Task scheduling method and system based on automatic driving simulation test workflow

Abstract

The invention relates to a task scheduling method and system based on an automatic driving simulation test workflow, and the method comprises the steps: W1, obtaining the instruction data information of a user operating a corresponding task flow in the simulation test work of an automatic driving vehicle, and collecting the anchor value data information of each work node of each task in the workflow; and W2, on the basis of the anchor value data information of each working node of each task in the workflow, representing the mutual information value of each working node of each task by adopting a maximum correlation minimum redundancy feature selection algorithm based on adaptive adjustment to obtain mutual information value data information of each working node of each task in the workflow. According to the method, task issuing and automatic execution are carried out on the platform, the whole operation process is controlled, the problem is quickly positioned, and the development efficiency is improved.

Description

Technical Field

[0001] The present invention relates to the field of autonomous driving technology, and in particular to a task scheduling method and system based on an autonomous driving simulation test workflow. Background Art

[0002] The research on autonomous driving technology began in the 1980s, when the research focused on automatic cruising and road condition recognition. With the improvement of computing power and the increase in data volume, autonomous driving technology began to enter the practical stage in the 2010s. At present, many companies and research institutions are actively developing autonomous driving technology. The autonomous driving system uses advanced communication, computer, network and control technologies to achieve real-time and continuous control of trains. Using modern communication means and directly facing the train, two-way data communication between the train and the ground can be achieved, with fast transmission rate and large amount of information. Subsequent tracking trains and control centers can know the exact location of the preceding train in time, making operation management more flexible, control more effective, and more adapted to the needs of train autonomous driving. From the above, it can be concluded that autonomous driving technology is developing rapidly, but the development and testing process of autonomous driving involves many professions and complex processes. How to simplify it has become a problem that we need to solve urgently. Summary of the invention

[0003] In view of the above problems, the present invention provides a task scheduling method and system based on the autonomous driving simulation test workflow, which not only issues tasks and automatically executes them on the platform, but also controls the entire operation process, quickly locates problems, and improves development efficiency.

[0004] In order to achieve the above-mentioned purpose and other related purposes, the technical solution provided by the present invention is as follows:

[0005] A task scheduling method based on an autonomous driving simulation test workflow, the method comprising:

[0006] W1. In the simulation test of the autonomous driving vehicle, obtain the instruction data information of the user's operation of the corresponding task flow, and collect the anchor value data information of each work node of each task in the workflow;

[0007] W2. Based on the anchor value data information of each working node of each task in the workflow, the mutual information value of each working node of each task is characterized by using the maximum relevance minimum redundancy feature selection algorithm based on adaptive adjustment to obtain the mutual information value data information of each working node of each task in the workflow;

[0008] W3. Based on the mutual information value data information of each work node of each task in the workflow and the instruction data information of the user operating the corresponding task flow, a hierarchical clustering algorithm based on the golden sine is used to characterize the dependency relationship between the instruction and each work node, and the dependency matrix data information of the user instruction and each work node is obtained;

[0009] W4. Based on the dependency matrix data information between the user instructions and each work node, a task scheduling function Q of the workflow is constructed, each task in the workflow is scheduled, and the task scheduling data information of the workflow is obtained.

[0010] Furthermore, the task scheduling function Q of the workflow is:

[0011]

[0012] Among them, x is the dependency matrix data information between user instructions and each working node, and α1, α2 and α3 are the task scheduling factors of the workflow.

[0013] Furthermore, the task scheduling factors α1, α2 and α3 of the workflow are:

[0014]

[0015] Among them, x is the dependency matrix data information between user instructions and each working node.

[0016] Furthermore, the constraint function f of the task scheduling factors α1, α2 and α3 of the workflow is,

[0017]

[0018] The value range of the constraint function f is (1,2).

[0019] Furthermore, in step W2, the use of the maximum relevance minimum redundancy feature selection algorithm based on adaptive adjustment to characterize the mutual information value of each working node of each task includes:

[0020] W21. Based on the anchor value data information of each working node of each task in the workflow, establish a feature selection function R of the anchor value of the working node,

[0021]

[0022] Among them, y is the anchor value data information of each work node of each task in the workflow, β1, β2 and β3 are the feature selection factors of the work node, and the feature relationship of each work node of each task is characterized to obtain the feature relationship data information of each work node of each task;

[0023] W22. Based on the characteristic relationship data information of each working node of each task, establish the mutual information value function P of the working node,

[0024]

[0025] Among them, z is the characteristic relationship data information of each working node of each task, δ1, δ2 and δ3 are the adaptive adjustment factors of the working nodes;

[0026] W23. Based on the mutual information value function P of the working nodes, the mutual information value of each working node of each task is characterized to obtain the mutual information value data information of each working node of each task in the workflow.

