Urban suburb fire emergency prevention internet of things large model system, method and medium

Abstract

The application provides a city suburb fire emergency prevention Internet of Things large model system, method and medium, relates to the fire prevention field, and the emergency supervision and management platform in the system is configured to execute every preset period: based on temperature distribution, crowd distribution, region type and historical danger data of a plurality of preset regions, a plurality of fire risk coefficients of the plurality of preset regions are determined; based on a plurality of fire risk coefficients and a plurality of risk thresholds, a target region is determined; and based on the fire risk coefficient of the target region, a monitoring instruction including a sampling point and a sampling amount is generated, a drone of an emergency supervision object platform is controlled to move to the sampling point, and monitoring data corresponding to the sampling amount is acquired at the sampling point. The application can quickly evaluate the fire risk coefficients of different regions, thereby determining the regions that need to be monitored in key, and collecting relevant data in the regions in time by using the drone, to provide data reference for fire emergency.

Description

Technical Field

[0001] This manual pertains to the field of fire prevention, and in particular to the IoT-based large-scale model system, methods, and media for fire emergency prevention in urban and suburban areas. Background Technology

[0002] With the acceleration of urbanization, the need for fire prevention in suburban areas is becoming increasingly urgent. Suburbs typically have high vegetation coverage, complex terrain, and limited rescue resources, making them prone to large-scale ecological and property damage once a fire breaks out.

[0003] Currently, emergency response to suburban fires relies on static monitoring networks comprised of fixed-location temperature sensors, smoke detectors, and surveillance cameras. These systems suffer from coverage blind spots, struggle to capture dynamic fire conditions (such as wildfire spread) in real time, and are susceptible to environmental interference (such as heavy fog or low nighttime illumination), leading to frequent false alarms or missed alarms. Fire monitoring through regular manual inspections and satellite thermal imaging suffers from significant response delays and struggles to pinpoint early fire sources.

[0004] To address the aforementioned issues, there is an urgent need for a large-scale IoT model system, methods, and media for fire emergency prevention in urban and suburban areas. This system could rapidly assess the fire risk coefficients of different regions, thereby identifying areas requiring focused monitoring, and utilize drones to collect relevant data within these areas in a timely manner, providing data references for fire emergency response. Summary of the Invention

[0005] The invention includes a large-scale IoT model system for emergency fire prevention in urban suburbs. The system comprises an emergency monitoring user platform, an emergency monitoring service platform, an emergency monitoring management platform, an emergency monitoring sensor network platform, and an emergency monitoring object platform. The emergency monitoring management platform is configured to execute the following every preset period: determine multiple fire risk coefficients for multiple preset areas based on temperature distribution, population distribution, area type, and historical hazard data; determine a target area based on the multiple fire risk coefficients and multiple risk thresholds; and generate monitoring instructions including sampling points and sampling quantities based on the fire risk coefficients of the target area. These instructions are then sent to the emergency monitoring object platform via the emergency monitoring sensor network platform to control a drone on the emergency monitoring object platform to move to the sampling point and acquire monitoring data corresponding to the sampling quantity at that point.

[0006] The invention includes a method for emergency fire prevention in suburban areas, executed by an emergency monitoring and management platform within a large-scale IoT model system for suburban fire emergency prevention at preset intervals. The method includes: determining multiple fire risk coefficients for multiple preset areas based on temperature distribution, population distribution, area type, and historical incident data; determining a target area based on the multiple fire risk coefficients and multiple risk thresholds; and generating monitoring instructions including sampling points and sampling quantities based on the fire risk coefficients of the target area, and sending the monitoring instructions to an emergency monitoring target platform via an emergency monitoring sensor network platform to control a drone on the emergency monitoring target platform to move to the sampling point and acquire monitoring data corresponding to the sampling quantity at the sampling point.

[0007] The invention includes a computer-readable storage medium that stores computer instructions. When a computer reads the computer instructions from the storage medium, the computer executes a method for emergency prevention of fires in urban suburbs.

[0008] The beneficial effects of the present invention include, but are not limited to: (1): Through the Internet of Things large model system for urban suburban fire emergency prevention, an information operation loop can be formed between various functional platforms, and coordinated and regular operation can be carried out under the unified management of the emergency supervision and management platform, so as to realize the informatization and intelligence of urban fire emergency prevention; (2): Through temperature distribution, population distribution, regional type and historical risk data, the fire risk coefficient of the preset area can be quickly assessed, so as to select the target area that needs to be further screened by drones from multiple preset areas for key monitoring; (3): Based on the vegetation abnormality data, the severity of the fire in the target area can be estimated, and the delivery dose and hovering height can be determined. The delivery dose and hovering height that can be more effective in extinguishing the fire can be determined according to the possible development of the fire, so as to ensure the fire extinguishing effect and reduce the wear and tear of the drone's service life. Attached Figure Description

[0009] This specification will be further described by way of exemplary embodiments, which will be described in detail with reference to the accompanying drawings. These embodiments are not limiting; in these embodiments, the same reference numerals denote the same structures, wherein:

[0010] Figure 1 This is an exemplary structural diagram of a large-scale IoT model system for urban suburban fire emergency prevention, as shown in some embodiments of this specification.

