AI-based emergency response intelligent decision-making and data security assurance system

Abstract

The present invention relates to the field of data security technology, and in particular to an AI-based emergency response intelligent decision-making and data security assurance system, including an AI emergency management center, an emergency response unit, a dynamic response unit, an information management unit, and a back-end feedback unit. The present invention obtains priority emergency plans through multi-source data collection and AI algorithm model analysis, accompanied by intelligent solution optimization analysis under comprehensive evaluation of resource allocation, response performance, and impact range, and tracks the execution of the plans and provides feedback on the results, so as to make overall dynamic adjustments to the plans or targeted adjustments to single parameters to minimize the negative impact of emergencies, and perform sensitive identification and management level division processing from the perspective of data security, that is, by automatically identifying sensitive data in the emergency response system and performing risk grading, a basis is provided for subsequent data management, significantly improving the data security protection capabilities during the emergency response process.

Description

Technical Field

[0001] The present invention relates to the field of data security technology, and in particular to an AI-based emergency response intelligent decision-making and data security assurance system. Background Art

[0002] When emergencies occur (such as natural disasters, public health incidents, and production accidents), rapid and accurate emergency response decisions are crucial and directly impact the safety of people's lives and property, as well as social stability. Traditional emergency response decisions rely primarily on manual experience, resulting in slow response times, low decision-making accuracy, and significant influence from subjective factors.

[0003] However, existing emergency response systems still have many problems: multi-source data is scattered across different departments or terminals, making it difficult to fully acquire multi-source data. This reduces the accuracy of AI decision-making model analysis, and makes it difficult to flexibly and dynamically adjust plans based on changes at the emergency site, making it difficult to sustain the effectiveness of emergency decisions. Furthermore, during the emergency response process, a large amount of sensitive data, such as personal privacy data and government confidential data, is involved, making it difficult to effectively protect sensitive data and increasing the risk of data leakage.

[0004] In view of the above technical defects, a solution is now proposed. Summary of the Invention

[0005] The purpose of the present invention is to provide an AI-based emergency response intelligent decision-making and data security assurance system to solve the technical defects mentioned above. The present invention obtains priority emergency plans through multi-source data collection and AI algorithm model analysis, accompanied by intelligent solution optimization analysis under comprehensive evaluation of resource allocation, response performance and impact range, and tracks the execution of the plan and provides feedback on the results so as to make overall dynamic adjustments to the plan or targeted adjustments to single parameters to minimize the negative impact of emergencies. From the perspective of data security, sensitive identification and management level division are performed, that is, by automatically identifying sensitive data in the emergency response system and performing risk classification, a basis is provided for subsequent data management, which significantly improves the data security protection capabilities during the emergency response process.

[0006] The purpose of the present invention can be achieved by the following technical solutions: an AI-based emergency response intelligent decision-making and data security assurance system, including an AI emergency management center, an emergency response unit, a dynamic response unit, an information management unit, and a back-end feedback unit;

[0007] The AI ​​emergency management center is used to retrieve and store multi-source data related to emergencies in the target area;

[0008] The emergency response unit is used to retrieve multi-source data and conduct emergency response feedback analysis, and obtain a priority emergency plan based on the obtained plan priority coefficient;

[0009] The dynamic response unit is used to analyze the flexibility of the implementation process of the priority emergency plan, determine whether the priority emergency plan should be adjusted, and obtain an over-deviation signal or a single parameter adjustment plan;

[0010] Further analyze the control effect of the single parameter adjustment scheme, perform discrimination processing on the obtained reduction rate, and obtain an invalid signal or a valid signal. At the same time, based on the valid signal, perform discrimination processing on the obtained continuous fit degree, and obtain a stable signal or an alarm signal.

[0011] The information management unit is used to perform data sensitivity identification and management level classification processing on the emergency response data related to the sudden incidents in the collected target area, and obtain high-level encryption, medium-level encryption and low-level encryption.

