Resource scheduling method, system and device based on Internet of Vehicles and storage medium

Abstract

The invention relates to a resource scheduling method, system and device based on the Internet of Vehicles, and a storage medium, and relates to the technical field of the Internet of Vehicles. The method comprises the steps that an Internet of Vehicles calculation task issued by a vehicle terminal is acquired, the Internet of Vehicles calculation task at least comprises an identification code, a sensor signal of a vehicle corresponding to the identification code is acquired, and communication connection with the corresponding vehicle is established based on the sensor signal; constructing a multi-stage mobile computing power network model based on the sensor signal, wherein the multi-stage mobile computing power network model comprises a cloud computing sub-model, an edge computing sub-model and an end computing sub-model; the method comprises the following steps: decomposing an Internet of Vehicles calculation task based on a multi-stage mobile computing power network model to obtain a plurality of calculation sub-tasks, and constructing a directed acyclic graph of the plurality of calculation sub-tasks; based on the directed acyclic graph, computing power resource scheduling calculation is conducted on the multiple calculation sub-tasks, and a sub-task execution priority sequence is obtained. The method has the advantages that the computing power resources are reasonably scheduled, and the computing efficiency of the car networking task is improved.

Description

Technical Field

[0001] The present application relates to the technical field of Internet of Vehicles, and in particular to a resource scheduling method, system, device and storage medium based on Internet of Vehicles. Background Art

[0002] The Internet of Vehicles is a new industry that places cars in the generalized Internet to achieve multi-dimensional interactions between cars, cars and roads, cars and people, cars and sensor devices, etc. It is based on the in-vehicle network, inter-vehicle network and in-vehicle mobile Internet, and in accordance with the agreed communication protocols and data interaction standards, it builds a large system network for wireless communication and information exchange between cars, people, roads and service platforms; with the continuous improvement of the degree of vehicle networking, the automatic driving of vehicles is gradually becoming popular, so the computing tasks of the Internet of Vehicles are becoming more and more complex. The main computing tasks in the Internet of Vehicles system are large-scale distributed dependent computing tasks. Existing Internet of Vehicles technologies are difficult to effectively raise computing power, so it is impossible to efficiently process such large-scale distributed dependent computing tasks. The low computing efficiency makes it difficult for vehicle equipment to respond quickly. Therefore, this application proposes a computing power resource scheduling platform, and based on the platform, adopts a resource scheduling method based on the Internet of Vehicles to reasonably schedule computing power resources and improve the computing efficiency of Internet of Vehicles tasks. Summary of the invention

[0003] In order to reasonably schedule computing resources and improve the computing efficiency of Internet of Vehicles tasks, the present application provides a resource scheduling method, system, device and storage medium based on the Internet of Vehicles.

[0004] The present application provides a resource scheduling method, system, device and storage medium based on the Internet of Vehicles, which adopts the following technical solutions: In a first aspect, the present application discloses a resource scheduling method based on the Internet of Vehicles, wherein the execution subject of the method is a computing resource scheduling platform, wherein the computing resource scheduling platform is communicatively connected to a cloud, wherein the cloud pre-stores identification codes for distinguishing different vehicle terminals, and the cloud pre-stores a correspondence between the identification codes and the sensor signals of all devices in the corresponding vehicle terminals; The method comprises: Acquire a vehicle networking computing task issued by a vehicle terminal, wherein the vehicle networking computing task at least includes an identification code for identifying different vehicle terminals, acquire a sensor signal of a vehicle corresponding to the identification code, and establish a communication connection with the corresponding vehicle based on the sensor signal; Constructing a multi-level mobile computing network model based on all the sensor signals, wherein the multi-level mobile computing network model includes but is not limited to a cloud computing sub-model, an edge computing sub-model and a terminal computing sub-model; Decomposing the Internet of Vehicles computing task based on a multi-level mobile computing power network model to obtain multiple computing subtasks, and constructing a directed acyclic graph of the multiple computing subtasks; Based on the directed acyclic graph, computing power resource scheduling calculation is performed on the multiple computing subtasks to obtain a subtask execution priority sequence.