[0011] Figure 2 This is an exemplary flowchart of an urban suburban fire emergency prevention method according to some embodiments of this specification;

[0012] Figure 3This is an exemplary schematic diagram illustrating the determination of whether a fire exists in a target area using an anomaly identification model, based on some embodiments of this specification.

[0013] Figure 4 This is an exemplary flowchart illustrating the generation of delivery instructions according to some embodiments of this specification. Detailed Implementation

[0014] The accompanying drawings used in the description of the embodiments will be briefly introduced below. The drawings do not represent all embodiments.

[0015] The embodiments in this specification are merely illustrative and not intended to limit the scope of this specification. Various modifications and changes that can be made under the guidance of this specification by those skilled in the art are still within the scope of this specification. Furthermore, certain features, structures, or characteristics in one or more embodiments of this specification can be appropriately combined.

[0016] Figure 1 This is an exemplary structural diagram of a large-scale Internet of Things (IoT) system for emergency fire prevention in urban and suburban areas, as shown in some embodiments of this specification.

[0017] In some embodiments, such as Figure 1 As shown, the large-scale Internet of Things (IoT) system 100 for emergency fire prevention in urban suburbs may include an emergency monitoring user platform 110, an emergency monitoring service platform 120, an emergency monitoring management platform 130, an emergency monitoring sensor network platform 140, and an emergency monitoring object platform 150.

[0018] Emergency monitoring user platform 110 refers to a platform for interaction with users (such as regulatory personnel). In some embodiments, emergency monitoring user platform 110 includes terminal devices. For example, terminal devices may include mobile devices, tablet computers, and consoles.

[0019] Emergency monitoring service platform 120 refers to a platform used for receiving and transmitting data and / or information. In some embodiments, emergency monitoring service platform 120 is configured as a server or processor, etc. Emergency monitoring service platform 120 can interact bidirectionally with emergency monitoring user platform 110 and emergency monitoring management platform 130.

[0020] The emergency monitoring and management platform 130 refers to a comprehensive management platform that manages and coordinates the connections and collaboration between multiple platforms. In some embodiments, the emergency monitoring and management platform 130 is configured as a server or processor, etc. The emergency monitoring and management platform 130 can interact bidirectionally with the emergency monitoring service platform 120 and the emergency monitoring sensor network platform 140. For example, the emergency monitoring and management platform 130 can obtain monitoring data from the emergency monitoring target platform 150 through the emergency monitoring sensor network platform 140; or, for another example, the emergency monitoring and management platform 130 can issue control instructions (such as monitoring instructions) to the emergency monitoring target platform 150 through the emergency monitoring sensor network platform 140 to control the emergency monitoring target platform 150 to send monitoring instructions to the emergency monitoring target platform through the emergency monitoring sensor network platform, thereby controlling the drone of the emergency monitoring target platform to move to the sampling point and obtain monitoring data corresponding to the sampling quantity at the sampling point.

[0021] The emergency monitoring sensor network platform 140 refers to a platform used for the comprehensive management of sensor information. In some embodiments, the emergency monitoring sensor network platform 140 is configured as a communication network or gateway, etc.

[0022] The emergency monitoring object platform 150 refers to a platform that generates sensor information and executes control commands. In some embodiments, the emergency monitoring object platform 150 includes drones, image acquisition devices, temperature sensors, and fire alarms, etc. The emergency monitoring object platform 150 can interact bidirectionally with the emergency monitoring sensor network platform 140.

[0023] Drones can be used to sample and acquire monitoring data from target areas. In some embodiments, the drone is equipped with a camera and gas sensors, among other things. For more information on target areas and monitoring data, please see [link to relevant documentation]. Figure 2 And its related descriptions.

[0024] Image acquisition devices can be devices used to acquire images of a specific area, such as cameras or industrial cameras.

[0025] Temperature sensors can be used to collect the temperature within a specific area.

[0026] In some embodiments, image acquisition devices and temperature sensors may be deployed at one or more preset points within a preset area.

[0027] For more information on drones, image acquisition equipment, and temperature sensors, please see [link to relevant documentation]. Figure 2 And its related descriptions.

[0028] In some embodiments of this specification, the Urban Suburban Fire Emergency Prevention Internet of Things Large Model System 100 can form an information operation closed loop among various functional platforms and operate in a coordinated and regular manner under the unified management of the emergency supervision and management platform, thereby realizing the informatization and intelligentization of urban fire emergency prevention.

[0029] Figure 2 This is an exemplary flowchart illustrating an urban suburban fire emergency prevention method according to some embodiments of this specification. Figure 2 As shown, the process 200 of the urban suburban fire emergency prevention method includes the following steps. In some embodiments, the process 200 of the urban suburban fire emergency prevention method can be executed by an emergency monitoring and management platform at preset intervals. The preset intervals can be pre-set based on historical experience.

[0030] Step 210: Based on the temperature distribution, population distribution, area type, and historical incident data of multiple preset areas, determine multiple fire risk coefficients for multiple preset areas.

