Vehicle emergency rescue dispatching method, device and equipment and computer program product

Abstract

This invention discloses a vehicle emergency rescue dispatch method, device, equipment, and computer program product, including: acquiring vehicle status data, occupant status data, and environmental scene data of a target vehicle, inputting them into a lightweight accident detection model deployed on the vehicle to determine whether an accident has occurred; when an accident occurs, inputting the data into an accident severity assessment model deployed in the cloud to obtain the current accident type and corresponding accident severity of the target vehicle; constructing a rescue resource dispatch optimization model based on the accident type and severity, aiming at minimizing rescue time, maximizing resource adaptability, and optimizing global resource utilization; solving the rescue resource dispatch optimization model using a genetic algorithm to obtain the optimal rescue resource dispatch scheme, and performing emergency rescue on the target vehicle according to the optimal rescue resource dispatch scheme. This invention improves the efficiency and reliability of vehicle emergency rescue and can be applied to the field of vehicle rescue technology.

Description

Technical Field

[0001] This invention relates to the field of vehicle rescue technology, and in particular to a vehicle emergency rescue dispatching method, device, equipment, and computer program product. Background Technology

[0002] Current technologies related to intelligent vehicle emergency rescue mainly revolve around the basic logic of "collision detection - automatic alarm". The core technical solutions can be divided into two categories: one is the vehicle-side passive triggering technology, which collects collision intensity data through vehicle collision sensors and airbag controllers, combines it with vehicle location information to trigger an emergency call, and provides guidance to rescue personnel on the vehicle unlocking path, thus achieving basic rescue connection after an accident; the other is the call-side matching optimization technology, which optimizes the matching rules for emergency call recipients by analyzing vehicle fault codes and collision parameters, thereby improving call response efficiency.

[0003] In addition, traditional emergency dispatch systems often adopt the principle of "allocation based on proximity." For example, the dispatch algorithm commonly used in urban emergency command platforms calculates the straight-line distance between rescue resources and accident sites and prioritizes dispatching the nearest rescue vehicles, without considering the actual adaptability of rescue resources and dynamic changes in the scenario.

[0004] In summary, while existing technologies have achieved basic emergency rescue functions, they have the following shortcomings in complex accident scenarios: 1) Insufficient accuracy in assessing the severity of accidents: Existing technologies rely on a single collision intensity parameter (such as acceleration value) to determine the accident level, which cannot cover complex scenarios such as "slight vehicle deformation but battery thermal runaway", "low collision intensity but occupant unconsciousness", and "differential damage in multi-vehicle chain collisions". This can easily lead to resource mismatch problems such as "over-adjusting minor injuries" or "missing to adjust serious injuries", resulting in reduced rescue efficiency.

[0005] 2) Rigid logic of rescue resource dispatch: The static dispatch mode of "proximity principle" is adopted without comprehensively considering dynamic factors such as the adaptability of rescue vehicle equipment (such as demolition equipment, battery fire extinguishing equipment), real-time traffic conditions (such as travel time on congested road sections), and the priority of multiple accidents. For example, rescue vehicles without the ability to extinguish fires in new energy vehicles are dispatched to the scene of battery fire accidents, delaying the golden rescue time.

[0006] 3) Weak data collaboration capabilities: There are "data silos" between vehicle sensor data (such as vehicle status and occupant information), roadside unit data (such as accident scene video and road conditions) and emergency platform data (such as rescue resource inventory and medical personnel status), which cannot achieve real-time fusion analysis, resulting in delayed dispatch decisions and an inability to adapt to the complex needs of intelligent vehicles after an accident.

[0007] In summary, existing technologies are insufficient to meet the needs of "accurate assessment, on-demand dispatch, and efficient collaboration" in vehicle emergency rescue, thus affecting the efficiency and reliability of vehicle emergency rescue. These issues urgently need to be addressed. Summary of the Invention

[0008] The purpose of this invention is to at least partially solve one of the technical problems existing in the prior art.

[0009] Therefore, one objective of this invention is to provide a vehicle emergency rescue dispatch method that improves the efficiency and reliability of vehicle emergency rescue.

[0010] Another objective of this invention is to provide a vehicle emergency rescue dispatch device.

[0011] To achieve the above-mentioned technical objectives, the technical solutions adopted in the embodiments of the present invention include: On one hand, embodiments of the present invention provide a vehicle emergency rescue dispatch method, including the following steps: The system acquires vehicle status data, occupant status data, and environmental scene data of the target vehicle. It then inputs these data into a lightweight accident detection model deployed on the vehicle and determines whether the target vehicle has been involved in an accident based on the accident detection results. When the target vehicle is involved in an accident, the vehicle status data, the occupant status data, and the environmental scene data are uploaded to the cloud and input into the accident severity assessment model deployed in the cloud to obtain the current accident type and corresponding accident severity of the target vehicle. Based on the accident type and the accident severity, a rescue resource scheduling optimization model is constructed with the objectives of minimizing rescue time, maximizing resource adaptability, and optimizing global resource utilization. The optimal rescue resource scheduling scheme is obtained by solving the rescue resource scheduling optimization model using a genetic algorithm, and emergency rescue is carried out on the target vehicle according to the optimal rescue resource scheduling scheme.

[0012] Furthermore, in one embodiment of the present invention, the lightweight accident detection model is obtained through the following steps: The first vehicle state sample, the first occupant state sample, and the first environmental scene sample of historical vehicles are obtained, and the corresponding accident occurrence labels are determined through manual annotation. The first vehicle state sample, the first occupant state sample, and the first environmental scene sample are input into a pre-constructed first multi-branch convolutional neural network to obtain the predicted accident detection result. A first loss value is determined based on the predicted accident detection results and the accident occurrence label; The parameters of the first multi-branch convolutional neural network are updated using the backpropagation algorithm based on the first loss value to obtain the trained accident detection teacher model. Knowledge distillation is performed on the accident detection teacher model to obtain the lightweight accident detection model.

