of claims 1-8.

13. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores at least one computer program, which is loaded and executed by a processor to enable the computer to implement the artificial intelligence-based dynamic allocation method for emergency communication resources as described in any one of claims 1-8.

Images