[0031] A preset area refers to an area where a fire may occur. Examples include warehouses, parks, or grasslands. In some embodiments, preset areas can be pre-defined based on historical experience. Each preset area corresponds to a set of temperature distribution, population distribution, area type, and historical hazard data; each preset area corresponds to a fire risk coefficient.

[0032] Temperature distribution is a parameter used to describe the temperature conditions at different locations within a preset area. In some embodiments, the temperature distribution includes the temperatures of multiple preset points within the preset area. These preset points can be pre-set based on experience.

[0033] In some embodiments, for a single preset area, the emergency monitoring and management platform can obtain the temperature distribution by acquiring the temperatures of multiple preset points within the preset area from the emergency monitoring object platform through the emergency sensor network platform. The emergency monitoring object platform can acquire the temperatures of multiple preset points by deploying multiple temperature sensors at multiple preset points within the preset area.

[0034] Crowd distribution is a parameter used to describe the crowd situation at different locations within a preset area. In some embodiments, crowd distribution includes the crowd density of multiple sub-regions within the preset area. These sub-regions can be pre-defined based on experience.

[0035] In some embodiments, for a single preset area, the emergency monitoring and management platform can obtain images acquired by image acquisition devices from the emergency monitoring object platform through the emergency sensor network platform, and identify the crowd density in multiple sub-regions of the image through image recognition algorithms (such as support vector machine algorithms or convolutional neural network models, etc.) to obtain the crowd distribution.

[0036] In some embodiments, the emergency monitoring and management platform can determine the monitoring period of the image acquisition device based on the visibility distribution of multiple preset areas.

[0037] Visibility distribution can reflect the distribution of visibility in a preset area over multiple preset monitoring periods. In some embodiments, the multiple preset monitoring periods may include multiple time periods of a day in historical time. The multiple preset monitoring periods can be preset based on historical experience; for example, the multiple preset monitoring periods may be 0-8 AM, 8-6 PM, and 6-12 AM of a day in historical time.

[0038] In some embodiments, for a single preset area, the emergency monitoring and management platform can statistically analyze the visibility of the preset area in historical data across multiple identical preset monitoring periods, and use the average visibility as the visibility within that preset monitoring period to obtain the visibility distribution. The emergency monitoring and management platform determines the visibility of a single preset area across multiple preset monitoring periods using the above method.

[0039] In some embodiments, the emergency monitoring and management platform can identify the edge sharpness of images acquired by image acquisition devices within a preset monitoring period using edge detection algorithms, and determine the visibility of a preset area within the preset monitoring period based on the edge sharpness. Higher edge sharpness corresponds to higher visibility. Edge detection algorithms may include, for example, the Canny edge detection algorithm.

[0040] The monitoring period refers to the time during which the image acquisition device collects monitoring data. In some embodiments, for a single preset area, the emergency monitoring and management platform can, based on the visibility distribution of the preset area, designate a preset monitoring period where the visibility is greater than a visibility threshold as the monitoring period for the image acquisition device within that preset area. The visibility threshold can be preset based on historical experience.

[0041] In some embodiments of this specification, since the image acquisition device acquires images with different accuracy under different visibility conditions, by setting the period with higher visibility as the monitoring period, the accuracy of the images acquired by the image acquisition device can be guaranteed, which in turn helps to improve the accuracy of the subsequently determined fire risk coefficient.

[0042] In some embodiments, when determining multiple fire risk coefficients for multiple preset areas, the emergency monitoring and management platform may also replace the temperature distribution and population distribution of the multiple preset areas with future temperature distribution and future population distribution.

[0043] In some embodiments, the emergency monitoring and management platform can determine the future temperature distribution and future population distribution of multiple preset areas based on the temperature distribution and population distribution of those areas. Each preset area corresponds to a set of future temperature distributions and future population distributions.

[0044] Future temperature distribution refers to the temperature distribution over a preset future time period. This preset future time period can be set in advance based on historical experience.

[0045] Future population distribution refers to the population distribution in a pre-defined future time period.

[0046] In some embodiments, the emergency monitoring and management platform can determine the future temperature distribution and future population distribution of multiple preset areas based on the temperature distribution and population distribution of multiple preset areas through a distribution prediction model.

[0047] In some embodiments, the distribution prediction model is a machine learning model, such as any one or a combination of a neural network (NN) model, a convolutional neural network (CNN) model, or other custom model structures.

[0048] In some embodiments, the inputs to the distribution prediction model include the temperature distribution and population distribution of a single preset area, and the outputs include the future temperature distribution and future population distribution of the single preset area.

[0049] In some embodiments, the emergency monitoring and management platform can obtain a distribution prediction model by training multiple first training samples with first labels using methods such as gradient descent. The first training samples include the sample temperature distribution and sample population distribution of a preset sample area in a first historical period, and the first label includes the temperature distribution and population distribution of the preset sample area in a second historical period. The first historical period is earlier than the second historical period.

[0050] In some embodiments, the first training sample and the first label can be obtained from historical data.