[0005] Optionally, the multi-level mobile computing power network model is used to decompose the Internet of Vehicles computing task to obtain multiple computing sub-tasks, and also includes: the computing sub-tasks at least include a task generation source and a nearby device sensor signal, and a suitable computing sub-model is selected based on the task generation source and the nearby device sensor signal to construct a directed acyclic graph of the multiple computing sub-tasks.

[0006] Optionally, the multi-level mobile computing power network model also includes computing power attributes corresponding to the cloud computing sub-model, the edge computing sub-model and the end computing sub-model, and the computing power attributes include at least computing speed and computing energy consumption. The computing power resource scheduling calculation includes at least computing power attribute calculation, and the sub-task execution priority sequence is obtained based on the computing speed and computing energy consumption.

[0007] Optionally, the vehicle network computing task is decomposed based on a multi-level mobile computing power network model to obtain multiple computing sub-tasks, and a directed acyclic graph of the multiple computing sub-tasks is constructed, which also includes: acquiring the dynamic changes of the network topology of the vehicle terminal in real time, and controlling the dynamic exit or entry of the multi-level mobile computing power network model based on the dynamic changes of the network topology.

[0008] Optionally, the method of obtaining the Internet of Vehicles computing tasks issued by the vehicle terminal, wherein the Internet of Vehicles computing tasks at least include an identification code for identifying different vehicle terminals, obtaining a sensor signal of a vehicle corresponding to the identification code, and establishing a communication connection with the corresponding vehicle based on the sensor signal, also includes: storing the Internet of Vehicles computing tasks in a historical task database preset in the cloud, obtaining the Internet of Vehicles computing tasks with the corresponding identification code in the historical task database, and constructing a deep reinforcement learning network. The computing power resource scheduling calculation also includes at least deep reinforcement learning network calculation.

[0009] In a second aspect, the present application further provides a resource scheduling system based on the Internet of Vehicles, the resource scheduling system comprising: An information interaction and communication module is used to obtain a vehicle networking computing task issued by a vehicle terminal, wherein the vehicle networking computing task at least includes an identification code for identifying different vehicle terminals, obtain a sensor signal of a vehicle corresponding to the identification code, and establish a communication connection with the corresponding vehicle based on the sensor signal; A computing power network construction module: used to construct a multi-level mobile computing power network model based on all the sensor signals, wherein the multi-level mobile computing power network model includes but is not limited to a cloud computing sub-model, an edge computing sub-model and a terminal computing sub-model; A computing task decomposition module is used to decompose the Internet of Vehicles computing task based on a multi-level mobile computing network model to obtain multiple computing subtasks and construct a directed acyclic graph of the multiple computing subtasks; Computing power resource scheduling module: used to perform computing power resource scheduling calculations on the multiple computing subtasks based on the directed acyclic graph to obtain a subtask execution priority sequence.

[0010] In a third aspect, the present application also provides a computing resource scheduling platform, the computing resource scheduling platform comprising a processor and a memory, the memory storing at least one instruction, at least one program, a code set or an instruction set, the at least one instruction, the at least one program, the code set or the instruction set being loaded and executed by the processor to implement the above-mentioned resource scheduling method based on the Internet of Vehicles. In a fourth aspect, the present application also provides a computer-readable storage medium, in which at least one instruction, at least one program, a code set or an instruction set is stored, and the at least one instruction, the at least one program, the code set or the instruction set is loaded and executed by a processor to implement the above-mentioned resource scheduling method based on the Internet of Vehicles.

[0011] In summary, the present application includes at least one of the following beneficial technical effects: 1. This application can distribute the computing power, storage, algorithm and other resource information of service nodes (devices in various parts of the vehicle, cloud, etc.) through the network control plane to achieve the optimal configuration and use of computing resources, thereby improving the execution efficiency of vehicle network computing tasks; 2. This application uses each subtask as a vertex and the dependency relationship between each subtask as an edge to generate a directed acyclic graph, so that the Internet of Vehicles computing tasks can be executed in a distributed manner, computing resources can be scheduled rationally, and the task can be quickly calculated and executed. BRIEF DESCRIPTION OF THE DRAWINGS

[0012] Figure 1 It is a flow chart of a resource scheduling method based on the Internet of Vehicles in an embodiment of the present application.