[0012] Preferably, the analysis process of the emergency response unit is as follows:

[0013] S1: Real-time acquisition of multi-source data related to emergencies in the target area;

[0014] S2: Preprocessing the collected multi-source data related to emergencies in the target area;

[0015] S3: Inputting the initial multi-source data into a pre-set AI emergency response decision model to obtain multiple emergency response plans output by the pre-set AI emergency response decision model;

[0016] S4: Obtain basic information of each emergency response plan, including a resource allocation indicator, a response time indicator, and an impact range indicator; obtain preset weight coefficients of the resource allocation indicator, the response time indicator, and the impact range indicator; and set the sum of the resource allocation indicator, the response time indicator, and the impact range indicator multiplied by the corresponding preset weight coefficients as the plan priority coefficient;

[0017] S5: Sort the obtained scheme priority coefficients in ascending order, and based on the sorting result, set the emergency response scheme corresponding to the minimum scheme priority coefficient as the priority emergency scheme.

[0018] Preferably, the resource allocation index represents the sum of the amount of redundant resources and the amount of resource gap, divided by the total amount of resource input; the response time index represents the ratio between the actual response time and the maximum allowable response time; the impact range index represents the ratio between the actual impact and the potential maximum impact, and the actual impact represents the sum of the number of affected people and the affected economy multiplied by the corresponding preset weight coefficient.

[0019] Preferably, the analysis process of the dynamic response unit is as follows:

[0020] Based on the implementation process of the priority emergency plan, obtain the on-site multi-source data of the target area during the implementation process and pre-process the on-site multi-source data;

[0021] Based on the pre-processed on-site multi-source data, the values ​​of various parameters in the on-site multi-source data are obtained, and each parameter is discriminated and processed to obtain the discrimination processing results of each parameter, where the discrimination processing results include whether the parameter meets the standard or not.

[0022] The numerical deviation of the corresponding parameter whose judgment result is not up to standard is obtained, and the numerical deviation of the corresponding parameter whose judgment result is not up to standard is set as the deviation degree, and whether the deviation degree exceeds the preset deviation degree threshold is judged. If so, the corresponding parameter is judged to be an over-deviation parameter, if not, the corresponding parameter is judged to be a low-deviation parameter.

[0023] Preferably, the sum of the excess deviation parameter and the low deviation parameter is obtained, and the sum of the excess deviation parameter and the low deviation parameter is set as the scheme deviation degree, and a judgment is made as to whether the scheme deviation degree exceeds a preset scheme deviation degree threshold. If so, an over-deviation signal is generated, and if not, a normal signal is generated.

[0024] Preferably, when a conventional signal is generated, a variation curve of the excess deviation parameter and the low deviation parameter is constructed based on the time series, and the variation curve of the excess deviation parameter and the low deviation parameter is input into a preset AI parameter adjustment model to obtain an output single parameter adjustment plan;

[0025] Based on a single parameter adjustment scheme, the reduction rate of the deviation degree of the over-deviation parameter and the under-deviation parameter is obtained, and the reduction rate is compared with the preset reduction rate range. If the reduction rate falls within the preset reduction rate range, a valid signal is generated; if the reduction rate does not fall within the preset reduction rate range, an invalid signal is generated.

[0026] Preferably, when a valid signal is generated, a reduction rate change curve is constructed based on the time series, and a preset reduction rate change curve is obtained at the same time. The difference between the reduction rate change curve and the preset reduction rate change curve is set as the continuous fit, and whether the continuous fit is less than the preset continuous fit threshold is judged. If so, a stable signal is generated; if not, an alarm signal is generated.

[0027] Preferably, the data sensitivity identification and management level classification process is as follows:

[0028] Obtain emergency response data related to emergencies in the target area, including text information and image information;

[0029] Performing sensitive information processing on the emergency response data to obtain the emergency response data after sensitive information processing, and setting the emergency response data after sensitive information processing as data to be managed;

[0030] The amount of data to be managed is obtained, as is the difficulty of data leakage recovery for the data to be managed. The difficulty of data leakage recovery represents the combination of the normalized values ​​of data recovery cost and data recovery time. The data leakage probability of the data to be managed is also obtained. The data leakage probability represents the ratio of the number of historical data leakages to the total number of storage times of the data storage device to be managed.