[0051] In some embodiments, the training process of the distribution prediction model includes: the emergency monitoring and management platform can input multiple first training samples into the initial distribution prediction model, construct a loss function based on the first label and the output of the initial distribution prediction model, and then iteratively update the parameters of the initial distribution prediction model based on the loss function. When the training conditions are met, the model training is completed, and a trained distribution prediction model is obtained. The training conditions may include the convergence of the loss function or the number of iterations reaching a threshold, etc.

[0052] In some embodiments of this specification, the future temperature distribution and population distribution are considered when determining the fire risk coefficient, which helps to determine a more accurate fire risk coefficient. A distribution prediction model can be used to determine a more accurate future temperature distribution and future population distribution.

[0053] The region type describes the type of a preset region. In some embodiments, the region type may include urban or suburban areas, etc. The region type can be obtained through user input.

[0054] Historical hazard data refers to data related to fires that have occurred within a pre-defined area. In some embodiments, historical hazard data includes historical temperature distribution, historical population distribution, and actual losses during each fire in the pre-defined area at historical times. Actual losses can be economic losses caused by the fire. Actual losses can be confirmed and labeled by technical personnel and then uploaded to the emergency monitoring and management platform.

[0055] The fire risk coefficient is an indicator used to comprehensively assess the probability of a fire occurring and the potential losses within a predetermined area. In some embodiments, a higher fire risk coefficient indicates a greater probability of a fire occurring and a greater potential loss.

[0056] In some embodiments, for a single preset area, the emergency monitoring and management platform can construct feature vectors based on historical data of temperature distribution, population distribution, and area type at multiple historical time points, resulting in multiple first-type vectors. The platform can also construct feature vectors based on historical fire incident data for that preset area, including historical temperature distribution, historical population distribution, and area type at each fire occurrence, resulting in multiple second-type vectors. Finally, the platform can construct feature vectors based on the current temperature distribution, population distribution, and area type, resulting in a third-type vector. Multiple historical time points can be pre-set based on historical experience. Neither the first-type nor the third-type vectors carry labels; the label for the second-type vectors is the actual loss from the historical fire incident data.

[0057] In some embodiments, the emergency monitoring and management platform uses a clustering algorithm to cluster the first, second, and third type vectors to determine at least one cluster center. For a cluster center containing a third type vector, the platform counts the number of second type vectors included in the cluster center and the total number of feature vectors. The ratio of the number of second type vectors to the total number of feature vectors is used as the probability of a fire occurring, and the average of the actual losses among the labels of all second type vectors is used as the potential loss. The clustering algorithm may include K-Means clustering, etc.

[0058] In some embodiments, the emergency monitoring and management platform can determine a fire risk coefficient based on the probability of a fire occurring and the potential losses. For example, the platform can normalize the probability of a fire occurring and the potential losses before weighted summation to obtain the fire risk coefficient. The weights for the probability of a fire occurring and the potential losses can be preset.

[0059] Step 220: Determine the target area based on multiple fire risk coefficients and multiple risk thresholds.

[0060] The risk threshold can be the maximum acceptable fire risk coefficient. In some embodiments, a preset area corresponds to a risk threshold. The risk thresholds for different preset areas can be the same or different. The risk threshold can be preset based on experience.

[0061] In some embodiments, the risk threshold can be negatively correlated with the amount of vegetation within a preset area; that is, the more vegetation there is, the lower the risk threshold. The amount of vegetation within the area can be determined by user input, or the emergency monitoring and management platform can obtain the amount of vegetation within the preset area by recognizing images captured by image acquisition devices using image recognition algorithms.

[0062] Understandably, the more vegetation in an area, the faster a fire can spread. In such cases, it is necessary to lower the risk threshold and respond to the fire in a timely manner to minimize losses.

[0063] The target area refers to the pre-defined area that needs to be monitored in multiple pre-defined areas.

[0064] In some embodiments, the emergency monitoring and management platform can determine the target area based on multiple fire risk coefficients and multiple risk thresholds. For example, the emergency monitoring and management platform can select a preset area with a fire risk coefficient greater than the corresponding risk threshold from among multiple preset areas as the target area.

[0065] Step 230: Based on the fire risk coefficient of the target area, generate monitoring instructions including sampling points and sampling quantities.

[0066] Monitoring instructions are commands used to guide drones in collecting monitoring data. In some embodiments, monitoring instructions include sampling points and sampling quantities. A sampling point refers to the location where monitoring data is collected. The sampling quantity refers to the amount of monitoring data collected.

[0067] The monitoring data can be data obtained after monitoring the target area. In some embodiments, the monitoring data may include at least one of image data (such as plant images) captured by the drone's camera and gas-related data (such as smoke concentration and the concentration of various substances in the air) obtained by the drone's gas sensor.

[0068] In some embodiments, the emergency monitoring and management platform can query the number and sampling volume of sampling points corresponding to the fire risk coefficient in a first preset table based on the fire risk coefficient of the target area, determine the locations of multiple sampling points according to the number of sampling points, and generate monitoring instructions based on the multiple sampling points and sampling volume. The emergency monitoring and management platform can set the locations of multiple sampling points within the target area in various ways according to the number of sampling points. These various methods include setting them at equal intervals, etc.

