Performance monitoring method and system for cloud platform of autonomous driving events
By counting and comparing the processing amount difference per second in the cloud platform of autonomous driving events, the problem of difficulty in positioning the bottleneck of data channel in the existing technology is solved, and automated performance monitoring and remote optimization functions are realized.
Patent Information
- Application Number
- CN202210967797.7
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Filing Date
- 2022-08-12
- Publication Date
- 2024-11-01
- Estimated Expiration
- 2042-08-12
AI Technical Summary
The existing technology is difficult to locate which part of the data channel has bottlenecks in the process of uploading data to the cloud when facing a huge amount of data, making it difficult for maintenance personnel to process remotely in a timely manner.
By setting up the first and second data channels between the autonomous driving controller, the on-board computer and the cloud platform, counting the processing volumes per second TPS1, TPS2, TPS3, and TPS4, and comparing the difference to judge the performance bottleneck position, and generating a message file for Positioning alarm.
It realizes performance monitoring of the data transmission process of uploading autonomous driving events to the cloud platform, and can automatically locate the performance bottleneck position of the data channel, making it easier for maintenance personnel to optimize remotely.
Smart Images

Figure CN115339465B_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to the technical field of autonomous driving data platforms, and in particular to a performance monitoring method and system for an autonomous driving event cloud platform. Background Art
[0002] As autonomous driving technology becomes more and more mature, the types and scale of data collected by autonomous driving controllers are becoming increasingly large. Therefore, ensuring that the data collected by autonomous driving controllers can be uploaded to the cloud without bottlenecks within a certain period of time has become a key link in the performance evaluation of autonomous driving data platforms.
[0003] At present, the architecture of the autonomous driving data platform includes three parts: the hardware end consisting of the autonomous driving controller (C2), the on-board computer (EDC), and the cloud. The three parts transmit autonomous driving events through data channels. The autonomous driving controller is the control hardware after the autonomous driving event occurs. It collects various autonomous driving events and then sends them to the on-board computer. The on-board computer acts as an intermediate forwarding device to send the autonomous driving events to the cloud. The specific receiving device in the cloud is the cloud platform server.
[0004] The prior art evaluates the performance of the data channel in the network architecture for uploading autonomous driving events to the cloud by making a rough estimate based on the total number of autonomous driving events uploaded from the hardware end to the cloud and the total time spent uploading all events. The estimation result is then used to determine whether there is performance degradation or bottleneck in the process of uploading autonomous driving events to the cloud.
[0005] However, when faced with huge amounts of data, existing technologies for evaluating the data channel and platform processing performance of the entire network architecture cannot be used to automatically locate which part of the data channel has a bottleneck in the process of uploading autonomous driving event data to the cloud, resulting in reduced transmission performance, making it difficult for maintenance personnel to perform remote processing in a timely manner. Summary of the invention
[0006] One of the purposes of the present invention is to provide a performance monitoring method for an autonomous driving event cloud platform to solve the problem in the prior art that it is difficult to locate which part of the data channel has a bottleneck; the second purpose is to provide a performance monitoring system for an autonomous driving event cloud platform.
[0007] In order to achieve the above object, the technical solution adopted by the present invention is as follows:
[0008] A performance monitoring method for an autonomous driving event on a cloud platform is performed on a system consisting of an autonomous driving controller, an on-board computer, and a cloud platform; a first data channel is provided between the autonomous driving controller and the on-board computer, and a second data channel is provided between the on-board computer and the cloud platform; the steps are as follows:
[0009] The autonomous driving controller counts the autonomous driving event processing volume per second sent through the first data channel, recorded as TPS1, and then sends TPS1 to the on-board computer through the first data channel;
[0010] The on-board computer counts the processing volume TPS2 of the autonomous driving events received through the first data channel per second;
[0011] The cloud platform counts the processing volume TPS3 of the autonomous driving events received through the second data channel, and then sends TPS3 to the on-board computer through the second data channel;
[0012] The on-board computer counts the processing volume TPS4 of the autonomous driving events received through the second data channel;
[0013] The on-board computer compares the differences among TPS1, TPS2, TPS3, and TPS4, and then determines whether performance bottlenecks occur in the first data channel and the second data channel based on the comparison results;
[0014] The specific process of comparing the difference is:
[0015] The on-board computer compares the differences between TPS1 and TPS2, TPS2 and TPS3, and TPS3 and TPS4 to see if they are within the set range;
[0016] The specific process of determining whether a performance bottleneck occurs is as follows:
[0017] If TPS1 is greater than TPS2 and the difference is beyond the set range, it means that the first data channel has a performance bottleneck;
[0018] If TPS2 is greater than TPS3 and the difference is beyond the set range, it means that the upload performance of the second data channel has a bottleneck;
[0019] If TPS3 is greater than TPS4 and the difference exceeds the set range, it means that a performance bottleneck occurs in the download of the second data channel.