[0013] Furthermore, in one embodiment of the present invention, the accident severity assessment model is obtained through the following steps: Acquire second vehicle state samples, second occupant state samples, and second environmental scene samples of historical vehicles at the time of accidents, and determine the corresponding first-level labels for accident type and second-level labels for severity through manual annotation; The second vehicle state sample, the second occupant state sample, and the second environmental scene sample are input into a pre-built second multi-branch convolutional neural network to obtain the predicted accident type and the predicted severity. A second loss value is determined based on the predicted accident type, the predicted severity, the primary label of the accident type, and the secondary label of the severity. The parameters of the second multi-branch convolutional neural network are updated using the backpropagation algorithm based on the second loss value to obtain the trained accident severity assessment model.

[0014] Furthermore, in one embodiment of the present invention, the step of constructing a rescue resource scheduling optimization model based on the accident type and the accident severity, with the objectives of minimizing rescue time, maximizing resource adaptability, and optimizing global resource utilization, specifically includes: Obtain the current vehicle location of the target vehicle, and obtain the resource type and current location of multiple alternative rescue resources; The rescue time for each of the candidate rescue resources is determined based on the current vehicle location and the current resource location, and a first objective function that minimizes the latest arrival time of the resources is constructed based on the rescue time. Based on the accident type, the accident severity, and the resource type, determine the resource suitability between each of the alternative rescue resources and the target vehicle, and construct a second objective function that maximizes the average resource suitability based on the resource suitability. The resource occupation cost of the corresponding alternative rescue resource is determined according to the resource type, and a third objective function that minimizes the total resource occupation cost is constructed based on the resource occupation cost. Based on the resource type, determine the resource capacity value of the corresponding alternative rescue resource; based on the accident type and the accident severity, determine the total resource capacity value of the target vehicle required for the accident; and then construct the constraint condition that the total resource capacity meets the accident requirements based on the resource capacity value and the total resource capacity value required for the accident. The rescue resource scheduling optimization model is obtained based on the first objective function, the second objective function, the third objective function, and the constraints.

[0015] Furthermore, in one embodiment of the present invention, the first objective function is:

[0016] in, Indicates the latest arrival time of the resource. Indicates alternative rescue resources The rescue time For alternative rescue resources The scheduling status, Indicates alternative rescue resources Scheduled Indicates alternative rescue resources Not scheduled; The second objective function is:

[0017] in, Indicates the average resource fit. Indicates alternative rescue resources Resource compatibility with the target vehicle; The third objective function is:

[0018] in, Indicates the total cost of resource usage. Indicates alternative rescue resources Resource occupancy costs; The constraints are as follows:

[0019] in, This represents the total resource capacity required for the target vehicle in case of an accident. Indicates alternative rescue resources Resource capability value; The objective function of the rescue resource scheduling optimization model is:

[0020] in, This indicates the priority score of the rescue resource allocation plan. , as well as This represents the preset weight parameters.

[0021] Furthermore, in one embodiment of the present invention, the step of solving the rescue resource scheduling optimization model using a genetic algorithm to obtain the optimal rescue resource scheduling scheme specifically includes: Multiple feasible solutions are randomly generated based on the constraints. The scheduling status of each of the alternative rescue resources in the feasible solutions is binary encoded to obtain the initial individual corresponding to the feasible solution. The population is initialized based on the initial individuals, and the fitness function is determined based on the rescue resource scheduling optimization model. The fitness value of each individual in the population is determined according to the fitness function. Based on the fitness value, individuals are selected using roulette wheel selection or tournament selection to obtain multiple parent individuals; Crossover and mutation operations are performed on parent individuals to obtain multiple offspring individuals, and the population is updated based on the offspring individuals; When the preset population iteration termination condition is met, the optimal rescue resource scheduling scheme is determined based on the best individual in the current population.

[0022] Furthermore, in one embodiment of the present invention, the step of providing emergency rescue to the target vehicle according to the optimal rescue resource scheduling scheme specifically includes: Multiple target rescue resources are determined based on the optimal rescue resource scheduling scheme; Dynamic path planning is performed based on the current resource location of each target rescue resource and the current vehicle location of each target vehicle to obtain the optimal path for each target rescue resource. Emergency rescue of the target vehicle is carried out by scheduling the target rescue resources according to the optimal path.

[0023] On the other hand, embodiments of the present invention provide a vehicle emergency rescue dispatch device, comprising: The vehicle-side detection module is used to acquire vehicle status data, occupant status data, and environmental scene data of the target vehicle, input the vehicle status data, occupant status data, and environmental scene data into a lightweight accident detection model deployed on the vehicle, and determine whether the target vehicle has been involved in an accident based on the accident detection results. The cloud-based assessment module is used to upload the vehicle status data, the occupant status data, and the environmental scene data to the cloud when the target vehicle is involved in an accident, and input them into the accident severity assessment model deployed in the cloud to obtain the current accident type and corresponding accident severity of the target vehicle. The optimization model building module is used to build a rescue resource scheduling optimization model based on the accident type and the accident severity, with the objectives of minimizing rescue time, maximizing resource adaptability, and optimizing global resource usage. The optimization solution module is used to solve the rescue resource scheduling optimization model through a genetic algorithm to obtain the optimal rescue resource scheduling scheme, and to carry out emergency rescue of the target vehicle according to the optimal rescue resource scheduling scheme.

[0024] On the other hand, embodiments of the present invention provide an electronic device, including: At least one processor; At least one memory for storing at least one program; When the at least one program is executed by the at least one processor, the at least one processor implements the above-described vehicle emergency rescue dispatch method.