[0013] Figure 2 This is a structural block diagram of an embodiment of the present application used to embody a resource scheduling system based on the Internet of Vehicles.

[0014] Explanation of the accompanying drawings: 201, information interaction and communication module; 202, computing power network construction module; 203, computing power task decomposition module; 204, computing power resource scheduling module. DETAILED DESCRIPTION

[0015] The following is combined with Figure 1-2 This application is described in further detail.

[0016] The embodiment of the present application discloses a resource scheduling method based on the Internet of Vehicles. The above method aims to rationally schedule computing resources and improve the computing efficiency of Internet of Vehicles tasks. The execution subject of the above method is a computing resource scheduling platform, which is connected to the Internet of Vehicles in the form of APP software. At the same time, the above platform is connected to the cloud in communication. The cloud pre-stores identification codes (which can be frame numbers, etc.) for distinguishing different vehicle terminals (vehicle computers can be selected). In addition, a database is preset in the cloud, which stores the corresponding relationship between the identification code and the sensor signal of the corresponding vehicle terminal corresponding to the vehicle equipment. The devices corresponding to the above sensor signals must have certain computing and task execution capabilities.

[0017] The specific processing flow of the resource scheduling method based on the Internet of Vehicles is as follows: Figure 1 , the following will be combined with specific implementation methods. Figure 1 The specific processing flow in is described in detail as follows: Step 101, obtaining the Internet of Vehicles computing task issued by the vehicle terminal, wherein the Internet of Vehicles computing task at least includes an identification code for identifying different vehicle terminals, obtaining a sensor signal of the vehicle corresponding to the identification code, and establishing a communication connection with the corresponding vehicle based on the sensor signal.

[0018] During the implementation process, when the vehicle is performing autonomous driving, the vehicle terminal will issue vehicle networking computing tasks to ensure the safety and stability of the vehicle during autonomous driving. The vehicle networking computing tasks must at least include identification codes and establish communication connections with the sensors of the corresponding vehicle equipment based on the identification codes to achieve data interaction so that each device of the vehicle can respond in a timely manner. The settings of the vehicle equipment sensors include but are not limited to the engine, gearbox, wheel steering, etc., so that the calculation results can be executed in all aspects to ensure the safety of the vehicle's autonomous driving.

[0019] Optionally, the Internet of Vehicles computing task may also include parameter information of the vehicle (such as length, width, height, wheelbase, etc.) to more accurately control the operation of various vehicle devices based on the parameter information and calculation results.

[0020] Optionally, step 101 further includes the following steps: Step 1011, store the Internet of Vehicles computing tasks in a historical task database preset in the cloud, obtain the Internet of Vehicles computing tasks with corresponding identification codes in the historical task database, and build a deep reinforcement learning network.

[0021] During the specific implementation process, each completed Internet of Vehicles computing task will be stored in the historical task database, so that when the subsequent vehicle performs autonomous driving, a deep reinforcement learning network can be built, which can more quickly improve the calculation of various information during the vehicle's autonomous driving process, further improve the scheduling speed of computing resources, thereby rationalizing the scheduling of computing resources and saving running time.

[0022] Step 102: construct a multi-level mobile computing network model based on all sensor signals, wherein the multi-level mobile computing network model includes but is not limited to a cloud computing sub-model, an edge computing sub-model, and an end computing sub-model.

