Fire rescue resource allocation system based on big data

Abstract

The application discloses a fire rescue resource allocation system based on big data, comprising: a data acquisition and processing module: acquiring various fire-related factor data of a fire site, and acquiring fire resource replenishment point big data; and a dynamic fire vector calculation module related to the technical field of data analysis and processing, wherein the fire range, fire size, fire spread speed, burning material type and surrounding environment information are brought into dynamic fire vector calculation, the weight of each factor is determined through an association rule mining algorithm and an entropy weight method, a weighted average method is adopted to fuse the first weight and the second weight to obtain a comprehensive weight, the different roles of each factor in the fire are fully considered, the values of the fire range, the fire size and the fire spread speed are predicted by means of a time series analysis method, and the comprehensive weight is dynamically adjusted according to the prediction result, so that the dynamic fire vector calculation is more in line with the actual situation, and the judgment accuracy is improved.

Description

Technical Field

[0001] This invention relates to the field of data analysis and processing technology, and in particular to a fire rescue resource allocation system based on big data. Background Technology

[0002] Fire and rescue resources refer to the sum of all human, material, and information resources used for rescue and firefighting operations in the event of an emergency such as a fire.

[0003] Publication number CN113744108A discloses a smart fire control method and system based on big data. It uses big data to determine and disseminate firsthand fire information, enabling the rational allocation of fire rescue resources; it also integrates alternative fire resources based on big data.

[0004] However, the above application still has the following problems: The above application simply combines the fire range and fire intensity into a dynamic fire vector, and the calculation formula used only considers the square root of the sum of the squares of the two factors, without fully considering the complex characteristics of fire, such as the fire spread speed and the characteristics of burning materials, which may lead to an inaccurate judgment of the severity of the fire and affect the rational allocation of rescue forces. Summary of the Invention

[0005] To address the technical problems existing in the background art, this invention proposes a fire rescue resource allocation system based on big data.

[0006] This invention proposes a big data-based fire and rescue resource allocation system, comprising:

[0007] Data acquisition and processing module: Acquires data on various fire-related factors at the fire location and obtains big data on fire resource replenishment points;

[0008] Fire-related factors include the fire's extent, size, spread rate, type of combustible material, and surrounding environmental information, including topography, vegetation cover, and water distribution.

[0009] Dynamic fire vector calculation module: Cleans, denoises, and standardizes the historical fire-related factors data obtained from the data acquisition and processing module.

[0010] Based on the association rule mining algorithm, different weights are assigned to each fire-related factor. The weights of each fire-related factor obtained in this study are the first weights.

[0011] The entropy weights of various fire-related factors are calculated using the entropy weight method, and the resulting entropy weights are the second weights.

[0012] The predicted values ​​of fire range, fire size, and fire spread rate are obtained by using time series analysis to predict the fire range, fire size, and fire spread rate over a future period.

[0013] A weighted average is used to combine the first and second weights to obtain the comprehensive weight;

[0014] The weights of fire extent, fire size, and fire spread rate in the comprehensive weighting are adjusted based on the prediction results of the time series model.

[0015] The standardized data of each fire-related factor currently acquired are multiplied by the comprehensive weight and summed to construct a dynamic fire vector calculation model. The dynamic fire vector calculation model is used to calculate the dynamic fire vector value.

[0016] The rescue resource allocation decision module compares the dynamic fire vector value calculated in the dynamic fire vector calculation module with a preset threshold. If the dynamic fire vector value is greater than the preset threshold, it generates instructions for multiple fire brigades to reinforce the fire location and sends them to multiple fire and rescue brigades; otherwise, it generates instructions for a single fire brigade based on the principle of proximity and sends them to a single fire and rescue team.

[0017] Based on big data on fire resource replenishment points, at least two fire resource replenishment points within a preset range of the fire location are identified, replenishment routes for each replenishment point are planned, and material information for each replenishment point is sent to the fire and rescue team in list form to facilitate the fire commander's selection of appropriate replenishment points.

[0018] Preferably, it also includes an information interaction module, which is used for real-time information interaction between the fire and rescue command center and the fire and rescue team and fire resource replenishment points.