[0027] Furthermore, the adaptive adjustment factors δ1, δ2 and δ3 of the working node are:

[0028]

[0029] Among them, z is the feature relationship data information of each working node of each task.

[0030] Furthermore, the constraint function g of the characteristic decision factors β1, β2 and β3 of the working node is,

[0031]

[0032] Among them, the value range of the constraint function g is (2,4).

[0033] Furthermore, in step W3, the characterization of the dependency relationship between the instructions and each work node using the golden sine-based hierarchical clustering algorithm includes:

[0034] W31. Based on the mutual information value data information of each work node of each task in the workflow, a cluster family data set of each work node of each task in the workflow is constructed, and based on the instruction data information of the user operating the corresponding task flow, a cluster family data set of the instruction of the user operating the corresponding task flow is constructed;

[0035] W32. Based on the clustering data set of the instructions of the user operating the corresponding task flow and the clustering data set of each working node of each task in the workflow, a Hausdorff distance function G between the working node and the instruction is established.

[0036]

[0037] Where X is the clustered dataset of the instructions of the user operating the corresponding task flow, Y is the clustered dataset of each working node of each task in the workflow, a is the element in the clustered dataset of the instructions of the user operating the corresponding task flow, b is the element in the clustered dataset of each working node of each task in the workflow, γ and η are the golden sine factors;

[0038] W33. Based on the Hausdorff distance function G between the working node and the instruction, the dependency relationship between the instruction and each working node is characterized to obtain the dependency matrix data information between the user instruction and each working node.

[0039] In order to achieve the above-mentioned objectives and other related objectives, the present invention also provides a task scheduling system based on an autonomous driving simulation test workflow, including a computer device, which is programmed or configured to execute any one of the steps of the task scheduling method based on the autonomous driving simulation test workflow.

[0040] In order to achieve the above-mentioned purpose and other related purposes, the present invention also provides a computer-readable storage medium, which stores a computer program programmed or configured to execute any one of the task scheduling methods based on the autonomous driving simulation test workflow.

[0041] The present invention has the following positive effects:

[0042] 1. The present invention characterizes the mutual information value of each working node of each task by adopting the maximum correlation minimum redundancy feature selection algorithm based on adaptive adjustment, and characterizes the dependency relationship between instructions and each working node by combining the hierarchical clustering algorithm based on golden sine, so as to obtain the dependency matrix data information between user instructions and each working node. It can not only issue tasks and automatically execute them on the platform, realize the control of the entire operation process, quickly locate problems, and improve development efficiency, but also adopt a distributed architecture to enable multiple nodes to collaborate to complete a complete workflow, get rid of the physical isolation of different servers, so that scattered servers can be combined to form a working cluster, saving the cost of purchasing expensive and high-performance servers.

[0043] 2. The present invention schedules each task in the workflow by constructing a task scheduling function Q of the workflow. It not only has the characteristics of high availability and high fault tolerance, but also can realize uninterrupted computing services by reasonably constructing a computing cluster. It can also realize elastic capacity expansion and increase computing resources without stopping to meet higher computing needs. BRIEF DESCRIPTION OF THE DRAWINGS

[0044] Figure 1 It is a schematic diagram of the method flow of the present invention;

[0045] Figure 2 A schematic flow chart of the maximum relevance minimum redundancy feature selection algorithm for adaptive adjustment of the present invention;

[0046] Figure 3 It is a schematic diagram of the process of the hierarchical clustering algorithm based on the golden sine of the present invention. DETAILED DESCRIPTION

[0047] The following is a description of exemplary embodiments of the present disclosure in conjunction with the accompanying drawings, including various details of the embodiments of the present disclosure to facilitate understanding, which should be considered as merely exemplary. Therefore, it should be recognized by those of ordinary skill in the art that various changes and modifications may be made to the embodiments described herein without departing from the scope and spirit of the present disclosure. Similarly, for the sake of clarity and conciseness, descriptions of well-known functions and structures are omitted in the following description.