[0031] The weight coefficient w1 of the data volume of the data to be managed, the weight coefficient w2 of the difficulty of leakage recovery, and the weight coefficient w3 of the data leakage probability are obtained, and w1, w2, and w3 are all greater than zero. The sum of the data volume of the data to be managed, the difficulty of leakage recovery, and the probability of data leakage multiplied by w1, w2, and w3 is set as the data protection requirement value, and the data protection requirement value is discriminated and processed to obtain high-level encryption, intermediate encryption, or low-level encryption.

[0032] The beneficial effects of the present invention are as follows:

[0033] Through multi-source data collection and AI algorithm model analysis, this invention can quickly generate scientific and reasonable emergency response decision-making plans. At the same time, it is accompanied by intelligent solution optimization analysis based on comprehensive evaluation of resource allocation, response performance, and impact range. That is, through multi-objective balance, the feasibility and effectiveness of the plan are ensured, providing scientific support for emergency command.

[0034] The present invention also tracks the execution of the plan and provides feedback on the results, so as to make dynamic adjustments to the plan as a whole or to target individual parameters, thereby improving the flexibility and effectiveness of emergency response and minimizing the negative impact of emergencies. At the same time, the adjusted parameters are continuously evaluated to ensure the feasibility of the adjustment plan and the accuracy of the adjustment model.

[0035] The present invention performs sensitive identification and management level classification from the perspective of data security. That is, by automatically identifying sensitive data in the emergency response system and performing risk classification, it provides a basis for subsequent data management and significantly improves the data security protection capability during the emergency response process. BRIEF DESCRIPTION OF THE DRAWINGS

[0036] The present invention will be further described below with reference to the accompanying drawings;

[0037] Figure 1 It is a flow chart of the system of the present invention;

[0038] Figure 2 This is a local analysis diagram of Example 1 of the present invention. DETAILED DESCRIPTION

[0039] The following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative efforts are within the scope of protection of the present invention.

[0040] References to "embodiments" herein mean that a particular feature, structure, or characteristic described in connection with the embodiment may be included in at least one embodiment of the present invention. The appearance of the phrase in various places in the specification does not necessarily refer to the same embodiment, nor does it constitute a separate or alternative embodiment that is mutually exclusive of other embodiments. It is understood, both explicitly and implicitly, by those skilled in the art that the embodiments described herein may be combined with other embodiments;

[0041] Example 1: Please refer to Figures 1 to 2 As shown, the present invention is an AI-based emergency response intelligent decision-making and data security assurance system, including an AI emergency management center, an emergency response unit, a dynamic response unit, an information management unit, and a back-end feedback unit. The AI ​​emergency management center is connected to the emergency response unit in a two-way communication manner, the AI ​​emergency management center is connected to the dynamic response unit and the information management unit in a one-way communication manner, the information management unit is connected to the dynamic response unit in a one-way communication manner, and the dynamic response unit is connected to the back-end feedback unit in a one-way communication manner;

[0042] The AI ​​emergency management center is used to retrieve and store multi-source data related to emergencies in the target area;

[0043] The emergency response unit is used to retrieve multi-source data and perform emergency response feedback analysis. The priority emergency plan is obtained based on the obtained plan priority coefficient. The specific emergency response feedback analysis process is as follows:

[0044] S1: Real-time acquisition of multi-source data related to emergencies in the target area, including environmental data, personnel data, etc.

[0045] Environmental data includes ambient temperature, humidity, etc. Personnel data includes the total number of people on site, number of injured people, etc.

[0046] For example, sensors, cameras, network interfaces, and other devices distributed across various scenarios can be used to collect various data related to emergencies in real time, including but not limited to environmental data, personnel data, and equipment operation data at the scene of the incident.