[0019] Preferably, in the rescue resource allocation decision module, when sending instruction information to the fire and rescue team, detailed data on fire-related factors and dynamic fire vector values ​​are also sent simultaneously.

[0020] Preferably, in the data acquisition and processing module, the collected missing data is processed using mean imputation, interpolation, or a machine learning-based missing value prediction method.

[0021] Preferably, in the dynamic fire vector calculation module, the entropy weights of various fire-related factors are calculated using the entropy weight method, as follows:

[0022] For the standardized data on fire extent, fire size, fire spread rate, type of combustible material, and surrounding environment, let the sample size be n:

[0023] The information entropy of each fire-related factor is calculated using the entropy weight method formula. : ,in , , Let be the standardized value of the j-th fire-related factor in the i-th sample;

[0024] Calculate the entropy weights of each indicator based on information entropy. : , where m is the number of fire-related factors.

[0025] Preferably, in the dynamic fire vector calculation module, the fire range, fire size, and fire spread rate over a future period are predicted using time series analysis, resulting in predicted values ​​for the fire range, fire size, and fire spread rate, as follows:

[0026] Obtain historical data on the fire's extent, size, and spread rate at different times;

[0027] Perform a stationarity test on these data. If the data is not stationary, use difference operations to make it stationary.

[0028] Choose an autoregressive moving average model, train the model using historical data, determine the model parameters, and construct a time series model;

[0029] Time series models are used to predict the fire extent, fire size, and fire spread rate over a future period, resulting in predicted values ​​for these parameters.

[0030] Preferably, in the dynamic fire vector calculation module, the weights of fire extent, fire size, and fire spread rate in the comprehensive weighting are adjusted based on the time series model prediction results, as follows:

[0031] According to time series models, in the near future:

[0032] If the predicted fire extent is greater than the set threshold, the weight of the fire extent is increased by the amount by which the predicted fire extent increases relative to the threshold; correspondingly, the weights of the other fire-related factors are reduced according to their own proportions, so that the sum of the weights of all fire-related factors is equal to 1.

[0033] If the predicted fire size is greater than the set threshold, the weight of the fire size will be increased by the amount by which the predicted fire size increases relative to the threshold.

[0034] If the predicted fire spread rate is greater than the set threshold, the weight of the fire size will be increased by the amount by which the predicted fire spread rate increases relative to the threshold.

[0035] Preferably, the maximum weight increase is set to Mq%, and the maximum weight increase will only take effect when the weights of fire range, fire size and fire spread speed all need to be increased;

[0036] Let the weight increase for the fire extent be a%, the weight increase for the fire size be b%, and the weight increase for the fire spread rate be c%.

[0037] When a%+b%+c%≤Mq%, the weights of fire range, fire size, and fire spread rate increase as usual;

[0038] When a%+b%+c%>Mq%, at this time...

[0039] The weighting increase for the fire extent is: ;

[0040] The weighting increase for fire size is: ;

[0041] The weighting increase for the fire spread rate is: .

[0042] A big data-based method for allocating fire and rescue resources includes the following steps:

[0043] S1. Obtain data on various fire-related factors at the fire location through artificial observation satellites and pre-installed fire-fighting equipment at the fire location;

[0044] Fire-related factors include the extent of the fire, the size of the fire, the rate of fire spread, the type of burning material, and information about the surrounding environment.

[0045] Collect big data on fire resource replenishment points;

[0046] Big data on fire resource replenishment points includes the location and material information of fire resource replenishment points within a pre-defined range;

[0047] Surrounding environmental information includes topography, vegetation cover, and water body distribution;

[0048] S2. Clean, denoise, and standardize the data on various historical fire-related factors;

[0049] Based on the association rule mining algorithm, different weights are assigned to each fire-related factor. The weights obtained from the association rule mining are normalized so that the sum of the weights of each fire-related factor is equal to 1. The weights of each fire-related factor obtained in this case are the first weights.

[0050] The entropy weights of various fire-related factors are calculated using the entropy weight method, and the resulting entropy weights are used as the second weights.

[0051] The predicted values ​​of fire range, fire size, and fire spread rate are obtained by using time series analysis to predict the fire range, fire size, and fire spread rate over a future period.