[0025] On the other hand, embodiments of the present invention also provide a computer-readable storage medium storing a processor-executable computer program that, when executed by a processor, implements the above-described vehicle emergency rescue dispatch method.

[0026] On the other hand, embodiments of the present invention also provide a computer program product, including a computer program that, when executed by a processor, implements the above-described vehicle emergency rescue dispatch method.

[0027] The advantages and beneficial effects of the present 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: This invention acquires vehicle status data, occupant status data, and environmental scene data of a target vehicle. These data are then input into a lightweight accident detection model deployed on the vehicle. Based on the accident detection results, it is determined whether an accident has occurred. If an accident occurs, the vehicle status data, occupant status data, and environmental scene data are uploaded to the cloud and input into an accident severity assessment model deployed in the cloud. This yields the current accident type and corresponding accident severity for the target vehicle. Based on the accident type and severity, a rescue resource scheduling optimization model is constructed with the objectives of minimizing rescue time, maximizing resource adaptability, and optimizing global resource usage. A genetic algorithm is used to solve the rescue resource scheduling optimization model to obtain the optimal rescue resource scheduling scheme. Emergency rescue of the target vehicle is then carried out according to the optimal rescue resource scheduling scheme. This invention deploys a lightweight accident detection model on the vehicle side, determining whether an accident has occurred based on vehicle status data, occupant status data, and environmental scene data. Simultaneously, an accident severity assessment model is deployed in the cloud, identifying the accident type and severity based on data uploaded from the vehicle side. This allows the simple task of accident detection and the complex task of accident assessment to be completed on the vehicle side and the cloud side respectively, ensuring accurate accident detection on the vehicle side while precise identification of the accident type and severity via the cloud. This facilitates cloud-based scheduling of rescue resources, improving the efficiency and reliability of vehicle emergency rescue. Furthermore, the cloud constructs a rescue resource scheduling optimization model based on accident type and severity, aiming for the shortest rescue time, highest resource adaptability, and optimal global resource utilization. This model is solved using a genetic algorithm, comprehensively considering the rescue time, resource adaptability, and resource utilization cost of each alternative rescue resource, further enhancing the efficiency and reliability of vehicle emergency rescue. Attached Figure Description

[0028] To more clearly illustrate the technical solutions in the embodiments of the present invention, the drawings used in the embodiments of the present invention are described below. It should be understood that the drawings described below are only for the convenience of clearly describing some embodiments of the technical solutions of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0029] Figure 1 A flowchart illustrating the steps of a vehicle emergency rescue dispatching method provided in an embodiment of the present invention; Figure 2 This is a structural block diagram of a vehicle emergency rescue dispatching device provided in an embodiment of the present invention; Figure 3 This is a structural block diagram of an electronic device provided in an embodiment of the present invention. Detailed Implementation

[0030] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative of the invention and are not intended to limit the invention. In the following description, when referring to the accompanying drawings, unless otherwise indicated, the same numbers in different drawings represent the same or similar elements. The embodiments described in the following exemplary embodiments do not represent all embodiments consistent with the embodiments of this invention; they are merely examples of apparatuses and methods consistent with some aspects of the embodiments of this invention as detailed in the appended claims.

[0031] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention pertains. The terminology used herein is for the purpose of describing embodiments of the invention only and is not intended to limit the invention.

[0032] The vehicle emergency rescue dispatch method provided in this invention can be applied to a terminal, a server, or software running on a terminal or server. In some embodiments, the terminal can be a smartphone, tablet, laptop, desktop computer, smart speaker, smartwatch, or vehicle terminal, but is not limited thereto; the server can be configured as an independent physical server, a server cluster or distributed system composed of multiple physical servers, or a cloud server providing basic cloud computing services such as cloud services, cloud databases, cloud computing, cloud functions, cloud storage, network services, cloud communication, middleware services, domain name services, security services, CDN, and big data and artificial intelligence platforms. The server can also be a node server in a blockchain network; the software can be an application that implements the vehicle emergency rescue dispatch method, but is not limited to the above forms.

[0033] This invention can be used in a wide variety of general-purpose or special-purpose computer system environments or configurations. Examples include: personal computers, server computers, handheld or portable devices, tablet devices, multiprocessor systems, microprocessor-based systems, set-top boxes, programmable consumer electronics, network PCs, minicomputers, mainframe computers, and distributed computing environments including any of the above systems or devices. This invention can be described in the general context of computer-executable instructions, such as program modules, that are executed by a computer. Generally, program modules include routines, programs, objects, components, data structures, etc., that perform specific tasks or implement specific abstract data types. This invention can also be practiced in distributed computing environments where tasks are performed by remote processing devices connected via a communication network. In distributed computing environments, program modules can reside in local and remote computer storage media, including storage devices.

[0034] It should be noted that in various specific embodiments of the present invention, when processing data related to user identity or characteristics, such as user information, user behavior data, user historical data, and user location information, user permission or consent is obtained first. Furthermore, the collection, use, and processing of this data comply with relevant laws, regulations, and standards. In addition, when embodiments of the present invention require access to sensitive personal information of users, separate permission or consent from the user is obtained through pop-ups or redirection to a confirmation page. Only after obtaining the user's separate permission or consent is the necessary user-related data for the normal operation of the embodiments of the present invention acquired.