[0047] S2: Preprocessing the collected multi-source data related to the emergency in the target area, including format conversion, noise filtering, etc., and setting the preprocessed multi-source data as the initial multi-source data;

[0048] S3: Inputting the initial multi-source data into a pre-set AI emergency response decision model to obtain multiple emergency response plans output by the pre-set AI emergency response decision model;

[0049] Leveraging the powerful analytical and learning capabilities of AI algorithms, it is possible to quickly mine useful information from large amounts of data, generate scientific and reasonable decision-making plans, and improve the efficiency and accuracy of emergency response decisions;

[0050] S4: Obtain basic information of each emergency response plan, including a resource allocation indicator, a response time indicator, and an impact range indicator; obtain preset weight coefficients of the resource allocation indicator, the response time indicator, and the impact range indicator; and set the sum of the resource allocation indicator, the response time indicator, and the impact range indicator multiplied by the corresponding preset weight coefficients as the plan priority coefficient;

[0051] S5: Sort the obtained plan priority coefficients in ascending order. Based on the sorting results, the emergency response plan corresponding to the plan with the smallest priority coefficient is set as the priority emergency plan. The AI ​​emergency management center stores and executes the priority emergency plan.

[0052] The resource allocation index represents the sum of the redundant resources and the resource gap, divided by the total resource input. It should be noted that the smaller the value of the resource allocation index, the better.

[0053] The response time indicator represents the ratio of the actual response time to the maximum allowed response time. It should be noted that the time from the occurrence of an event to the implementation of emergency measures is the shorter the actual response time, the better. Therefore, the smaller the value of the response time indicator, the better.

[0054] The impact range index represents the ratio of actual impact to potential maximum impact. The actual impact is the sum of the number of affected people and the affected economy multiplied by the corresponding preset weight coefficient. It should be noted that the smaller the impact range index value, the higher the priority of the emergency response plan.

[0055] That is, the present invention can quickly generate an optimal set of solutions that take into account resources, time, and impact in emergencies, avoiding the subjectivity of manual decision-making while ensuring the feasibility and effectiveness of the solutions through multi-objective balance, providing scientific support for emergency command.

[0056] Embodiment 2: The dynamic response unit is used to perform flexibility dynamic adjustment analysis on the implementation process of the priority emergency plan to determine whether the priority emergency plan should be adjusted. The specific flexibility dynamic adjustment analysis process is as follows:

[0057] Based on the implementation process of the priority emergency plan, we obtain and pre-process the on-site multi-source data of the target area during the implementation process. The on-site multi-source data includes the number of newly trapped people, the remaining amount of supplies, the rate of supply consumption, the evacuation completion rate, the number of trapped people rescued, etc.

[0058] For example, the collected data can be filtered for noise (e.g., removing outliers caused by sensor failures), aligned in time and space (unifying the timestamps and geographic locations of different devices into a unified coordinate system), and format standardized to convert multimodal data such as text, images, and videos into structured features.

[0059] Based on the pre-processed on-site multi-source data, the values ​​of various parameters in the on-site multi-source data are obtained, and each parameter is discriminated and processed to obtain the discrimination processing results of each parameter, where the discrimination processing results include whether the parameter meets the standard or not.

[0060] For example, if the number of newly trapped people exceeds the preset threshold for the number of newly trapped people, it is determined that the standard is not met; if the number of newly trapped people does not exceed the preset threshold for the number of newly trapped people, it is determined that the standard is met; if the remaining amount of supplies falls within the preset range for remaining supplies, it is determined that the standard is met; if the remaining amount of supplies does not fall within the preset range for remaining supplies, it is determined that the standard is not met, and so on;

[0061] Obtaining the numerical deviation of the parameter corresponding to the judgment processing result that does not meet the standard, setting the numerical deviation of the parameter corresponding to the judgment processing result that does not meet the standard as the deviation degree, and performing judgment processing on whether the deviation degree exceeds a preset deviation degree threshold; if so, judging the corresponding parameter as an over-deviation parameter; if not, judging the corresponding parameter as a low-deviation parameter;