[0052] A weighted average is used to combine the first and second weights to obtain the comprehensive weight;

[0053] The weights of fire extent, fire size, and fire spread rate in the comprehensive weighting are adjusted based on the prediction results of the time series model.

[0054] The standardized data of each fire-related factor currently acquired are multiplied by the comprehensive weight and summed to construct a dynamic fire vector calculation model. The dynamic fire vector calculation model is used to calculate the dynamic fire vector value.

[0055] S3. Compare the calculated dynamic fire vector value with a preset threshold. If the dynamic fire vector value is greater than the preset threshold, generate instructions for multiple fire brigades to reinforce the fire location and send them to multiple fire and rescue brigades. If the dynamic fire vector value is less than or equal to the preset threshold, generate instructions for a single fire brigade based on the principle of proximity and send them to a single fire and rescue team. Based on big data on fire resource replenishment points, obtain at least two fire resource replenishment points within a preset range of the fire location, plan the replenishment path for each replenishment point, and send the material information of each replenishment point to the fire and rescue team in list form.

[0056] The big data-based fire rescue resource allocation system proposed in this invention has the following beneficial technical effects:

[0057] 1. By incorporating information such as fire range, fire size, fire spread rate, type of burning material, and surrounding environment into the dynamic fire vector calculation, the weights of each factor are determined through association rule mining algorithms and entropy weighting. A weighted average is then used to fuse the first and second weights to obtain a comprehensive weight. This approach fully considers the different roles of each factor in the fire. Time series analysis is used to predict the values ​​of fire range, fire size, and fire spread rate. The comprehensive weights are dynamically adjusted based on the prediction results, making the dynamic fire vector calculation more realistic and improving the accuracy of judgment.

[0058] 2. Time series models are used to predict the fire's extent, size, and spread rate over a future period, yielding predicted values ​​for these parameters. These predictions are integrated into the dynamic fire vector calculation, enabling the dynamic fire vector to reflect not only the current fire situation but also its future development. This makes the dynamic fire vector more forward-looking, allowing rescue personnel more time to prepare response measures and allocate more resources, thus responding to fires more effectively. The weights of fire extent, size, and spread rate in the comprehensive weighting of the time series model prediction results are adjusted. By optimizing the weight allocation, the adaptability and accuracy of the dynamic fire vector calculation model are enhanced, providing stronger support for fire safety decision-making.

[0059] 3. In the dynamic fire vector calculation module, the weights of fire extent, fire size, and fire spread rate in the comprehensive weighting are adjusted based on the prediction results of the time series model. The approach of "increasing the weight of the fire extent by the amount of increase relative to the threshold if the increase in fire extent exceeds a set threshold" has the following advantages compared to "directly increasing the weight of the fire extent based on the increase in fire extent":

[0060] Directly increasing the weight based on the increase in the fire's extent may result in an excessively high increase in weight due to a large increase. This would severely compress the weights of other fire-related factors, making it difficult to fully reflect the actual fire situation. In contrast, increasing the weight based on the increase relative to a threshold allows for control over the magnitude of the weight increase by setting a threshold, ensuring a relatively balanced weighting of all factors.

[0061] A threshold is set, which acts as a "warning line." Weight adjustments are only made when the increase in fire extent exceeds this standard. This allows the model to focus on extent changes that truly have a significant impact on the fire situation, rather than reacting to all extent changes. Normal fluctuations in fire extent generally do not trigger weight adjustments; only when the fire shows an abnormal expansion trend exceeding the threshold is the fire extent weight increased, making it more targeted.

[0062] Adjusting weights directly based on the increase in value may make the model overly sensitive to fluctuations in fire extent data; even slight changes in data can drastically alter the weights and assessment results. However, adjusting weights based on thresholds ensures that data fluctuations within the threshold range do not affect the weights; adjustments are only made when the threshold is exceeded. This reduces the volatility of the dynamic fire vector calculation model and enhances its stability.