[0035] Reference Figure 1 This invention provides a vehicle emergency rescue dispatch method, which specifically includes the following steps: S101. Obtain vehicle status data, occupant status data and environmental scene data of the target vehicle, input the vehicle status data, occupant status data and environmental scene data into the lightweight accident detection model deployed on the vehicle, and determine whether the target vehicle has been involved in an accident based on the accident detection results. S102. When the target vehicle is involved in an accident, the vehicle status data, occupant status data, and environmental scene data are uploaded to the cloud and input into the accident severity assessment model deployed in the cloud to obtain the current accident type and corresponding accident severity of the target vehicle. S103. Construct a rescue resource scheduling optimization model based on the accident type and severity, with the objectives of minimizing rescue time, maximizing resource adaptability, and optimizing global resource utilization. S104. Solve the rescue resource scheduling optimization model using a genetic algorithm to obtain the optimal rescue resource scheduling scheme, and carry out emergency rescue on the target vehicle according to the optimal rescue resource scheduling scheme.

[0036] This invention deploys a lightweight accident detection model on the vehicle side, determining whether an accident has occurred based on vehicle status data, occupant status data, and environmental scene data. Simultaneously, an accident severity assessment model is deployed in the cloud, identifying the accident type and severity based on data uploaded from the vehicle side. This allows the simple task of accident detection and the complex task of accident assessment to be completed on the vehicle side and the cloud side respectively, ensuring accurate accident detection on the vehicle side while precise identification of the accident type and severity via the cloud. This facilitates cloud-based scheduling of rescue resources, improving the efficiency and reliability of vehicle emergency rescue. Furthermore, the cloud constructs a rescue resource scheduling optimization model based on accident type and severity, aiming for the shortest rescue time, highest resource adaptability, and optimal global resource utilization. This model is solved using a genetic algorithm, comprehensively considering the rescue time, resource adaptability, and resource utilization cost of each alternative rescue resource, further enhancing the efficiency and reliability of vehicle emergency rescue.

[0037] Specifically, the vehicle-mounted perception layer and edge computing layer are deployed on the vehicle side to acquire vehicle-mounted perception data and perform accident detection.

[0038] The vehicle-mounted perception layer integrates three types of perception modules to achieve comprehensive capture of accident scene data, specifically including: Vehicle Status Module: Composed of millimeter-wave radar, lidar, inertial measurement unit (IMU) and vehicle control unit (VCU), it collects parameters such as collision acceleration, vehicle body deformation, battery pack temperature / voltage, and chassis attitude, with a sampling frequency of up to 100Hz to ensure the integrity of data at the moment of an accident; Occupant Status Module: Collects data such as the number of occupants, sitting posture, respiratory rate, and heart rate through in-vehicle infrared cameras, seat pressure sensors, and vital sign monitoring belts. Uses image recognition technology to determine whether occupants are unconscious or injured, solving the problem of "false alarms of no injury". Environmental scenario module: Combining vehicle-mounted cameras and roadside unit (RSU) data, it collects environmental information such as accident scene videos, road type (highway / urban area), weather conditions, and distribution of surrounding obstacles, providing basic data for dispatch route planning.

[0039] The edge computing layer is deployed in the vehicle's intelligent gateway, and its core functions include: Data cleaning and fusion: Kalman filtering algorithm is used to remove noise from multi-source data. Vehicle, occupant and environmental data are standardized into a unified format through data fusion protocol to remove abnormal data (such as instantaneous sensor interference values). Preliminary accident detection: The lightweight accident detection model deployed on the vehicle is used to perform preliminary identification of occupant status images and key vehicle parameters to determine whether an accident has occurred. When an accident occurs, vehicle status data, occupant status data, and environmental scene data are uploaded to the cloud.

[0040] As an optional implementation, the lightweight accident detection model is obtained through the following steps: S201. Obtain the first vehicle status sample, the first occupant status sample, and the first environmental scene sample of the historical vehicles, and determine the corresponding accident occurrence label through manual annotation. S202. Input the first vehicle state sample, the first occupant state sample, and the first environmental scene sample into the pre-constructed first multi-branch convolutional neural network to obtain the predicted accident detection result. S203. Determine the first loss value based on the predicted accident detection results and the accident occurrence label; S204. Update the parameters of the first multi-branch convolutional neural network based on the first loss value using the backpropagation algorithm to obtain the trained accident detection teacher model. S205. Perform knowledge distillation on the accident detection teacher model to obtain a lightweight accident detection model.

[0041] Specifically, an accident detection teacher model is pre-trained based on a multi-branch convolutional neural network to detect whether a vehicle has been involved in an accident. Then, a lightweight accident detection model is obtained through knowledge distillation and deployed on the vehicle.

[0042] As an optional implementation, the accident severity assessment model is obtained through the following steps: S301. Obtain the second vehicle status sample, the second occupant status sample, and the second environmental scene sample of the historical vehicle at the time of the accident, and determine the corresponding first-level label of accident type and second-level label of severity through manual annotation. S302. Input the second vehicle state sample, the second occupant state sample, and the second environmental scene sample into the pre-constructed second multi-branch convolutional neural network to obtain the predicted accident type and the predicted severity. S303. Determine the second loss value based on the predicted accident type, predicted severity, primary accident type label, and secondary severity label; S304. Update the parameters of the second multi-branch convolutional neural network using the backpropagation algorithm based on the second loss value to obtain the trained accident severity assessment model.

[0043] Specifically, an accident severity assessment model, pre-trained based on a multi-branch convolutional neural network, can accurately identify and assess the severity of sudden accidents such as scratches, collisions, rollovers, and fires, facilitating subsequent cloud-based dispatch of rescue resources.

[0044] It should be noted that the severity of the accident in this application is a secondary label under the accident type. That is, different accident types correspond to different accident severity systems. For example, when the accident type is a scratch accident, the corresponding accident severity level can be the vehicle damage level (level 1-5). When the accident type is a collision accident, the corresponding accident severity level can be the occupant injury level (level 1-4), the secondary risk level (such as battery fire risk level 1-3), etc.