[0062] The sum of the over-deviation parameter and the under-deviation parameter is obtained, and the sum of the over-deviation parameter and the under-deviation parameter is set as the scheme deviation degree, and whether the scheme deviation degree exceeds the preset scheme deviation degree threshold is judged and processed. If so, an over-deviation signal is generated, and the back-end feedback unit is used to respond to the over-deviation signal, and then the priority emergency plan is adjusted as a whole, and multi-source data related to the emergency event in the target area is re-collected, the priority emergency plan is re-planned, or the preset weight coefficients of the resource allocation index, the response time index and the impact range index are readjusted. If not, a normal signal is generated;

[0063] When a regular signal is generated, a change curve of the over-deviation parameter and the under-deviation parameter is constructed based on the time series, and the change curve of the over-deviation parameter and the under-deviation parameter is input into the pre-set AI parameter adjustment model to obtain a single parameter adjustment plan as output. The back-end feedback unit is used to respond to the single parameter adjustment plan, and the single parameter adjustment plan is used to adjust the parameters to improve the flexibility and effectiveness of the emergency response and minimize the negative impact of the emergency;

[0064] For example, by leveraging the powerful analysis and learning capabilities of AI algorithms, we can process, train, and verify a large number of historical excess deviation parameter change curves to build an AI parameter adjustment model.

[0065] Based on the single parameter adjustment scheme, the reduction rate of the deviation degree of the over-deviation parameter and the under-deviation parameter is obtained, and the reduction rate is compared with the preset reduction rate range. If the reduction rate falls within the preset reduction rate range, a valid signal is generated; if the reduction rate does not fall within the preset reduction rate range, an invalid signal is generated. The back-end feedback unit is used to respond to the invalid signal, immediately execute the preset early warning operation corresponding to the invalid signal, and further adjust the single parameter adjustment scheme corresponding to the invalid signal to ensure the effectiveness of the single parameter adjustment scheme;

[0066] When a valid signal is generated, a reduction rate change curve is constructed based on the time series, and a preset reduction rate change curve is obtained at the same time. The difference between the reduction rate change curve and the preset reduction rate change curve is set as the continuous fit, and whether the continuous fit is less than the preset continuous fit threshold is judged. If so, a stable signal is generated; if not, an alarm signal is generated. The back-end feedback unit is used to respond to the stable signal or the alarm signal, and immediately display the preset warning text corresponding to the stable signal or the alarm signal, so as to adjust the preset AI parameter adjustment model based on the feedback information to ensure the continuous stability effect of the single parameter adjustment scheme output by the preset AI parameter adjustment model.

[0067] Example 3: The information management unit is used to perform data sensitivity identification and management level classification processing on the emergency response data related to the sudden incident in the target area, and obtain high-level encryption, medium-level encryption and low-level encryption. The specific data sensitivity identification and management level classification processing process is as follows:

[0068] Obtain emergency response data related to emergencies in the target area. Emergency response data includes text information and image information. Text information includes ID numbers, medical records, etc., and image information includes pictures of military facilities and facial images.

[0069] Use existing technologies (BERT text classification model, CNN image recognition model) to process the emergency response data for sensitive information, obtain the processed emergency response data, and set the processed emergency response data as data to be managed;

[0070] The amount of data to be managed is obtained, as is the difficulty of data leakage recovery for the data to be managed. The difficulty of data leakage recovery represents the combination of the normalized values ​​of data recovery cost and data recovery time. The data leakage probability of the data to be managed is also obtained. The data leakage probability represents the ratio of the number of historical data leakages to the total number of storage times of the data storage device to be managed.

[0071] The weight coefficient w1 of the data volume of the data to be managed, the weight coefficient w2 of the difficulty of leakage recovery and the weight coefficient w3 of the probability of data leakage are obtained, and w1, w2 and w3 are all greater than zero. The sum of the data volume of the data to be managed, the difficulty of leakage recovery and the probability of data leakage multiplied by w1, w2 and w3 is set as the data protection requirement value, and the data protection requirement value is judged. If the data protection requirement value is greater than the maximum value in the preset data protection requirement value range, it is determined to be high-level encryption. If the data protection requirement value belongs to the preset data protection requirement value range, it is determined to be medium-level encryption. If the data protection requirement value is less than the minimum value in the preset data protection requirement value range, it is determined to be low-level encryption. The back-end feedback unit is used to respond to high-level encryption, medium-level encryption and low-level encryption, and perform different degrees of encryption management on the data to be managed according to the encryption level, that is, to realize the automatic sensitive discovery and intelligent encryption management classification of emergency response data, which significantly improves the data security protection capability during the emergency response process;