[0063] 4. By setting a maximum weight increase of Mq%, and only when the weights of fire range, fire size, and fire spread rate all need to be increased, the maximum weight increase will take effect. This design avoids drastically reducing the weights of combustible material type and surrounding environment information when adjusting the weights of fire-related factors. This is because although fire range, fire size, and fire spread rate have a significant impact on the development of the fire, combustible material type and surrounding environment information are also crucial. Ignoring them will weaken the model's adaptability to complex fire scenarios.

[0064] Additional aspects and advantages of the invention will be set forth in part in the description which follows, and in part will be obvious from the description, or may be learned by practice of the invention. Attached Figure Description

[0065] Figure 1 This is a schematic diagram of the system of the present invention;

[0066] Figure 2 This is a flowchart of the method of the present invention. Detailed Implementation

[0067] Embodiments of the present invention are described in detail below. Examples of these embodiments are illustrated in the accompanying drawings, wherein the same or similar symbols denote the same or similar elements or elements having the same or similar functions throughout. The embodiments described below with reference to the accompanying drawings are exemplary and are only used to explain the present invention, and should not be construed as limiting the present invention.

[0068] like Figure 1 The fire and rescue resource allocation system shown includes:

[0069] Data acquisition and processing module:

[0070] Data on various fire-related factors at the fire location are obtained through artificial observation satellites and pre-installed fire-fighting equipment at the fire location;

[0071] Fire-related factors include the fire's extent, size, spread rate, type of combustible material, and surrounding environmental information, including topography, vegetation cover, and water distribution.

[0072] Collect big data on fire resource replenishment points, which includes the location and material information of fire resource replenishment points within a preset range;

[0073] Pre-installed fire protection equipment includes temperature sensors, smoke concentration sensors, and heat source tracking sensors;

[0074] Artificial observation satellites carry high-resolution optical and infrared sensors to observe fires from space. Optical sensors capture images of the fire scene, and by analyzing these images, they measure the area and brightness of the flames to infer the size of the fire. Infrared sensors detect the intensity of thermal radiation at the fire scene; the intensity of thermal radiation is closely related to the size of the fire—higher radiation intensity generally indicates a larger fire. Furthermore, by comparing and analyzing images taken by the satellite at different times, the spread rate of the fire can be calculated, and the extent and direction of its spread over a period of time can be observed.

[0075] The satellite's onboard multispectral and hyperspectral sensors can monitor large areas. By analyzing the spectral characteristics of fire areas in satellite images, the type of burning material can be identified.

[0076] At the same time, satellite imagery can provide a wide range of environmental information around a fire, such as topography, vegetation cover, and water distribution, helping fire command departments to fully understand the surrounding environment and formulate reasonable rescue strategies.

[0077] Dynamic fire vector calculation module:

[0078] The historical fire-related data obtained from the data acquisition and processing module are cleaned, denoised, and standardized.

[0079] Based on the association rule mining algorithm, different weights are assigned to each fire-related factor. The weights obtained from the association rule mining are normalized so that the sum of the weights of each fire-related factor is equal to 1. The weights of each fire-related factor obtained in this case are the first weights.

[0080] In the association rule mining process, minimum support and minimum confidence thresholds are set, and the Apriori algorithm is run to mine association rules between the dynamic fire vector and various fire-related factors. Based on the mined association rules, the frequency and importance of each factor in the strong association rules are statistically analyzed. Factors with high frequency and significant impact on the dynamic fire vector are assigned higher weights; conversely, those with low frequency are assigned lower weights. In this way, the primary weights are determined for fire extent, fire size, fire spread rate, type of combustible material, and surrounding environmental information.

[0081] The entropy weights of various fire-related factors are calculated using the entropy weight method, and the resulting entropy weights are the second weights.

[0082] In the dynamic fire vector calculation module, the entropy weights of various fire-related factors are calculated using the entropy weight method, as follows:

[0083] For the standardized data on fire extent, fire size, fire spread rate, type of combustible material, and surrounding environment, let the sample size be n:

[0084] The information entropy of each fire-related factor is calculated using the entropy weight method formula. :

[0085] ,in , , Let be the standardized value of the j-th fire-related factor in the i-th sample;

[0086] Calculate the entropy weights of each indicator based on information entropy. : , where m is the number of fire-related factors, i.e., m=5;

[0087] Entropy weight This reflects the degree of dispersion of fire-related factor data. The greater the dispersion, the greater the entropy weight, and the greater the impact of the fire-related factor on the dynamic fire vector results.