[0048] Example 1: Figure 1 As shown, a task scheduling method based on an autonomous driving simulation test workflow comprises:

[0049] W1. In the simulation test of the autonomous driving vehicle, obtain the instruction data information of the user's operation of the corresponding task flow, and collect the anchor value data information of each work node of each task in the workflow;

[0050] W2. Based on the anchor value data information of each working node of each task in the workflow, the mutual information value of each working node of each task is characterized by using the maximum relevance minimum redundancy feature selection algorithm based on adaptive adjustment to obtain the mutual information value data information of each working node of each task in the workflow;

[0051] W3. Based on the mutual information value data information of each work node of each task in the workflow and the instruction data information of the user operating the corresponding task flow, a hierarchical clustering algorithm based on the golden sine is used to characterize the dependency relationship between the instruction and each work node, and the dependency matrix data information of the user instruction and each work node is obtained;

[0052] W4. Based on the dependency matrix data information between the user instructions and each work node, a task scheduling function Q of the workflow is constructed, each task in the workflow is scheduled, and the task scheduling data information of the workflow is obtained.

[0053] In this embodiment, the task scheduling function Q of the workflow is:

[0054]

[0055] Among them, x is the dependency matrix data information between user instructions and each working node, and α1, α2 and α3 are the task scheduling factors of the workflow.

[0056] In this embodiment, the task scheduling factors α1, α2 and α3 of the workflow are:

[0057]

[0058] Among them, x is the dependency matrix data information between user instructions and each working node.

[0059] In this embodiment, the constraint function f of the task scheduling factors α1, α2 and α3 of the workflow is,

[0060]

[0061] The value range of the constraint function f is (1,2).

[0062] In this embodiment, if Figure 2 As shown, in step W2, the use of the maximum relevance minimum redundancy feature selection algorithm based on adaptive adjustment to characterize the mutual information value of each working node of each task includes:

[0063] W21. Based on the anchor value data information of each working node of each task in the workflow, establish a feature selection function R of the anchor value of the working node,

[0064]

[0065] Among them, y is the anchor value data information of each work node of each task in the workflow, β1, β2 and β3 are the feature selection factors of the work node, and the feature relationship of each work node of each task is characterized to obtain the feature relationship data information of each work node of each task;

[0066] W22. Based on the characteristic relationship data information of each working node of each task, establish the mutual information value function P of the working node,

[0067]

[0068] Among them, z is the characteristic relationship data information of each working node of each task, δ1, δ2 and δ3 are the adaptive adjustment factors of the working nodes;

[0069] W23. Based on the mutual information value function P of the working nodes, the mutual information value of each working node of each task is characterized to obtain the mutual information value data information of each working node of each task in the workflow.

[0070] In this embodiment, the adaptive adjustment factors δ1, δ2 and δ3 of the working node are:

[0071]

[0072] Among them, z is the feature relationship data information of each working node of each task.

[0073] In this embodiment, the constraint function g of the feature selection factors β1, β2 and β3 of the working node is,

[0074]

[0075] Among them, the value range of the constraint function g is (2,4).

[0076] Example 2: Based on the task scheduling method based on the autonomous driving simulation test workflow in Example 1, the present invention is further illustrated and described below.

[0077] like Figure 1 As shown, a task scheduling method based on an autonomous driving simulation test workflow comprises:

[0078] W1. In the simulation test of the autonomous driving vehicle, obtain the instruction data information of the user's operation of the corresponding task flow, and collect the anchor value data information of each work node of each task in the workflow;

[0079] W2. Based on the anchor value data information of each working node of each task in the workflow, the mutual information value of each working node of each task is characterized by using the maximum relevance minimum redundancy feature selection algorithm based on adaptive adjustment to obtain the mutual information value data information of each working node of each task in the workflow;

[0080] W3. Based on the mutual information value data information of each work node of each task in the workflow and the instruction data information of the user operating the corresponding task flow, a hierarchical clustering algorithm based on the golden sine is used to characterize the dependency relationship between the instruction and each work node, and the dependency matrix data information of the user instruction and each work node is obtained;

[0081] W4. Based on the dependency matrix data information between the user instructions and each work node, a task scheduling function Q of the workflow is constructed, each task in the workflow is scheduled, and the task scheduling data information of the workflow is obtained.