[0045] As an optional implementation, a rescue resource scheduling optimization model is constructed based on the accident type and severity, aiming to minimize rescue time, maximize resource adaptability, and optimize global resource utilization. This model specifically includes: S1031. Obtain the current vehicle location of the target vehicle, and obtain the resource type and current location of multiple alternative rescue resources; S1032. Determine the rescue time of each alternative rescue resource based on the current vehicle location and the current resource location, and construct a first objective function that minimizes the latest arrival time of the resource based on the rescue time. S1033. Determine the resource fit between each alternative rescue resource and the target vehicle based on the accident type, accident severity, and resource type, and construct a second objective function to maximize the average resource fit based on the resource fit. S1034. Determine the resource occupation cost of the corresponding alternative rescue resources according to the resource type, and construct a third objective function to minimize the total resource occupation cost based on the resource occupation cost; S1035. Determine the resource capacity value of the corresponding alternative rescue resources according to the resource type, determine the total resource capacity value of the target vehicle according to the accident type and the severity of the accident, and then construct the constraint condition that the total resource capacity meets the accident demand based on the resource capacity value and the total resource capacity value of the accident demand. S1036. Based on the first objective function, the second objective function, the third objective function, and the constraints, the rescue resource scheduling optimization model is obtained.

[0046] Specifically, with the objectives of "shortest rescue time, highest resource adaptability, and optimal global resource utilization," a rescue resource scheduling optimization model is constructed. Input parameters include a comprehensive accident assessment report (including accident type and severity), a rescue resource database (including rescue vehicle locations, equipment lists, medical personnel configurations, and medical resource allocations), and real-time traffic data (accessed via traffic department APIs and roadside unit data). The core objectives of the rescue resource scheduling optimization model in this embodiment are to minimize rescue time, maximize resource adaptability, and optimize global resource utilization. These three objectives involve certain trade-offs (for example, the fastest-arriving resource may not have high adaptability). Therefore, a multi-objective optimization algorithm is needed to find the Pareto optimal solution.

[0047] As an optional implementation, the first objective function is:

[0048] in, Indicates the latest arrival time of the resource. Indicates alternative rescue resources The rescue time For alternative rescue resources The scheduling status, Indicates alternative rescue resources Scheduled Indicates alternative rescue resources Not scheduled; The second objective function is:

[0049] in, Indicates the average resource fit. Indicates alternative rescue resources Resource compatibility with the target vehicle; The third objective function is:

[0050] in, Indicates the total cost of resource usage. Indicates alternative rescue resources Resource occupancy costs; The constraints are:

[0051] in, This represents the total resource capacity required for the target vehicle in case of an accident. Indicates alternative rescue resources Resource capability value; The objective function of the rescue resource scheduling optimization model is:

[0052] in, This indicates the priority score of the rescue resource allocation plan. , as well as This represents the preset weight parameters.

[0053] Specifically, the rescue resource scheduling optimization model of this invention quantifies the impact of the latest arrival time of resources, the average resource adaptability, and the total resource occupancy cost on the priority score of the rescue resource scheduling scheme based on preset weight parameters, and finally obtains the Pareto optimal solution.

[0054] As a further optional implementation, a genetic algorithm is used to solve the rescue resource scheduling optimization model to obtain the optimal rescue resource scheduling scheme, which specifically includes: S1041. Generate multiple feasible solutions randomly according to the constraints, and encode the scheduling status of each alternative rescue resource in the feasible solution in binary to obtain the initial individual corresponding to the feasible solution. S1042. Initialize the population based on the initial individuals and determine the fitness function based on the rescue resource scheduling optimization model; S1043. Determine the fitness value of each individual in the population based on the fitness function; S1044. Select individuals based on fitness values ​​using roulette wheel selection or tournament selection to obtain multiple parent individuals; S1045. Perform crossover and mutation operations on the parent individuals to obtain multiple offspring individuals, and update the population based on the offspring individuals; S1046. When the preset population iteration termination condition is reached, determine the optimal rescue resource scheduling scheme based on the best individual in the current population.

[0055] Specifically, multiple feasible solutions are randomly generated based on constraints and known alternative rescue resources. The scheduling status of each alternative rescue resource in the feasible solutions is then binary-coded, with 1 indicating scheduling and 0 indicating no scheduling. Each gene corresponds to a scheduling decision for a rescue resource, ultimately yielding initial individuals for each feasible solution. The population is initialized, and a rescue resource scheduling optimization model is obtained based on the aforementioned steps to determine the fitness function. The fitness value of each individual is calculated according to the fitness function, and individuals are selected based on roulette wheel selection or tournament selection to obtain multiple parent individuals. Crossover and mutation operations are performed on the parent individuals to obtain multiple offspring individuals, and the population is updated based on the offspring individuals, thus completing one population iteration. When the preset population iteration termination condition is met (the number of iterations reaches a threshold or there is no improvement in fitness for multiple consecutive generations), the gene encoding of the best individual in the current population is restored to the scheduling status of each alternative rescue resource, thus determining the optimal rescue resource scheduling scheme.

[0056] As an optional implementation method, emergency rescue of the target vehicle is carried out according to the optimal rescue resource dispatch plan, which specifically includes: S1047. Determine multiple target rescue resources based on the optimal rescue resource scheduling plan; S1048. Based on the current resource location of each target rescue resource and the current vehicle location of the target vehicle, perform dynamic path planning to obtain the optimal path for each target rescue resource. S1049. Dispatch rescue resources from each target according to the optimal path to carry out emergency rescue of the target vehicle.