[0088] The predicted values ​​of fire range, fire size, and fire spread rate are obtained by using time series analysis to predict the fire range, fire size, and fire spread rate over a future period.

[0089] In the dynamic fire vector calculation module, the time series analysis method is used to predict the fire range, fire size, and fire spread rate over a future period. The predicted values ​​for fire range, fire size, and fire spread rate are as follows:

[0090] Obtain historical data on the fire's extent, size, and spread rate at different times;

[0091] Perform a stationarity test on these data. If the data is not stationary, use difference operations to make it stationary.

[0092] Choose an autoregressive moving average model, train the model using historical data, determine the model parameters, and construct a time series model;

[0093] Time series models are used to predict the fire extent, fire size, and fire spread rate over a future period, resulting in predicted values ​​for these parameters.

[0094] A weighted average is used to combine the first and second weights to obtain the comprehensive weight;

[0095] The weights of fire extent, fire size, and fire spread rate in the comprehensive weighting are adjusted based on the prediction results of the time series model.

[0096] In the dynamic fire vector calculation module, the weights of fire extent, fire size, and fire spread rate in the comprehensive weighting are adjusted based on the time series model prediction results, as follows:

[0097] According to time series models, in the near future:

[0098] If the predicted fire extent is greater than the set threshold, the weight of the fire extent is increased by the amount by which the predicted fire extent increases relative to the threshold; correspondingly, the weights of the other fire-related factors are reduced according to their own proportions, so that the sum of the weights of all fire-related factors is equal to 1.

[0099] If the predicted fire size is greater than the set threshold, the weight of the fire size will be increased by the amount by which the predicted fire size increases relative to the threshold.

[0100] If the predicted fire spread rate is greater than the set threshold, the weight of the fire size will be increased by the amount by which the predicted fire spread rate increases relative to the threshold.

[0101] Time series models are used to predict the extent, size, and spread rate of a fire over a future period, yielding predicted values ​​for these parameters. These predictions are integrated into the dynamic fire vector calculation, enabling the dynamic fire vector to reflect not only the current fire situation but also its future development. This makes the dynamic fire vector more forward-looking, allowing rescue personnel more time to prepare response measures and allocate more resources, thus enabling more effective fire response. The weights of fire extent, size, and spread rate in the comprehensive weighting of the time series model prediction results are adjusted. By optimizing the weight allocation, the adaptability and accuracy of the dynamic fire vector calculation model are enhanced, providing stronger support for fire safety decision-making.

[0102] Based on the time series model, the predicted values ​​of fire range, fire size, and fire spread rate over a future period are obtained, and the comprehensive weights are dynamically adjusted. If the predicted values ​​of fire range, fire size, or fire spread rate are greater than the set threshold, the weights are increased according to the amount of increase of the predicted values ​​relative to the thresholds.

[0103] "If the increase in fire extent exceeds a set threshold, then the weight of the fire extent is increased by the amount of increase relative to the threshold" has the following advantages over "directly increasing the weight of the fire extent based on the increase in fire extent":

[0104] Directly increasing the weight based on the increase in the fire's extent may result in an excessively high increase in weight due to a large increase. This would severely compress the weights of other fire-related factors, making it difficult to fully reflect the actual fire situation. In contrast, increasing the weight based on the increase relative to a threshold allows for control over the magnitude of the weight increase by setting a threshold, ensuring a relatively balanced weighting of all factors.

[0105] A threshold is set, which acts as a "warning line." Weight adjustments are only made when the increase in fire extent exceeds this standard. This allows the model to focus on extent changes that truly have a significant impact on the fire situation, rather than reacting to all extent changes. Normal fluctuations in fire extent generally do not trigger weight adjustments; only when the fire shows an abnormal expansion trend exceeding the threshold is the fire extent weight increased, making it more targeted.