[0023] In the specific implementation process, three computing sub-models are used to form a multi-level mobile computing network model. Among them, the cloud computing sub-model is used to obtain and calculate shared information in the Internet of Vehicles, such as driving routes, road conditions, apocalypse conditions, etc. Driving tasks that do not require immediate response from the vehicle can obtain relevant information by accessing the network to calculate large-scale tasks during the vehicle's autonomous driving process; the edge computing sub-model is used to place data processing and computing resources on edge devices and sensors close to the source of data generation to provide faster and real-time computing and data analysis capabilities. For example, during vehicle driving, the ACC adaptive cruise control system is used to obtain the conditions of the road ahead (such as curves, etc.) so that the vehicle's steering system can respond when it arrives at the corresponding location; the end computing sub-model is used to perform data processing and calculations on terminal devices without relying on the cloud or central server. For example, the vehicle's on-board computer is used for calculations to reduce dependence and eliminate the need for network data transmission, thereby improving autonomous computing and decision-making capabilities and overall real-time response capabilities.

[0024] Optionally, step 102 further includes the following steps: Step 1021, the multi-level mobile computing power network model also includes computing power attributes corresponding to the cloud computing sub-model, the edge computing sub-model and the end computing sub-model, and the computing power attributes include at least computing speed and computing energy consumption.

[0025] During the specific implementation process, it is necessary to consider the computing speed and computing energy consumption of each computing sub-model based on the urgency of the computing task, the computing type, and the computing amount, so as to select the appropriate computing sub-model and achieve a more reasonable computing resource scheduling.

[0026] Step 103, based on the multi-level mobile computing power network model, the Internet of Vehicles computing task is decomposed to obtain multiple computing subtasks, and a directed acyclic graph of the multiple computing subtasks is constructed.

[0027] In the specific implementation process, a directed acyclic graph is generated with each subtask as a vertex and the dependency relationship between each subtask as an edge, so that the Internet of Vehicles computing tasks can be executed in a distributed manner, the computing resources can be scheduled rationally, and the tasks can be quickly calculated and executed.

[0028] Optionally, step 103 further includes the following steps: Step 1031, the computing subtask includes at least a task generating source and a neighboring device sensor signal, and a suitable computing submodel is selected based on the task generating source and the neighboring device sensor signal to construct a directed acyclic graph of multiple computing subtasks.

[0029] During the specific implementation process, the task source of each computing subtask is obtained to determine the urgency of the computing task. Whether the sensors of nearby devices have computing capabilities will also be included in the calculation scope. If the computing task is urgent, edge computing can be performed through the sensors of nearby devices to reduce the time spent on data transmission, so that each device in the vehicle can respond quickly.

[0030] Step 1032, obtain the dynamic changes of the network topology of the vehicle terminal in real time, and control the dynamic exit or entry of the multi-level mobile computing power network model based on the dynamic changes of the network topology.

[0031] During the specific implementation process, if an emergency occurs in the vehicle, such as dangerous situation avoidance and other sudden problems, which are not mentioned in the Internet of Vehicles computing tasks, in order to avoid accidents, it is necessary to dynamically increase the application of multi-level mobile computing network models, and control the intervention of cloud computing sub-models, edge computing sub-models and end computing sub-models based on actual conditions, so as to further improve the rationality of computing resource scheduling.

[0032] Step 104: Perform computing resource scheduling calculations on the multiple computing subtasks based on the directed acyclic graph to obtain a subtask execution priority sequence.

[0033] During the specific implementation process, computing subtasks with shorter time limits are sorted and computing resources are allocated preferentially, making the scheduling of computing resources more reasonable.

[0034] Optionally, step 104 further includes the following steps: Step 1041, the computing power resource scheduling calculation includes at least computing power attribute calculation and deep reinforcement learning network calculation, and the subtask execution priority sequence is obtained based on the computing speed, computing energy consumption and deep reinforcement learning network.

[0035] In the specific implementation process, it is necessary to consider the computing speed and computing energy consumption of each computing sub-model according to the urgency of the computing task, the type of computing, and the amount of computing, so as to select the appropriate computing sub-model, so as to achieve a more reasonable scheduling of computing resources; through the deep reinforcement learning network, it is possible to more quickly improve the calculation of various information in the process of autonomous driving of the vehicle, further improve the scheduling speed of computing resources, thereby rationalizing the scheduling of computing resources and saving running time.