[0072] In summary, the present invention can quickly generate scientific and reasonable emergency response decision-making plans through multi-source data collection and AI algorithm model analysis, and at the same time, it is accompanied by intelligent plan optimization analysis under comprehensive evaluation of resource allocation, response performance and impact range, that is, through multi-objective balance to ensure the feasibility and effectiveness of the plan, provide scientific support for emergency command, and track the implementation of the plan and provide feedback on the results, so as to make overall dynamic adjustments to the plan or targeted adjustments to single parameters to improve the flexibility and effectiveness of the emergency response and minimize the negative impact of emergencies. At the same time, the adjusted parameters are continuously evaluated to ensure the feasibility of adjustment prevention and the accuracy of the adjustment model, and sensitive identification and management level division are carried out from the perspective of data security, that is, by automatically identifying sensitive data in the emergency response system and performing risk classification, a basis is provided for subsequent data management, which significantly improves the data security protection capabilities during the emergency response process.

[0073] The threshold is set for result comparison and analysis in order to determine whether it is good or bad. The value of the threshold is set based on a combination of large-scale model analysis of sample data and manual experience to enter and store it. It can also be appropriately adjusted based on seasonal or common sense influencing conditions.

[0074] The size of the coefficient is to quantify each parameter to obtain a specific numerical value, which is convenient for subsequent comparison. The size of the coefficient depends on the amount of sample data and the preliminary setting of the corresponding operating coefficient for each set of sample data by technical personnel in this field; as long as it does not affect the proportional relationship between the parameter and the quantized value.

[0075] The above description is only a preferred specific embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any technician familiar with the technical field, within the technical scope disclosed by the present invention, who makes equivalent replacements or changes based on the technical solution and inventive concept of the present invention, should be covered by the scope of protection of the present invention.

Claims

1. AI-based emergency response intelligent decision-making and data security assurance system, characterized by: Including AI emergency management center, emergency response unit, dynamic response unit, information management unit and back-end feedback unit; The AI ​​emergency management center is used to retrieve and store multi-source data related to emergencies in the target area; The emergency response unit is used to retrieve multi-source data and conduct emergency response feedback analysis, and obtain a priority emergency plan based on the obtained plan priority coefficient; The dynamic response unit is used to analyze the flexibility of the implementation process of the priority emergency plan, determine whether the priority emergency plan should be adjusted, and obtain an over-deviation signal or a single parameter adjustment plan; Further analyze the control effect of the single parameter adjustment scheme, perform discrimination processing on the obtained reduction rate, and obtain an invalid signal or a valid signal. At the same time, based on the valid signal, perform discrimination processing on the obtained continuous fit degree, and obtain a stable signal or an alarm signal. The information management unit is used to perform data sensitivity identification and management level classification processing on the emergency response data related to the sudden incidents in the target area, and obtain high-level encryption, medium-level encryption and low-level encryption; The analysis process of the dynamic response unit is as follows: Based on the implementation process of the priority emergency plan, obtain the on-site multi-source data of the target area during the implementation process and pre-process the on-site multi-source data; Based on the pre-processed on-site multi-source data, the values ​​of various parameters in the on-site multi-source data are obtained, and each parameter is discriminated and processed to obtain the discrimination processing results of each parameter, where the discrimination processing results include whether the parameter meets the standard or not. Obtaining the numerical deviation of the parameter corresponding to the judgment processing result that does not meet the standard, setting the numerical deviation of the parameter corresponding to the judgment processing result that does not meet the standard as the deviation degree, and performing judgment processing on whether the deviation degree exceeds a preset deviation degree threshold; if so, judging the corresponding parameter as an over-deviation parameter; if not, judging the corresponding parameter as a low-deviation parameter; The sum of the over-deviation parameter and the under-deviation parameter is obtained, and the sum of the over-deviation parameter and the under-deviation parameter is set as the scheme deviation degree. It is then determined whether the scheme deviation degree exceeds a preset scheme deviation degree threshold. If so, an over-deviation signal is generated; if not, a normal signal is generated. When a regular signal is generated, the variation curves of the excess deviation parameter and the low deviation parameter are constructed based on the time series, and the variation curves of the excess deviation parameter and the low deviation parameter are input into the pre-set AI parameter adjustment model to obtain the output of a single parameter adjustment plan; Based on a single parameter adjustment scheme, the reduction rate of the deviation degree of the over-deviation parameter and the under-deviation parameter is obtained, and the reduction rate is compared with the preset reduction rate range. If the reduction rate falls within the preset reduction rate range, a valid signal is generated; if the reduction rate does not fall within the preset reduction rate range, an invalid signal is generated.