[0020] Furthermore, when comparing TPS1 and TPS2, the range is set such that the difference between TPS1 and TPS2 accounts for less than 5 percent of the size of TPS1;
[0021] When comparing TPS2 and TPS3, the setting range is that the difference between TPS2 and TPS3 accounts for less than 5% of the size of TPS2;
[0022] When comparing TPS3 and TPS4, the setting range is that the difference between TPS3 and TPS4 accounts for less than 5% of the size of TPS3.
[0023] Furthermore, after the on-board computer compares and finds the performance bottleneck, the on-board computer generates a message file corresponding to the performance bottleneck, and sends it to the cloud platform through the second data channel for positioning and warning.
[0024] Furthermore, a delay process is performed before TPS1, TPS2, TPS3, and TPS4 are compared. The specific process is as follows:
[0025] Before comparing TPS1 and TPS2, add the first set value to the original size of TPS2;
[0026] Before comparing TPS2 and TPS3, add the second set value to the original size of TPS3;
[0027] Before comparing TPS3 and TPS4, the third setting value is added to the original size of TPS4.
[0028] A performance monitoring system for an autonomous driving event on a cloud platform includes an autonomous driving controller, an on-board computer, and a cloud platform; a first data channel is provided between the autonomous driving controller and the on-board computer, and a second data channel is provided between the on-board computer and the cloud platform; the autonomous driving controller is provided with a first statistical module; the on-board computer is provided with a second statistical module and a comparison and judgment module;
[0029] The first statistical module is used to count the processing volume TPS1 of the autonomous driving events sent by the autonomous driving controller through the first data channel per second;
[0030] The second statistical module is used to count the number of autonomous driving events processed per second TPS2 received by the on-board computer through the first data channel, the number of autonomous driving events processed per second TPS4 received by the on-board computer through the second data channel, and the number of autonomous driving events processed per second TPS3 received by the cloud platform through the second data channel;
[0031] The comparison and judgment module is used to compare the differences among TPS1, TPS2, TPS3, and TPS4, and judge whether the first data channel and the second data channel have performance bottlenecks according to the comparison results;
[0032] The specific process of comparing the difference in the comparison and judgment module is as follows:
[0033] Compare the differences between TPS1 and TPS2, TPS2 and TPS3, and TPS3 and TPS4 to see if they are within the set range;
[0034] The specific process of the comparison and judgment module to determine whether a performance bottleneck occurs is as follows:
[0035] If TPS1 is greater than TPS2 and the difference is beyond the set range, it means that the first data channel has a performance bottleneck;
[0036] If TPS2 is greater than TPS3 and the difference is beyond the set range, it means that the upload performance of the second data channel has a bottleneck;
[0037] If TPS3 is greater than TPS4 and the difference exceeds the set range, it means that a performance bottleneck occurs in the download of the second data channel.
[0038] Further, the comparison and judgment module includes a first calculation module, a second calculation module, a third calculation module, and a judgment module;
[0039] The first calculation module is used to compare whether the ratio of the difference between TPS1 and TPS2 to the value of TPS1 exceeds a set range;
[0040] The second calculation module is used to compare whether the ratio of the difference between TPS2 and TPS3 to the value of TPS1 exceeds a set range;
[0041] The third calculation module is used to compare whether the ratio of the difference between TPS3 and TPS4 to the value of TPS1 exceeds a set range;
[0042] The judgment module is used to judge whether the calculation results of the first comparison module, the second comparison module and the third comparison module are beyond the set range, so as to locate whether the first data channel and the second data channel have performance bottlenecks.
[0043] Furthermore, the onboard computer is also provided with a delay processing module;
[0044] The delay processing module is used to add the first set value to the original size of TPS2, the second set value to the original size of TPS3, and the third set value to the original size of TPS4 before the comparison and judgment module compares the sizes of TPS1, TPS2, TPS3, and TPS4.