[0036] The present application also discloses a resource scheduling system based on the Internet of Vehicles, comprising: Information interaction communication module 201: used to obtain vehicle networking computing tasks issued by vehicle terminals, wherein the vehicle networking computing tasks at least include identification codes for identifying different vehicle terminals, obtaining sensor signals of vehicles corresponding to the identification codes, and establishing communication connections with corresponding vehicles based on the sensor signals.

[0037] Optionally, the information interaction communication module is also used to store the Internet of Vehicles computing tasks in a historical task database preset in the cloud, obtain the Internet of Vehicles computing tasks with corresponding identification codes in the historical task database, and build a deep reinforcement learning network.

[0038] Optionally, the information interaction communication module is also used to obtain dynamic changes in the network topology of the vehicle terminal in real time.

[0039] Computing power network construction module 202: used to construct a multi-level mobile computing power network model based on all sensor signals, wherein the multi-level mobile computing power network model includes but is not limited to a cloud computing sub-model, an edge computing sub-model and an end computing sub-model.

[0040] The computing task decomposition module 203 is used to decompose the Internet of Vehicles computing task based on a multi-level mobile computing network model to obtain multiple computing subtasks and construct a directed acyclic graph of the multiple computing subtasks.

[0041] Optionally, the computing task decomposition module is also used to select a suitable computing sub-model based on the task generation source and the sensor signal of the adjacent device, and construct a directed acyclic graph of multiple computing sub-tasks.

[0042] Optionally, the computing power task decomposition module is also used to control the dynamic exit or entry of a multi-level mobile computing power network model based on dynamic changes in network topology.

[0043] The computing power resource scheduling module 204 is used to perform computing power resource scheduling calculations on the multiple computing subtasks based on the directed acyclic graph to obtain a subtask execution priority sequence.

[0044] Optionally, the computing resource scheduling module is also used to obtain a subtask execution priority sequence based on computing speed, computing energy consumption and a deep reinforcement learning network.

[0045] An embodiment of the present application also discloses a computing resource scheduling platform, including a memory and a processor. The memory stores a computer program that can be loaded by the processor and execute the resource scheduling method based on the Internet of Vehicles as described above. The computer program can be at least one instruction, at least one program, a code set or an instruction set.

[0046] An embodiment of the present application also discloses a computer-readable storage medium, which stores a computer program that can be loaded by a processor and execute the resource scheduling method based on the Internet of Vehicles as described above. The computer-readable storage medium includes, for example: a USB flash drive, a mobile hard disk, a read-only memory (ROM), a random access memory (ROM), a magnetic disk or an optical disk, and other media that can store program codes.

[0047] The above embodiments are only used to illustrate the technical solutions of the present application, rather than to limit the protection scope of the application. Obviously, the described embodiments are only some embodiments of the present application, rather than all embodiments. Based on these embodiments, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present application.

Claims

1. A resource scheduling method based on Internet of Vehicles, characterized in that: The execution subject of the resource scheduling method based on the Internet of Vehicles is a computing resource scheduling platform, and the computing resource scheduling platform is communicatively connected to a cloud, and the cloud pre-stores identification codes for distinguishing different vehicle terminals, and the cloud also pre-stores the corresponding relationship between the identification codes and the sensor signals of all devices in the corresponding vehicle terminals; The method comprises: Acquire a vehicle networking computing task issued by a vehicle terminal, wherein the vehicle networking computing task at least includes an identification code for identifying different vehicle terminals, acquire a sensor signal of a vehicle corresponding to the identification code, and establish a communication connection with the corresponding vehicle based on the sensor signal; Constructing a multi-level mobile computing network model based on all the sensor signals, wherein the multi-level mobile computing network model includes but is not limited to a cloud computing sub-model, an edge computing sub-model and a terminal computing sub-model; Decomposing the Internet of Vehicles computing task based on a multi-level mobile computing power network model to obtain multiple computing subtasks, and constructing a directed acyclic graph of the multiple computing subtasks; Based on the directed acyclic graph, computing power resource scheduling calculation is performed on the multiple computing subtasks to obtain a subtask execution priority sequence.