[0069] In some embodiments, the first preset table can be pre-set based on historical experience, including the correspondence between multiple fire risk coefficients and the number of sampling points and the amount of sampling.

[0070] In some embodiments, the emergency monitoring management platform sends the monitoring instruction to the emergency monitoring object platform through the emergency monitoring sensor network platform to control the drone of the emergency monitoring object platform to move to the sampling point and acquire monitoring data corresponding to the sampling amount at the sampling point.

[0071] In some embodiments of this specification, the fire risk coefficient of a preset area can be quickly assessed by using temperature distribution, population distribution, area type, and historical hazard data. This allows for the selection of target areas from multiple preset areas that require further screening by drones for focused monitoring. Simultaneously, the fire risk coefficient can determine more appropriate monitoring instructions, enabling the drone to collect sufficient data from the target area, avoiding insufficient data and facilitating timely response to fires in the target area.

[0072] In some embodiments, the emergency monitoring and management platform can determine air-related data and vegetation-related data of the target area based on monitoring data, and determine whether there is a fire in the target area based on the air-related data and vegetation-related data.

[0073] Air-related data refers to data related to air quality. In some embodiments, air-related data may include smoke concentration and target substance concentration, etc.

[0074] Smoke concentration refers to the concentration of smoke within the target area.

[0075] Target substance concentration refers to the concentration of a target substance in the air within a target area. The target substance can be a substance that reflects the likelihood of a fire. In some embodiments, the target substance may include volatile organic compounds released by oily plants. In some embodiments, oily plants may include eucalyptus or pine trees, and volatile organic compounds may include isoprene or monoterpenes.

[0076] In some embodiments, the emergency monitoring and management platform can determine the smoke concentration and target substance concentration in the target area based on the smoke concentration and the concentration of various substances in the air collected by the gas sensors on the drone.

[0077] Vegetation-related data refers to data related to plants. In some embodiments, vegetation-related data may include the degree of leaf curling and the distribution of dead leaves within a target area. The degree of leaf curling may include the direction and extent of leaf curling.

[0078] The distribution of dead leaves is used to characterize the distribution of dead leaves within a target area. In some embodiments, the distribution of dead leaves can be represented by the amount of dead leaves in multiple sub-regions within the target area. Dead leaves can be dark-colored (e.g., dark brown).

[0079] In some embodiments, the emergency monitoring and management platform can analyze plant images captured by drones to obtain vegetation-related data. For example, based on plant images captured by drones, the platform can identify the amount of dead leaves in multiple sub-regions of the plant images using image recognition algorithms to obtain the distribution of dead leaves. As another example, based on plant images captured by drones, the platform can identify the direction and degree of leaf curling in the plant images using image recognition algorithms to obtain the degree of leaf curling. Plant images can be images of plants as the subject of the photographs.

[0080] In some embodiments, the emergency monitoring and management platform can determine whether a fire exists in a target area based on air-related data and vegetation-related data. For example, the platform can construct a target feature vector based on air-related data and vegetation-related data, and determine whether a fire exists in the target area by querying a first feature vector in a first vector library that meets the matching conditions, based on the label corresponding to the first feature vector. Matching conditions include the highest similarity between vectors. Vector similarity is negatively correlated with vector distance. Vector distance includes Euclidean distance, etc.

[0081] In some embodiments, the first vector library can be pre-configured based on historical data, including multiple first feature vectors and a label corresponding to each first feature vector. The first feature vector refers to a feature vector constructed based on historical airborne data and historical vegetation data. The label can be whether a fire occurred in the historical target area corresponding to the first feature vector.

[0082] In some embodiments, the emergency monitoring and management platform can calculate the probability of a fire occurring in the historical target area corresponding to the first feature vector during a third historical period. If the probability of a fire occurring is greater than a fire threshold, the platform determines that the label corresponding to the first feature vector indicates a fire exists. If the probability of a fire occurring is not greater than the fire threshold, the platform determines that the label corresponding to the first feature vector indicates no fire exists. The first historical period can be a time interval following the acquisition of historical air quality data and historical vegetation data for the historical target area. The third historical period is pre-set based on historical experience and includes multiple historical moments.

[0083] In some embodiments, the emergency monitoring and management platform can obtain the number of fire alarms triggered at multiple historical moments within a third historical period, and determine the probability of a fire as the ratio of the number of alarms triggered to the total number of alarms at each historical moment. Fire alarms can be deployed at multiple preset locations within a preset area and a target area.

[0084] In some embodiments, the emergency monitoring and management platform can also determine whether a fire exists in the target area through an anomaly identification model. See [link to relevant section] for more information. Figure 3 And its explanation.

[0085] In some embodiments of this specification, based on the air quality of the target area and the possible withering of vegetation, it is possible to comprehensively and accurately determine whether a fire has occurred in the target area.

[0086] In some embodiments, the emergency monitoring and management platform can also determine vegetation anomaly data 330 in the target area based on vegetation-related data 311 and target substance concentration 312 using anomaly identification model 320, and determine whether a fire exists in the target area based on vegetation anomaly data 330 and smoke concentration 340. For explanations of vegetation-related data, target substance concentration, and smoke concentration, please refer to... Figure 2 And its related descriptions.