[0057] Specifically, based on the optimal rescue resource scheduling plan, multiple target rescue resources that need to be scheduled are determined. Then, based on the current resource location of the target rescue resources and the current vehicle location of the target vehicles, dynamic path planning is performed based on real-time traffic conditions to obtain the optimal path for each target rescue resource. Finally, the optimal path and relevant information of the target vehicles are sent to the corresponding target rescue resources (rescue vehicles, rescue personnel terminals) to complete the scheduling of target rescue resources.

[0058] In some optional embodiments, when multiple accidents occur in a region, the cloud can prioritize risks based on the accident type and severity of each vehicle involved, thereby dynamically adjusting the allocation order of rescue resources and avoiding resource congestion.

[0059] The method steps of the embodiments of the present invention have been described above. It can be understood that the embodiments of the present invention deploy a lightweight accident detection model on the vehicle side, determining whether an accident has occurred based on vehicle status data, occupant status data, and environmental scene data. An accident severity assessment model is deployed in the cloud, identifying the accident type and severity based on data uploaded from the vehicle side. This allows the simple task of accident detection and the complex task of accident assessment to be completed on the vehicle side and the cloud side respectively, ensuring that the vehicle side can accurately detect the occurrence of an accident while the cloud side accurately identifies the accident type and severity, facilitating the scheduling of rescue resources in the cloud and improving the efficiency and reliability of vehicle emergency rescue. Furthermore, the cloud constructs a rescue resource scheduling optimization model based on the accident type and severity, aiming for the shortest rescue time, highest resource adaptability, and optimal global resource utilization. This model is solved using a genetic algorithm, comprehensively considering the rescue time, resource adaptability, and resource utilization cost of each alternative rescue resource, further improving the efficiency and reliability of vehicle emergency rescue.

[0060] Reference Figure 2 This invention provides a vehicle emergency rescue dispatch device, comprising: The vehicle-side detection module is used to acquire vehicle status data, occupant status data, and environmental scene data of the target vehicle. It inputs the vehicle status data, occupant status data, and environmental scene data into a lightweight accident detection model deployed on the vehicle side, and determines whether the target vehicle has been involved in an accident based on the accident detection results. The cloud-based assessment module is used to upload vehicle status data, occupant status data, and environmental scene data to the cloud when an accident occurs. These data are then input into an accident severity assessment model deployed in the cloud to obtain the current accident type and corresponding accident severity of the target vehicle. The optimization model building module is used to build an optimization model for rescue resource scheduling based on the accident type and severity, with the goals of minimizing rescue time, maximizing resource adaptability, and optimizing global resource usage. The optimization solution module is used to solve the rescue resource scheduling optimization model through a genetic algorithm to obtain the optimal rescue resource scheduling scheme, and to carry out emergency rescue of the target vehicle according to the optimal rescue resource scheduling scheme.

[0061] It is understood that the content of the above method embodiments is applicable to the present device embodiments. The specific functions implemented by the present device embodiments are the same as those of the above method embodiments, and the beneficial effects achieved are also the same as those achieved by the above method embodiments.

[0062] Reference Figure 3 This invention provides an electronic device, comprising: At least one processor; At least one memory for storing at least one program; When the above-mentioned at least one program is executed by the above-mentioned at least one processor, the above-mentioned at least one processor implements the above-mentioned vehicle emergency rescue dispatch method.

[0063] It is understood that the content of the above method embodiments is applicable to this device embodiment. The specific functions implemented by this device embodiment are the same as those of the above method embodiments, and the beneficial effects achieved are also the same as those achieved by the above method embodiments.

[0064] This invention also provides a computer-readable storage medium storing a processor-executable computer program that, when executed by a processor, implements the aforementioned vehicle emergency rescue dispatch method.

[0065] This invention provides a computer-readable storage medium that can execute a vehicle emergency rescue dispatch method provided in the method embodiment of this invention. It can execute any combination of the implementation steps of the method embodiment and has the corresponding functions and beneficial effects of the method.

[0066] This invention also provides a computer program product, including a computer program that, when executed by a processor, implements the above-described vehicle emergency rescue dispatch method.

[0067] It is understood that the content of the above method embodiments is applicable to the embodiments of this program product. The specific functions implemented by the embodiments of this program product are the same as those of the above method embodiments, and the beneficial effects achieved are also the same as those achieved by the above method embodiments.

[0068] Memory, as a non-transitory computer-readable storage medium, can be used to store non-transitory software programs and non-transitory computer-executable programs. Furthermore, memory may include high-speed random access memory, and may also include non-transitory memory, such as at least one disk storage device, flash memory device, or other non-transitory solid-state storage device. In some embodiments, memory may optionally include memory remotely located relative to the processor, and these remote memories can be connected to the processor via a network. Examples of such networks include, but are not limited to, the Internet, intranets, local area networks, mobile communication networks, and combinations thereof.

[0069] The embodiments described in this invention are for the purpose of more clearly illustrating the technical solutions of the embodiments of this invention, and do not constitute a limitation on the technical solutions provided by the embodiments of this invention. As those skilled in the art will know, with the evolution of technology and the emergence of new application scenarios, the technical solutions provided by the embodiments of this invention are also applicable to similar technical problems.

[0070] The terms "first," "second," "third," "fourth," etc. (if present) in the specification and accompanying drawings of this invention are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that such data can be interchanged where appropriate so that embodiments of the invention described herein can be implemented in orders other than those illustrated or described herein. Furthermore, the terms "comprising" and "having," and any variations thereof, are intended to cover a non-exclusive inclusion; for example, a process, method, system, product, or apparatus that comprises a series of steps or units is not necessarily limited to those steps or units explicitly listed, but may include other steps or units not explicitly listed or inherent to such processes, methods, products, or apparatus.