2. The AI-based emergency response intelligent decision-making and data security assurance system according to claim 1 is characterized in that: The analysis process of the emergency response unit is as follows: S1: Real-time acquisition of multi-source data related to emergencies in the target area; S2: Preprocessing the collected multi-source data related to emergencies in the target area; S3: Inputting the initial multi-source data into a pre-set AI emergency response decision model to obtain multiple emergency response plans output by the pre-set AI emergency response decision model; S4: Obtain basic information of each emergency response plan, including a resource allocation indicator, a response time indicator, and an impact range indicator; obtain preset weight coefficients of the resource allocation indicator, the response time indicator, and the impact range indicator; and set the sum of the resource allocation indicator, the response time indicator, and the impact range indicator multiplied by the corresponding preset weight coefficients as the plan priority coefficient; S5: Sort the obtained scheme priority coefficients in ascending order, and based on the sorting result, set the emergency response scheme corresponding to the minimum scheme priority coefficient as the priority emergency scheme.

3. The AI-based emergency response intelligent decision-making and data security assurance system according to claim 2 is characterized in that: The resource allocation index represents the sum of the redundant resources and the resource gap, divided by the total resource input; the response time index represents the ratio between the actual response time and the maximum allowable response time; the impact range index represents the ratio between the actual impact and the potential maximum impact, where the actual impact represents the sum of the number of affected people and the affected economy multiplied by the corresponding preset weight coefficient.

4. The AI-based emergency response intelligent decision-making and data security assurance system according to claim 1 is characterized in that: When a valid signal is generated, a reduction rate change curve is constructed based on the time series, and a preset reduction rate change curve is obtained at the same time. The difference between the reduction rate change curve and the preset reduction rate change curve is set as the continuous fit, and whether the continuous fit is less than the preset continuous fit threshold is judged. If so, a stable signal is generated; if not, an alarm signal is generated.

5. The AI-based emergency response intelligent decision-making and data security assurance system according to claim 1 is characterized in that: The data sensitivity identification and management level classification process is as follows: Obtain emergency response data related to emergencies in the target area, including text information and image information; Performing sensitive information processing on the emergency response data to obtain the emergency response data after sensitive information processing, and setting the emergency response data after sensitive information processing as data to be managed; The amount of data to be managed is obtained, as is the difficulty of data leakage recovery for the data to be managed. The difficulty of data leakage recovery represents the combination of the normalized values ​​of data recovery cost and data recovery time. The data leakage probability of the data to be managed is also obtained. The data leakage probability represents the ratio of the number of historical data leakages to the total number of storage times of the data storage device to be managed. The weight coefficient w1 of the data volume of the data to be managed, the weight coefficient w2 of the difficulty of leakage recovery, and the weight coefficient w3 of the data leakage probability are obtained, and w1, w2, and w3 are all greater than zero. The sum of the data volume of the data to be managed, the difficulty of leakage recovery, and the probability of data leakage multiplied by w1, w2, and w3 is set as the data protection requirement value, and the data protection requirement value is discriminated and processed to obtain high-level encryption, intermediate encryption, or low-level encryption.

Images