[0106] Adjusting weights directly based on the increase in value may make the model overly sensitive to fluctuations in fire extent data; even slight changes in data can drastically alter the weights and evaluation results. However, adjusting weights based on thresholds ensures that fluctuations within the threshold range do not affect the weights; adjustments are only made when the threshold is exceeded. This reduces model volatility and enhances model stability.

[0107] The maximum weight increase is set to Mq%, and this maximum weight increase takes effect when the weights for fire range, fire size, and fire spread rate all need to be increased.

[0108] Let the weight increase for the fire extent be a%, the weight increase for the fire size be b%, and the weight increase for the fire spread rate be c%.

[0109] When a%+b%+c%≤Mq%, the weights of fire range, fire size, and fire spread rate increase as usual;

[0110] When a%+b%+c%>Mq%, at this time...

[0111] The weighting increase for the fire extent is: ;

[0112] The weighting increase for fire size is: ;

[0113] The weighting increase for the fire spread rate is: ;

[0114] This design avoids drastically reducing the weights of combustible material type and surrounding environmental information when adjusting the weights of fire-related factors. While the fire's extent, size, and spread rate significantly impact its development, combustible material type and environmental information are also crucial; ignoring them weakens the model's adaptability to complex fire scenarios. Fire scenarios are complex and varied, and a single factor cannot fully reflect the reality. Significantly adjusting weight thresholds can cause the model to over-rely on certain factors, reducing its adaptability to different fire scenarios.

[0115] The standardized data of each fire-related factor currently acquired are multiplied by the comprehensive weight and summed to construct a dynamic fire vector calculation model, which is used to calculate the dynamic fire vector value.

[0116] By incorporating information such as fire extent, fire size, fire spread rate, type of burning material, and surrounding environment into the dynamic fire vector calculation, and determining the weights of each factor through association rule mining algorithms and entropy weighting, a weighted average method is used to fuse the first and second weights to obtain a comprehensive weight. This fully considers the different roles of each factor in the fire. Time series analysis is used to predict the values ​​of fire extent, fire size, and fire spread rate. The comprehensive weight is dynamically adjusted based on the prediction results, making the dynamic fire vector calculation more realistic and improving the accuracy of judgment.

[0117] Rescue resource allocation decision-making module:

[0118] The calculated dynamic fire vector value is compared with a preset threshold. If the dynamic fire vector value is greater than the preset threshold, instructions for multiple fire brigades to reinforce the fire location are generated and sent to multiple fire and rescue teams. If the dynamic fire vector value is less than or equal to the preset threshold, instructions for single-brigade rescue are generated based on the principle of proximity and sent to a single fire and rescue team.

[0119] Based on big data on fire resource replenishment points, at least two fire resource replenishment points within a preset range of the fire location are identified, replenishment routes for each replenishment point are planned, and material information for each replenishment point is sent to the fire and rescue team in list form to facilitate the fire commander's selection of appropriate replenishment points.

[0120] It also includes a data storage module, which stores various types of collected data, calculated weights, constructed models, and prediction results.

[0121] It also includes an information interaction module, which is used to enable real-time information exchange between the fire and rescue command center, fire and rescue teams, and fire resource replenishment points.

[0122] In the rescue resource allocation decision-making module, when sending instruction information to the fire and rescue team, detailed data on fire-related factors and dynamic fire vector values ​​are also sent simultaneously.

[0123] In the data acquisition and processing module, the collected missing data is processed using mean imputation, interpolation, or machine learning-based missing value prediction methods.

[0124] In the dynamic fire vector calculation module, the time series analysis method is used to predict the fire range, fire size, and fire spread rate over a future period. The predicted values ​​for fire range, fire size, and fire spread rate are as follows:

[0125] Obtain historical data on the fire's extent, size, and spread rate at different times;

[0126] Perform a stationarity test on these data. If the data is not stationary, use difference operations to make it stationary.

[0127] Choose an autoregressive moving average model, train the model using historical data, determine the model parameters, and construct a time series model;

[0128] Time series models are used to predict the fire extent, fire size, and fire spread rate over a future period, resulting in predicted values ​​for these parameters.