[0087] Vegetation anomaly data can reflect abnormal vegetation conditions within a target area over a predetermined future time period. In some embodiments, vegetation anomaly data may include the location and severity of abnormal vegetation within the target area. For an explanation of the predetermined future time period, please refer to [link to documentation]. Figure 2 And its contents. Abnormal vegetation can be plants whose leaves curl and wither due to high temperatures.

[0088] In some embodiments, the anomaly detection model is a machine learning model, such as any one or a combination of a deep neural network (DNN) model, a recurrent neural network (RNN) model, or other custom model structures.

[0089] In some embodiments, the emergency monitoring and management platform can obtain an anomaly identification model by training multiple second training samples with second labels using methods such as gradient descent.

[0090] In some embodiments, the second training sample includes sample vegetation-related data and sample target substance concentration. The second label includes the actual vegetation anomaly data corresponding to the second training sample.

[0091] In some embodiments, the second training sample and the second label can be determined based on historical data. For example, the emergency monitoring and management platform can obtain historical vegetation-related data and historical target substance concentrations from historical data as the second training sample, and obtain vegetation anomaly data within a fourth historical time period corresponding to the second training sample as the second label. The fourth historical time period can be a period of time after obtaining the historical vegetation-related data in the second training sample. The fourth historical time period has the same duration as a preset future time period.

[0092] In some embodiments, for multiple sub-regions of the historical target area corresponding to the second training sample, the emergency monitoring and management platform can, based on historical vegetation-related data and historical target substance concentrations within the second historical period, perform a weighted summation of the historical amount of dead leaves, historical leaf curling degree, and historical target substance concentration for each sub-region. Sub-regions with a weighted summation value greater than an anomaly threshold are identified as locations of abnormal vegetation in the second label, and the weighted summation value is used as the degree of anomaly in the second label. The weights of the historical amount of dead leaves, historical leaf curling degree, and historical target substance concentration can be determined based on empirical presets. The anomaly threshold can also be determined based on empirical presets.

[0093] In some embodiments, the emergency monitoring and management platform can normalize the historical amount of dead leaves, the historical degree of leaf curling, and the historical concentration of target substances before weighted summation.

[0094] In some embodiments, the training method of the anomaly detection model is similar to that of the distribution prediction model, and will not be described in detail here.

[0095] In some embodiments, the emergency monitoring and management platform can determine whether a fire exists in a target area based on vegetation anomaly data and smoke concentration. For example, the platform can calculate the average density and severity of abnormal vegetation in the vegetation anomaly data, query the tags corresponding to the abnormal vegetation density, average severity of abnormality, and smoke concentration in a second preset table, and determine whether a fire exists in the target area based on the tags. Abnormal vegetation density can be the ratio of the quantity of abnormal vegetation to the area of ​​the target region.

[0096] In some embodiments, the second preset table can be pre-set based on historical data, including multiple sets of abnormal location densities, average abnormality levels, and labels corresponding to the smoke concentration and the multiple sets of data. The labels may indicate whether a fire exists in the target area. The second preset table can be determined based on empirical presets. For an explanation of smoke concentration, see [link to documentation]. Figure 2 And its contents.

[0097] In some embodiments, for areas with fires, the emergency monitoring and management platform can also determine the placement location of fire extinguishing materials and generate placement instructions based on vegetation-related data and target substance concentrations. These instructions are then sent to the emergency monitoring target platform via the emergency monitoring sensor network platform to control drones to move to the placement location and place fire extinguishing materials for fire suppression. See also [link to relevant content] for more details. Figure 4 And its related descriptions.

[0098] In some embodiments of this specification, by using an anomaly identification model to predict vegetation anomaly data within a preset future time period, it is possible to determine whether a fire has already occurred in the current target area by predicting the anomaly situation of vegetation in the future, thereby reducing the error of subjective judgment and improving the efficiency and reliability of fire determination.

[0099] Figure 4 This is an exemplary flowchart illustrating the generation of delivery instructions according to some embodiments of this specification. For example... Figure 4 As shown, the process 400 for generating delivery instructions includes the following steps.

[0100] Step 410: For target areas where fires are present, determine the location and distribution of abnormal vegetation based on vegetation-related data and target substance concentrations. See [link to documentation] for an explanation of vegetation-related data and target substance concentrations. Figure 3 And its contents.

[0101] The current location distribution of anomalous vegetation reflects the distribution of anomalous vegetation at the current moment. For an explanation of anomalous vegetation, please refer to [link to documentation / reference]. Figure 3 And its contents.

[0102] In some embodiments, the emergency monitoring and management platform can perform a weighted summation of current vegetation-related data and target substance concentration within the target area. Multiple locations where the weighted summation value exceeds an anomaly threshold are identified as locations of current anomalous vegetation. Based on these locations, the distribution of current anomalous vegetation locations is obtained. For an explanation of the weighted summation process and the anomaly threshold, please refer to [link to relevant documentation]. Figure 3 And its contents.

[0103] Step 420: Determine the placement location of fire extinguishing materials based on location distribution and vegetation anomaly data.