[0082] In this embodiment, if Figure 3 As shown, in step W3, the use of the golden sine-based hierarchical clustering algorithm to characterize the dependency relationship between the instructions and each work node includes:

[0083] W31. Based on the mutual information value data information of each work node of each task in the workflow, a cluster family data set of each work node of each task in the workflow is constructed, and based on the instruction data information of the user operating the corresponding task flow, a cluster family data set of the instruction of the user operating the corresponding task flow is constructed;

[0084] W32. Based on the clustering data set of the instructions of the user operating the corresponding task flow and the clustering data set of each working node of each task in the workflow, a Hausdorff distance function G between the working node and the instruction is established.

[0085]

[0086] Where X is the clustered dataset of the instructions of the user operating the corresponding task flow, Y is the clustered dataset of each working node of each task in the workflow, a is the element in the clustered dataset of the instructions of the user operating the corresponding task flow, b is the element in the clustered dataset of each working node of each task in the workflow, γ and η are the golden sine factors;

[0087] W33. Based on the Hausdorff distance function G between the working node and the instruction, the dependency relationship between the instruction and each working node is characterized to obtain the dependency matrix data information between the user instruction and each working node.

[0088] In this embodiment, the present invention provides a task scheduling system based on an autonomous driving simulation test workflow, including a computer device that is programmed or configured to execute any one of the steps of the task scheduling method based on an autonomous driving simulation test workflow.

[0089] In this embodiment, a task scheduling system based on an autonomous driving simulation test workflow is provided, the system comprising:

[0090] The data acquisition module is used to obtain the instruction data information of the user's operation of the corresponding task flow, and collect the anchor value data information of each working node of each task in the workflow;

[0091] The mutual information value module of the working node is connected to the data acquisition module and is used to characterize the mutual information value of each working node of each task by using the maximum relevance minimum redundancy feature selection algorithm based on adaptive adjustment, so as to obtain the mutual information value data information of each working node of each task in the workflow;

[0092] The instruction-to-work node relationship module is connected to the mutual information value module of the work node, and is used to characterize the dependency relationship between the instruction and each work node by using a hierarchical clustering algorithm based on the golden sine, and obtain the dependency relationship matrix data information between the user instruction and each work node;

[0093] The task scheduling module is connected to the instruction and work node relationship module, and is used to construct a task scheduling function Q of the workflow, schedule each task in the workflow, and obtain task scheduling data information of the workflow.

[0094] In this embodiment, the present invention provides a computer-readable storage medium, on which is stored a computer program programmed or configured to execute any one of the task scheduling methods based on the autonomous driving simulation test workflow.

[0095] Any reference to memory, storage, database or other media used in the embodiments provided in the present application may include non-volatile and / or volatile memory. Non-volatile memory may include read-only memory (ROM), programmable ROM (PROM), electrically programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM) or flash memory. Volatile memory may include random access memory (RAM) or external cache memory. As an illustration and not limitation, RAM is available in various forms, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), double data rate SDRAM (DDRSDRAM), enhanced SDRAM (ESDRAM), synchronous link (Synchlink) DRAM (SLDRAM), memory bus (Rambus) direct RAM (RDRAM), direct memory bus dynamic RAM (DRDRAM) and memory bus dynamic RAM (RDRAM).

[0096] In summary, the present invention not only issues tasks and automatically executes them on the platform, but also controls the entire operation process, quickly locates problems, and improves development efficiency.

[0097] The above specific implementations do not constitute a limitation on the protection scope of the present disclosure. It should be understood by those skilled in the art that various modifications, combinations, sub-combinations and substitutions can be made according to design requirements and other factors. Any modification, equivalent substitution and improvement made within the spirit and principle of the present disclosure shall be included in the protection scope of the present disclosure.