[0129] like Figure 2 The method for allocating fire and rescue resources based on big data, as shown, is characterized by the following steps:

[0130] S1. Obtain data on various fire-related factors at the fire location through artificial observation satellites and pre-installed fire-fighting equipment at the fire location;

[0131] Fire-related factors include the extent of the fire, the size of the fire, the rate of fire spread, the type of burning material, and information about the surrounding environment.

[0132] Collect big data on fire resource replenishment points;

[0133] Big data on fire resource replenishment points includes the location and material information of fire resource replenishment points within a pre-defined range;

[0134] Surrounding environmental information includes topography, vegetation cover, and water body distribution;

[0135] S2. Clean, denoise, and standardize the data on various historical fire-related factors;

[0136] Based on the association rule mining algorithm, different weights are assigned to each fire-related factor. The weights obtained from the association rule mining are normalized so that the sum of the weights of each fire-related factor is equal to 1. The weights of each fire-related factor obtained in this case are the first weights.

[0137] The entropy weights of various fire-related factors are calculated using the entropy weight method, and the resulting entropy weights are used as the second weights.

[0138] The predicted values ​​of fire range, fire size, and fire spread rate are obtained by using time series analysis to predict the fire range, fire size, and fire spread rate over a future period.

[0139] A weighted average is used to combine the first and second weights to obtain the comprehensive weight;

[0140] The weights of fire extent, fire size, and fire spread rate in the comprehensive weighting are adjusted based on the prediction results of the time series model.

[0141] The standardized data of each fire-related factor currently acquired are multiplied by the comprehensive weight and summed to construct a dynamic fire vector calculation model. The dynamic fire vector calculation model is used to calculate the dynamic fire vector value.

[0142] S3. Compare the calculated dynamic fire vector value with a preset threshold. If the dynamic fire vector value is greater than the preset threshold, generate instructions for multiple fire brigades to reinforce the fire location and send them to multiple fire and rescue brigades. If the dynamic fire vector value is less than or equal to the preset threshold, generate instructions for a single fire brigade based on the principle of proximity and send them to a single fire and rescue team. Based on big data on fire resource replenishment points, obtain at least two fire resource replenishment points within a preset range of the fire location, plan the replenishment path for each replenishment point, and send the material information of each replenishment point in a list format to the fire and rescue team so that the fire commander can select the appropriate replenishment point.

[0143] Furthermore, any content not described in detail in this specification is existing technology known to those skilled in the art.

[0144] In the embodiments provided by this invention, it should be understood that the disclosed system or method can be implemented in other ways. For example, the embodiments of the invention described above are merely illustrative; for instance, the division of modules is only a logical functional division, and there may be other division methods in actual implementation.

[0145] The modules described as separate components may or may not be physically separate. The components shown as modules may or may not be physical modules; they may be located in one place or distributed across multiple network modules. Some or all of the modules can be selected to achieve the purpose of this embodiment according to actual needs.

[0146] Furthermore, the functional modules in the various embodiments of the present invention can be integrated into one processing module, or each module can exist physically separately, or two or more modules can be integrated into one module. The integrated module can be implemented in hardware or in the form of hardware plus software functional modules.

[0147] For those skilled in the art, it is obvious that the present invention is not limited to the details of the above exemplary embodiments, and that the present invention can be implemented in other specific forms without departing from the basic characteristics of the present invention.

[0148] The above are merely preferred embodiments of the present invention, but the scope of protection of the present invention is not limited thereto. Any equivalent substitutions or modifications made by those skilled in the art within the scope of the technology disclosed in the present invention, based on the technical solution and inventive concept of the present invention, should be covered within the scope of protection of the present invention.