[0071] In some alternative embodiments, the functions / operations mentioned in the block diagrams may not occur in the order shown in the operation diagrams. For example, depending on the functions / operations involved, two consecutively shown blocks may actually be executed substantially simultaneously, or the aforementioned blocks may sometimes be executed in reverse order. Furthermore, the embodiments presented and described in the flowcharts of this invention are provided by way of example to provide a more comprehensive understanding of the technology. The disclosed methods are not limited to the operations and logic flows presented herein. Alternative embodiments are contemplated in which the order of various operations is changed and sub-operations described as part of a larger operation are executed independently.

[0072] Furthermore, although the invention has been described in the context of functional modules, it should be understood that, unless otherwise stated, one or more of the aforementioned functions and / or features may be integrated into a single physical device and / or software module, or one or more functions and / or features may be implemented in a separate physical device or software module. It is also understood that a detailed discussion of the actual implementation of each module is unnecessary for understanding the invention. Rather, given the properties, functions, and internal relationships of the various functional modules in the apparatus disclosed herein, the actual implementation of the module will be understood within the scope of conventional skill of an engineer. Therefore, those skilled in the art can implement the invention as set forth in the claims using ordinary techniques without excessive experimentation. It is also understood that the specific concepts disclosed are merely illustrative and not intended to limit the scope of the invention, which is determined by the full scope of the appended claims and their equivalents.

[0073] If the aforementioned functions are implemented as software functional units and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of this invention, or the part that contributes to the prior art, or a portion of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of this invention. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0074] The logic and / or steps represented in the flowchart or otherwise described herein, for example, can be considered as a sequenced list of executable instructions for implementing logical functions, and can be embodied in any computer-readable medium for use by, or in conjunction with, an instruction execution system, apparatus, or device (such as a computer-based system, a processor-including system, or other system that can fetch and execute instructions from, an instruction execution system, apparatus, or device). For the purposes of this specification, "computer-readable medium" can be any means that can contain, store, communicate, propagate, or transmit programs for use by, or in conjunction with, an instruction execution system, apparatus, or device.

[0075] More specific examples (a non-exhaustive list) of computer-readable media include: electrical connections (electronic devices) having one or more wires, portable computer disk drives (magnetic devices), random access memory (RAM), read-only memory (ROM), erasable and editable read-only memory (EPROM or flash memory), fiber optic devices, and portable optical disc read-only memory (CDROM). Furthermore, computer-readable media can even be paper or other suitable media on which the aforementioned program can be printed, because the aforementioned program can be obtained electronically, for example, by optically scanning the paper or other medium, followed by editing, interpreting, or otherwise processing as necessary, and then stored in computer memory.

[0076] It should be understood that various parts of the present invention can be implemented in hardware, software, firmware, or a combination thereof. In the above embodiments, multiple steps or methods can be implemented in software or firmware stored in memory and executed by a suitable instruction execution system. For example, if implemented in hardware, as in another embodiment, it can be implemented using any one or a combination of the following techniques known in the art: discrete logic circuits having logic gates for implementing logical functions on data signals, application-specific integrated circuits (ASICs) having suitable combinational logic gates, programmable gate arrays (PGAs), field-programmable gate arrays (FPGAs), etc.

[0077] In the foregoing description of this specification, references to terms such as "one embodiment," "another embodiment," or "some embodiments" indicate that a specific feature, structure, material, or characteristic described in connection with an embodiment or example is included in at least one embodiment or example of the present invention. In this specification, illustrative expressions of the above terms do not necessarily refer to the same embodiment or example. Furthermore, the specific features, structures, materials, or characteristics described may be combined in any suitable manner in one or more embodiments or examples.

[0078] Although embodiments of the invention have been shown and described, those skilled in the art will understand that various changes, modifications, substitutions and alterations can be made to these embodiments without departing from the principles and spirit of the invention, the scope of which is defined by the claims and their equivalents.

[0079] The above is a detailed description of the preferred embodiments of the present invention. However, the present invention is not limited to the above embodiments. Those skilled in the art can make various equivalent modifications or substitutions without departing from the spirit of the present invention. All such equivalent modifications or substitutions are included within the scope defined by the claims of the present invention.

Claims

1. A vehicle emergency rescue dispatch method, characterized in that, Includes the following steps: The system acquires vehicle status data, occupant status data, and environmental scene data of the target vehicle. It then inputs these data into a lightweight accident detection model deployed on the vehicle and determines whether the target vehicle has been involved in an accident based on the accident detection results. When the target vehicle is involved in an accident, the vehicle status data, the occupant status data, and the environmental scene data are uploaded to the cloud and input into the accident severity assessment model deployed in the cloud to obtain the current accident type and corresponding accident severity of the target vehicle. Based on the accident type and the accident severity, a rescue resource scheduling optimization model is constructed with the objectives of minimizing rescue time, maximizing resource adaptability, and optimizing global resource utilization. The optimal rescue resource scheduling scheme is obtained by solving the rescue resource scheduling optimization model using a genetic algorithm, and emergency rescue is carried out on the target vehicle according to the optimal rescue resource scheduling scheme.

2. The vehicle emergency rescue dispatch method according to claim 1, characterized in that, The lightweight accident detection model is obtained through the following steps: The first vehicle state sample, the first occupant state sample, and the first environmental scene sample of historical vehicles are obtained, and the corresponding accident occurrence labels are determined through manual annotation. The first vehicle state sample, the first occupant state sample, and the first environmental scene sample are input into a pre-constructed first multi-branch convolutional neural network to obtain the predicted accident detection result. A first loss value is determined based on the predicted accident detection results and the accident occurrence label; The parameters of the first multi-branch convolutional neural network are updated using the backpropagation algorithm based on the first loss value to obtain the trained accident detection teacher model. Knowledge distillation is performed on the accident detection teacher model to obtain the lightweight accident detection model.