[0104] Fire extinguishing materials refer to substances used for fire extinguishing, such as fire extinguishing agents.

[0105] The delivery location refers to the location where the drone delivers fire extinguishing materials.

[0106] In some embodiments, the emergency monitoring and management platform can determine the placement location of fire extinguishing materials based on location distribution and vegetation anomaly data. For example, the emergency monitoring and management platform can record both the current location of abnormal vegetation in the location distribution and the location of abnormal vegetation in the vegetation anomaly data as placement locations.

[0107] In some embodiments, the emergency monitoring and management platform can send the delivery location to a drone, which then delivers the fire extinguishing materials to the designated location. There can be multiple delivery locations.

[0108] Step 430: Based on the deployment location, generate a deployment instruction and send the deployment instruction to the emergency monitoring target platform through the emergency monitoring sensor network platform to control the drone to move to the deployment location and deploy fire extinguishing materials.

[0109] A deployment instruction is a directive to guide the deployment of fire extinguishing materials. In some embodiments, the deployment instruction may include the deployment location of the fire extinguishing materials. The emergency monitoring and management platform can generate a deployment instruction based on the deployment location and send the deployment instruction to the emergency monitoring target platform through the emergency monitoring sensor network platform to control the drone to move to the deployment location and deploy the fire extinguishing materials.

[0110] In some embodiments of this specification, by using the current location of abnormal vegetation and the predicted location of abnormal vegetation as the placement location for fire extinguishing materials, a more comprehensive placement location can be determined, the fire can be extinguished in a timely manner, and some areas can be avoided from being missed.

[0111] In some embodiments, the delivery instructions may also include the dosage of the fire extinguishing material and the hovering altitude of the drone.

[0112] In some embodiments, the emergency monitoring and management platform can determine the severity of the fire in the target area based on vegetation anomaly data, and determine the application dosage and hovering altitude based on the fire severity. The platform generates a deployment command based on the dosage and hovering altitude, and sends the command to the monitored platform via an emergency monitoring sensor network platform. This command controls the drone to hover at the deployment location at the specified altitude and deliver the fire extinguishing material at the appropriate dosage.

[0113] Fire severity can reflect the severity of a fire occurring in a target area within a preset future time period. In some embodiments, fire severity may include the fire severity of multiple sub-areas within the target area. For an explanation of the preset future time period, see [link to documentation]. Figure 2 And its related descriptions.

[0114] In some embodiments, the emergency monitoring and management platform can determine the fire severity of each sub-region within a target area based on vegetation anomaly data. For example, for a single sub-region, the emergency monitoring and management platform calculates the average anomaly level corresponding to the location of one or more abnormal vegetation within the sub-region, and uses the average anomaly level as the fire severity of that sub-region.

[0115] Dosage refers to the amount of extinguishing material used.

[0116] Hovering altitude refers to the hovering height of the drone relative to the ground of the target area. In some embodiments, a sub-area corresponds to a set of delivery doses and hovering altitudes.

[0117] In some embodiments, the emergency monitoring and management platform can determine the dosage and hovering height based on the severity of the fire. For example, for each sub-area, the emergency monitoring and management platform can construct a matching vector based on the severity of the fire in the sub-area, query a second reference vector in a second vector library that meets the matching conditions with the matching vector, and determine the label of the second reference vector as the dosage and hovering height.

[0118] In some embodiments, the second vector library can be pre-configured based on historical data and includes multiple second reference vectors and a label corresponding to each second reference vector. For example, for each sub-region, the emergency monitoring and management platform can construct multiple second reference vectors based on multiple historical fire severity levels in the historical data for that sub-region, and use the historical delivery dose and historical hovering height corresponding to each historical fire severity level as the label of the corresponding second reference vector.

[0119] In some embodiments, the emergency monitoring and management platform can filter the second reference vectors. The filtering process includes: obtaining the fire suppression time and the damage level suffered by the drone when dropping fire extinguishing materials corresponding to the second reference vector; normalizing the fire suppression time and damage level and then performing a weighted summation; retaining the second reference vectors whose weighted summation value is greater than a preset effect threshold; and removing the second reference vectors whose weighted summation value is not greater than the preset effect threshold. The fire suppression time can be the duration it takes for the fire to disappear, determined based on historical data. The damage can be represented by the area of ​​the drone charred by the fire, determined based on historical data. The weights of the fire suppression time and the damage level can be determined based on empirical presets.

[0120] In some embodiments, the emergency monitoring and management platform can send a delivery instruction to the emergency monitoring target platform through the emergency monitoring sensor network platform to control the drone to hover at the delivery location at a hovering altitude and deliver fire extinguishing materials in the specified dosage.

[0121] In some embodiments of this specification, the severity of the fire in the target area is estimated based on abnormal vegetation data, and the delivery dose and hovering altitude are determined. This allows for the determination of more effective delivery doses and hovering altitudes for fire suppression based on the potential development of the fire, ensuring fire suppression effectiveness and reducing the lifespan of the drone.

[0122] This specification also provides a computer-readable storage medium that stores computer instructions. When a computer reads the computer instructions from the storage medium, the computer executes the urban suburban fire emergency prevention method described in the above embodiments.