2. The resource scheduling method based on the Internet of Vehicles according to claim 1, characterized in that: The multi-level mobile computing power network model is used to decompose the Internet of Vehicles computing task to obtain multiple computing subtasks, and also includes: the computing subtasks at least include a task generation source and a nearby device sensor signal, and a suitable computing submodel is selected based on the task generation source and the nearby device sensor signal to construct a directed acyclic graph of the multiple computing subtasks.

3. The resource scheduling method based on the Internet of Vehicles according to claim 1, characterized in that: The multi-level mobile computing power network model also includes computing power attributes corresponding to the cloud computing sub-model, the edge computing sub-model and the end computing sub-model, and the computing power attributes include at least computing speed and computing energy consumption. The computing power resource scheduling calculation includes at least computing power attribute calculation, and the sub-task execution priority sequence is obtained based on the computing speed and computing energy consumption.

4. The resource scheduling method based on the Internet of Vehicles according to claim 1, characterized in that: The method of decomposing the Internet of Vehicles computing task based on a multi-level mobile computing power network model to obtain multiple computing sub-tasks and constructing a directed acyclic graph of the multiple computing sub-tasks also includes: acquiring the dynamic changes of the network topology of the vehicle terminal in real time, and controlling the dynamic exit or entry of the multi-level mobile computing power network model based on the dynamic changes of the network topology.

5. The resource scheduling method based on the Internet of Vehicles according to claim 1, characterized in that: The method of obtaining the Internet of Vehicles computing tasks issued by the vehicle terminal, wherein the Internet of Vehicles computing tasks at least include an identification code for identifying different vehicle terminals, obtaining a sensor signal of a vehicle corresponding to the identification code, and establishing a communication connection with the corresponding vehicle based on the sensor signal, also includes: storing the Internet of Vehicles computing tasks in a historical task database preset in the cloud, obtaining the Internet of Vehicles computing tasks with corresponding identification codes in the historical task database, and constructing a deep reinforcement learning network. The computing power resource scheduling calculation also includes at least deep reinforcement learning network calculation.

6. A resource scheduling system based on Internet of Vehicles, characterized by: The resource scheduling system comprises: An information interaction communication module (201) is used to obtain a vehicle networking computing task issued by a vehicle terminal, wherein the vehicle networking computing task at least includes an identification code for identifying different vehicle terminals, obtain a sensor signal of a vehicle corresponding to the identification code, and establish a communication connection with the corresponding vehicle based on the sensor signal; A computing power network construction module (202): used to construct a multi-level mobile computing power network model based on all the sensor signals, wherein the multi-level mobile computing power network model includes but is not limited to a cloud computing sub-model, an edge computing sub-model and a terminal computing sub-model; A computing task decomposition module (203): used to decompose the Internet of Vehicles computing task based on a multi-level mobile computing network model to obtain a plurality of computing subtasks and construct a directed acyclic graph of the plurality of computing subtasks; A computing power resource scheduling module (204): used to perform computing power resource scheduling calculations on the multiple computing subtasks based on the directed acyclic graph to obtain a subtask execution priority sequence.

7. A computing resource scheduling platform, characterized by: The computing power resource scheduling platform includes a processor and a memory, wherein the memory stores at least one instruction, at least one program, a code set or an instruction set, and the at least one instruction, the at least one program, the code set or the instruction set are loaded and executed by the processor to implement the processing of the resource scheduling method based on the Internet of Vehicles as described in any one of claims 1 to 7.

8. A computer-readable storage medium, characterized in that: The storage medium stores at least one instruction, at least one program, a code set or an instruction set, and the at least one instruction, the at least one program, the code set or the instruction set are loaded and executed by the processor to implement the processing of the resource scheduling method based on the Internet of Vehicles as described in any one of claims 1 to 7.

Images