Claims

1. A task scheduling method based on an autonomous driving simulation test workflow, characterized in that: The method comprises: W1. In the simulation test of the autonomous driving vehicle, obtain the instruction data information of the user's operation of the corresponding task flow, and collect the anchor value data information of each work node of each task in the workflow; W2. Based on the anchor value data information of each working node of each task in the workflow, the mutual information value of each working node of each task is characterized by using the maximum relevance minimum redundancy feature selection algorithm based on adaptive adjustment to obtain the mutual information value data information of each working node of each task in the workflow; W3. Based on the mutual information value data information of each work node of each task in the workflow and the instruction data information of the user operating the corresponding task flow, a hierarchical clustering algorithm based on the golden sine is used to characterize the dependency relationship between the instruction and each work node, and the dependency matrix data information of the user instruction and each work node is obtained; W4. Based on the dependency matrix data information between the user instructions and each work node, a task scheduling function Q of the workflow is constructed, each task in the workflow is scheduled, and the task scheduling data information of the workflow is obtained.

2. The task scheduling method based on the autonomous driving simulation test workflow according to claim 1 is characterized in that: The task scheduling function Q of the workflow is: Among them, x is the dependency matrix data information between user instructions and each working node, and α1, α2 and α3 are the task scheduling factors of the workflow.

3. The task scheduling method based on the autonomous driving simulation test workflow according to claim 2 is characterized in that: The task scheduling factors α1, α2 and α3 of the workflow are: Among them, x is the dependency matrix data information between user instructions and each working node.

4. The task scheduling method based on the autonomous driving simulation test workflow according to claim 3 is characterized in that: The constraint function f of the task scheduling factors α1, α2 and α3 of the workflow is, The value range of the constraint function f is (1,2).

5. The task scheduling method based on the autonomous driving simulation test workflow according to claim 1 is characterized in that: In step W2, the use of the maximum relevance minimum redundancy feature selection algorithm based on adaptive adjustment to characterize the mutual information value of each working node of each task includes: W21. Based on the anchor value data information of each working node of each task in the workflow, establish a feature selection function R of the anchor value of the working node, Among them, y is the anchor value data information of each work node of each task in the workflow, β1, β2 and β3 are the feature selection factors of the work node, and the feature relationship of each work node of each task is characterized to obtain the feature relationship data information of each work node of each task; W22. Based on the characteristic relationship data information of each working node of each task, establish the mutual information value function P of the working node, Among them, z is the characteristic relationship data information of each working node of each task, δ1, δ2 and δ3 are the adaptive adjustment factors of the working nodes; W23. Based on the mutual information value function P of the working nodes, the mutual information value of each working node of each task is characterized to obtain the mutual information value data information of each working node of each task in the workflow.

6. The task scheduling method based on the autonomous driving simulation test workflow according to claim 5 is characterized in that: The adaptive adjustment factors δ1, δ2 and δ3 of the working node are: Among them, z is the feature relationship data information of each working node of each task.

7. The task scheduling method based on the autonomous driving simulation test workflow according to claim 5 is characterized in that: The constraint function g of the characteristic decision factors β1, β2 and β3 of the working node is, Among them, the value range of the constraint function g is (2,4).

8. The task scheduling method based on the autonomous driving simulation test workflow according to claim 1 is characterized in that: In step W3, the use of the golden sine-based hierarchical clustering algorithm to characterize the dependency relationship between instructions and each work node includes: W31. Based on the mutual information value data information of each work node of each task in the workflow, a cluster family data set of each work node of each task in the workflow is constructed, and based on the instruction data information of the user operating the corresponding task flow, a cluster family data set of the instruction of the user operating the corresponding task flow is constructed; W32. Based on the clustering data set of the instructions of the user operating the corresponding task flow and the clustering data set of each working node of each task in the workflow, a Hausdorff distance function G between the working node and the instruction is established. Where X is the clustered dataset of the instructions of the user operating the corresponding task flow, Y is the clustered dataset of each working node of each task in the workflow, a is the element in the clustered dataset of the instructions of the user operating the corresponding task flow, b is the element in the clustered dataset of each working node of each task in the workflow, γ and η are the golden sine factors; W33. Based on the Hausdorff distance function G between the working node and the instruction, the dependency relationship between the instruction and each working node is characterized to obtain the dependency matrix data information between the user instruction and each working node.

9. A task scheduling system based on an autonomous driving simulation test workflow, comprising a computer device, characterized in that: The computer device is programmed or configured to execute the steps of the task scheduling method based on the autonomous driving simulation test workflow as described in any one of claims 1 to 8.

10. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores a computer program that is programmed or configured to execute the task scheduling method based on the autonomous driving simulation test workflow as described in any one of claims 1 to 8.

Images