[0123] The basic concepts have been described above. Obviously, for those skilled in the art, the detailed disclosure above is merely illustrative and does not constitute a limitation of this specification. Although not explicitly stated herein, those skilled in the art may make various modifications, improvements, and corrections to this specification. Such modifications, improvements, and corrections are suggested in this specification and therefore remain within the spirit and scope of the exemplary embodiments described herein.

[0124] It should be understood that the embodiments described in this specification are merely illustrative of the principles of the embodiments described herein. Other variations may also fall within the scope of this specification. Therefore, alternative configurations of the embodiments described herein are intended to be consistent with the teachings of this specification, rather than as examples or limitations. Accordingly, the embodiments described herein are not limited to those explicitly introduced and described herein.

Claims

1. A large-scale Internet of Things (IoT) model system for emergency fire prevention in suburban areas, characterized in that, The system includes an emergency monitoring user platform, an emergency monitoring service platform, an emergency monitoring management platform, an emergency monitoring sensor network platform, and an emergency monitoring object platform. The emergency monitoring management platform is configured to execute at preset intervals. Based on the temperature distribution, population distribution, area type, and historical incident data of multiple preset areas, multiple fire risk coefficients of the multiple preset areas are determined. The target area is determined based on the multiple fire risk coefficients and multiple risk thresholds. Based on the fire risk coefficient of the target area, a monitoring instruction including sampling point and sampling amount is generated, and the monitoring instruction is sent to the emergency monitoring object platform through the emergency monitoring sensor network platform to control the drone of the emergency monitoring object platform to move to the sampling point and acquire monitoring data corresponding to the sampling amount at the sampling point; The emergency monitoring and management platform is also configured as follows: Based on the monitoring data, air-related data and vegetation-related data for the target area are determined, wherein the air-related data includes smoke concentration and target substance concentration; Based on the air-related data and the vegetation-related data, it is determined whether there is a fire in the target area; The emergency monitoring and management platform is also configured as follows: Based on the vegetation-related data and the target substance concentration, the vegetation anomaly data of the target area is determined by an anomaly identification model, wherein the anomaly identification model is a machine learning model. Based on the abnormal vegetation data and the smoke concentration, it is determined whether there is a fire in the target area; The emergency monitoring and management platform is also configured as follows: For target areas where fires are present, the location and distribution of the current abnormal vegetation are determined based on the vegetation-related data and the concentration of the target substance. Based on the aforementioned location distribution and vegetation anomaly data, the placement locations for fire extinguishing materials are determined; and, Based on the deployment location, a deployment instruction is generated and sent to the emergency monitoring object platform through the emergency monitoring sensor network platform to control the drone to move to the deployment location and deploy the fire extinguishing materials.

2. The system as described in claim 1, characterized in that, The delivery command includes the dosage of the fire extinguishing material and the hovering altitude of the drone. The emergency monitoring and management platform is also configured to: Based on the vegetation anomaly data, the severity of the fire in the target area is determined; The dosage and hovering height are determined based on the severity of the fire. as well as, Based on the dosage and the hovering height, a delivery command is generated and sent to the emergency monitoring target platform through the emergency monitoring sensor network platform to control the drone to hover at the delivery location at the specified height and deliver the fire extinguishing material at the specified dosage.

3. A method for emergency fire prevention in suburban areas, characterized in that, The emergency monitoring and management platform within the urban suburban fire emergency prevention IoT big data model system executes at preset intervals, including: Based on the temperature distribution, population distribution, area type, and historical incident data of multiple preset areas, multiple fire risk coefficients of the multiple preset areas are determined. The target area is determined based on the multiple fire risk coefficients and multiple risk thresholds. Based on the fire risk coefficient of the target area, a monitoring instruction including sampling point and sampling amount is generated, and the monitoring instruction is sent to the emergency monitoring object platform through the emergency monitoring sensor network platform to control the drone of the emergency monitoring object platform to move to the sampling point and acquire monitoring data corresponding to the sampling amount at the sampling point; The method further includes: Based on the monitoring data, air-related data and vegetation-related data for the target area are determined, wherein the air-related data includes smoke concentration and target substance concentration; Based on the air-related data and the vegetation-related data, it is determined whether there is a fire in the target area; The method further includes: Based on the vegetation-related data and the target substance concentration, vegetation anomalies in the target area are determined by an anomaly identification model, which is a machine learning model. Based on the abnormal vegetation data and the smoke concentration, it is determined whether there is a fire in the target area; The method further includes: For target areas where fires are present, the location and distribution of the current abnormal vegetation are determined based on the vegetation-related data and the concentration of the target substance. Based on the aforementioned location distribution and vegetation anomaly data, the placement locations for fire extinguishing materials are determined; and, Based on the deployment location, a deployment instruction is generated and sent to the emergency monitoring object platform through the emergency monitoring sensor network platform to control the drone to move to the deployment location and deploy the fire extinguishing materials.

4. A computer-readable storage medium, characterized in that, The storage medium stores computer instructions. When the computer reads the computer instructions from the storage medium, the computer executes the method as described in claim 3.

Images