3. The vehicle emergency rescue dispatch method according to claim 1, characterized in that, The accident severity assessment model is obtained through the following steps: Acquire second vehicle state samples, second occupant state samples, and second environmental scene samples of historical vehicles at the time of accidents, and determine the corresponding first-level labels for accident type and second-level labels for severity through manual annotation; The second vehicle state sample, the second occupant state sample, and the second environmental scene sample are input into a pre-built second multi-branch convolutional neural network to obtain the predicted accident type and the predicted severity. A second loss value is determined based on the predicted accident type, the predicted severity, the primary label of the accident type, and the secondary label of the severity. The parameters of the second multi-branch convolutional neural network are updated using the backpropagation algorithm based on the second loss value to obtain the trained accident severity assessment model.

4. The vehicle emergency rescue dispatch method according to claim 1, characterized in that, The rescue resource scheduling optimization model, constructed based on the accident type and severity, aims to minimize rescue time, maximize resource adaptability, and optimize global resource utilization. Specifically, it includes: Obtain the current vehicle location of the target vehicle, and obtain the resource type and current location of multiple alternative rescue resources; The rescue time for each of the candidate rescue resources is determined based on the current vehicle location and the current resource location, and a first objective function that minimizes the latest arrival time of the resources is constructed based on the rescue time. Based on the accident type, the accident severity, and the resource type, determine the resource suitability between each of the alternative rescue resources and the target vehicle, and construct a second objective function that maximizes the average resource suitability based on the resource suitability. The resource occupation cost of the corresponding alternative rescue resource is determined according to the resource type, and a third objective function that minimizes the total resource occupation cost is constructed based on the resource occupation cost. Based on the resource type, determine the resource capacity value of the corresponding alternative rescue resource; based on the accident type and the accident severity, determine the total resource capacity value of the target vehicle required for the accident; and then construct the constraint condition that the total resource capacity meets the accident requirements based on the resource capacity value and the total resource capacity value required for the accident. The rescue resource scheduling optimization model is obtained based on the first objective function, the second objective function, the third objective function, and the constraints.

5. A vehicle emergency rescue dispatching method according to claim 4, characterized in that, The first objective function is: in, Indicates the latest arrival time of the resource. Indicates alternative rescue resources The rescue time For alternative rescue resources The scheduling status, Indicates alternative rescue resources Scheduled Indicates alternative rescue resources Not scheduled; The second objective function is: in, Indicates the average resource fit. Indicates alternative rescue resources Resource compatibility with the target vehicle; The third objective function is: in, Indicates the total cost of resource usage. Indicates alternative rescue resources Resource occupancy costs; The constraints are as follows: in, This represents the total resource capacity required for the target vehicle in case of an accident. Indicates alternative rescue resources Resource capability value; The objective function of the rescue resource scheduling optimization model is: in, This indicates the priority score of the rescue resource allocation plan. , as well as This represents the preset weight parameters.

6. The vehicle emergency rescue dispatch method according to claim 4, characterized in that, The step of solving the rescue resource scheduling optimization model using a genetic algorithm to obtain the optimal rescue resource scheduling scheme specifically includes: Multiple feasible solutions are randomly generated based on the constraints. The scheduling status of each of the alternative rescue resources in the feasible solutions is binary encoded to obtain the initial individual corresponding to the feasible solution. The population is initialized based on the initial individuals, and the fitness function is determined based on the rescue resource scheduling optimization model. The fitness value of each individual in the population is determined according to the fitness function. Based on the fitness value, individuals are selected using roulette wheel selection or tournament selection to obtain multiple parent individuals; Crossover and mutation operations are performed on parent individuals to obtain multiple offspring individuals, and the population is updated based on the offspring individuals; When the preset population iteration termination condition is met, the optimal rescue resource scheduling scheme is determined based on the best individual in the current population.

7. A vehicle emergency rescue dispatching method according to any one of claims 1 to 6, characterized in that, The step of providing emergency rescue to the target vehicle according to the optimal rescue resource dispatch plan specifically includes: Multiple target rescue resources are determined based on the optimal rescue resource scheduling scheme; Dynamic path planning is performed based on the current resource location of each target rescue resource and the current vehicle location of each target vehicle to obtain the optimal path for each target rescue resource. Emergency rescue of the target vehicle is carried out by scheduling the target rescue resources according to the optimal path.

8. A vehicle emergency rescue dispatch device, characterized in that, include: The vehicle-side detection module is used to acquire vehicle status data, occupant status data, and environmental scene data of the target vehicle, input the vehicle status data, occupant status data, and environmental scene data into a lightweight accident detection model deployed on the vehicle, and determine whether the target vehicle has been involved in an accident based on the accident detection results. The cloud-based assessment module is used to upload the vehicle status data, the occupant status data, and the environmental scene data to the cloud when the target vehicle is involved in an accident, and input them into the accident severity assessment model deployed in the cloud to obtain the current accident type and corresponding accident severity of the target vehicle. The optimization model building module is used to build a rescue resource scheduling optimization model based on the accident type and the accident severity, with the objectives of minimizing rescue time, maximizing resource adaptability, and optimizing global resource usage. The optimization solution module is used to solve the rescue resource scheduling optimization model through a genetic algorithm to obtain the optimal rescue resource scheduling scheme, and to carry out emergency rescue of the target vehicle according to the optimal rescue resource scheduling scheme.

9. An electronic device, characterized in that, include: At least one processor; At least one memory for storing at least one program; When the at least one program is executed by the at least one processor, the at least one processor implements a vehicle emergency rescue dispatch method as described in any one of claims 1 to 7.

10. A computer program product, comprising a computer program, characterized in that, When the computer program is executed by the processor, it implements a vehicle emergency rescue dispatch method as described in any one of claims 1 to 7.

Images