Claims

1. A fire and rescue resource allocation system based on big data, characterized in that, include: Data acquisition and processing module: Acquires data on various fire-related factors at the fire location and obtains big data on fire resource replenishment points; Dynamic fire vector calculation module: used to calculate dynamic fire vector values; The rescue resource allocation decision module compares the dynamic fire vector value calculated in the dynamic fire vector calculation module with a preset threshold. If the dynamic fire vector value is greater than the preset threshold, it generates instructions for multiple fire brigades to reinforce the fire location and sends them to multiple fire and rescue brigades; otherwise, it generates instructions for a single fire brigade based on the principle of proximity and sends them to a single fire and rescue team. Based on big data on fire resource replenishment points, at least two fire resource replenishment points within a preset range of the fire location are identified, replenishment routes for each replenishment point are planned, and material information for each replenishment point is sent to the fire and rescue team in a list format to facilitate the fire commander's selection of appropriate replenishment points. The historical fire-related data obtained from the data acquisition and processing module are cleaned, denoised, and standardized. Based on the association rule mining algorithm, different weights are assigned to each fire-related factor. The weights of each fire-related factor obtained in this case are the first weights. The entropy weights of various fire-related factors are calculated using the entropy weight method, and the resulting entropy weights are the second weights. The predicted values ​​of fire range, fire size, and fire spread rate are obtained by using time series analysis to predict the fire range, fire size, and fire spread rate over a future period. A weighted average is used to combine the first and second weights to obtain the comprehensive weight; The weights of fire extent, fire size, and fire spread rate in the comprehensive weighting are adjusted based on the prediction results of the time series model. The standardized data of each fire-related factor currently acquired are multiplied by the comprehensive weight and summed to construct a dynamic fire vector calculation model. The dynamic fire vector calculation model is used to calculate the dynamic fire vector value. In the dynamic fire vector calculation module, the weights of fire extent, fire size, and fire spread rate in the comprehensive weighting are adjusted based on the time series model prediction results, as follows: According to time series models, in the near future: If the predicted fire extent is greater than the set threshold, the weight of the fire extent is increased by the amount by which the predicted fire extent increases relative to the threshold; correspondingly, the weights of the other fire-related factors are reduced according to their own proportions, so that the sum of the weights of all fire-related factors is equal to 1. If the predicted fire size is greater than the set threshold, the weight of the fire size will be increased by the amount by which the predicted fire size increases relative to the threshold. If the predicted fire spread rate is greater than the set threshold, the weight of the fire size will be increased by the amount by which the predicted fire spread rate increases relative to the threshold. The maximum weight increase is set to Mq%, and this maximum weight increase takes effect when the weights for fire range, fire size, and fire spread rate all need to be increased. Let the weight increase for the fire extent be a%, the weight increase for the fire size be b%, and the weight increase for the fire spread rate be c%. When a%+b%+c%≤Mq%, the weights of fire range, fire size, and fire spread rate increase as usual; When a%+b%+c%>Mq%, at this time... The weighting increase for the fire extent is: ; The weighting increase for fire size is: ; The weighting increase for the fire spread rate is: .

2. The big data-based fire and rescue resource allocation system according to claim 1, characterized in that, It also includes an information interaction module, which is used for real-time information exchange between the fire and rescue command center and the fire and rescue team and fire resource replenishment points.

3. The big data-based fire and rescue resource allocation system according to claim 1, characterized in that, In the rescue resource allocation decision-making module, when sending instruction information to the fire and rescue team, detailed data on fire-related factors and dynamic fire vector values ​​are also sent simultaneously.

4. The big data-based fire and rescue resource allocation system according to claim 1, characterized in that, In the dynamic fire vector calculation module, the entropy weights of various fire-related factors are calculated using the entropy weight method, as follows: For the standardized data on fire extent, fire size, fire spread rate, type of combustible material, and surrounding environment, let the sample size be n: The information entropy of each fire-related factor is calculated using the entropy weight method formula. : ,in , , Let be the standardized value of the j-th fire-related factor in the i-th sample; Calculate the entropy weights of each indicator based on information entropy. : , where m is the number of fire-related factors.

5. The big data-based fire and rescue resource allocation system according to claim 4, characterized in that, In the dynamic fire vector calculation module, the time series analysis method is used to predict the fire range, fire size, and fire spread rate over a future period. The predicted values ​​for fire range, fire size, and fire spread rate are as follows: Obtain historical data on the fire's extent, size, and spread rate at different times; Perform a stationarity test on these data. If the data is not stationary, use difference operations to make it stationary. Choose an autoregressive moving average model, train the model using historical data, determine the model parameters, and construct a time series model; Time series models are used to predict the fire extent, fire size, and fire spread rate over a future period, resulting in predicted values ​​for these parameters.

Images