[0045] Furthermore, the onboard computer is also provided with a message file generation module;
[0046] The message file generation module is used to generate a message file corresponding to the performance bottleneck.
[0047] The performance monitoring method and system of the cloud platform for autonomous driving events of the present invention have the following beneficial effects:
[0048] Faced with the increasing demand for autonomous driving control with huge amounts of data, compared with the problem that existing technologies cannot automatically locate performance bottlenecks, the present invention obtains the specific location of the performance bottleneck in the data channel by comparing the difference between the TPS of autonomous driving events in the links, and realizes performance monitoring of the data transmission process of uploading autonomous driving events to the cloud platform, which is convenient for maintenance personnel to remotely optimize performance bottlenecks. BRIEF DESCRIPTION OF THE DRAWINGS
[0049] Figure 1 This is a flow chart of the performance monitoring method of the cloud platform for autonomous driving events of the present invention;
[0050] Figure 2 This is a structural framework diagram of the performance monitoring system of the autonomous driving event cloud platform of the present invention. DETAILED DESCRIPTION
[0051] The following will describe the embodiments of the present invention with reference to the accompanying drawings and preferred embodiments. Those skilled in the art can easily understand other advantages and effects of the present invention from the contents disclosed in this specification. The present invention can also be implemented or applied through other different specific embodiments, and the details in this specification can also be modified or changed in various ways based on different viewpoints and applications without departing from the spirit of the present invention. It should be understood that the preferred embodiments are only for illustrating the present invention, not for limiting the scope of protection of the present invention.
[0052] It should be noted that the illustrations provided in the following embodiments are only schematic illustrations of the basic concept of the present invention. The drawings only show components related to the present invention rather than being drawn according to the number, shape and size of components in actual implementation. In actual implementation, the type, quantity and proportion of each component may be changed arbitrarily, and the component layout may also be more complicated.
[0053] Example 1
[0054] This embodiment proposes a performance monitoring method for an autonomous driving event cloud platform, such as Figure 1 shown.
[0055] The performance monitoring method for uploading autonomous driving events to a cloud platform in this embodiment is performed based on a system consisting of an autonomous driving controller, an onboard computer, and a cloud platform. A first data channel is provided between the autonomous driving controller and the onboard computer, and a second data channel is provided between the onboard computer and the cloud platform. Both the first data channel and the second data channel are used to transmit autonomous driving events and related messages. The specific steps are as follows:
[0056] S1, the autonomous driving controller counts the number of autonomous driving events processed per second sent through the first data channel, recorded as TPS1, and then sends TPS1 to the on-board computer through the first data channel;
[0057] S2, the vehicle computer counts the processing volume of the autonomous driving events received through the first data channel per second, recorded as TPS2;
[0058] S3: The cloud platform counts the processing volume of autonomous driving events received through the second data channel per second, recorded as TPS3, and then sends TPS3 to the on-board computer through the second data channel;
[0059] S4, the vehicle computer counts the processing volume of the autonomous driving events received through the second data channel per second, recorded as TPS4;
[0060] S5, perform delay processing in the vehicle computer, and then compare TPS1 with TPS2 after adding the first set value, TPS2 with TPS3 after adding the second set value, and TPS3 with TPS4 after adding the third set value; the specific process is:
[0061] S51, before comparing TPS1 and TPS2, add the first set value to the original counted value of TPS2, calculate the difference between TPS1 and TPS2 after adding the first set value, and then calculate the ratio of the difference to TPS1, which is recorded as ratio 1;
[0062] S52, before comparing TPS2 and TPS3, add the second set value to the original counted value of TPS3, calculate the difference between TPS2 and TPS3 after adding the second set value, and then calculate the ratio of the difference to TPS2, which is recorded as ratio 2;
[0063] S53, before comparing TPS3 and TPS4, add the third set value to the original counted value of TPS4, calculate the difference between TPS3 and TPS4 after adding the third set value, and then calculate the ratio of the difference to TPS3, which is recorded as ratio 3;
[0064] S6. According to the calculation result of step S5, the on-board computer determines whether the size of ratio 1, ratio 2, and ratio 3 exceeds the set range, so as to locate the position where the performance bottleneck occurs. The setting range is zero to five percent. The performance bottleneck includes performance bottleneck and performance bottleneck. The specific process is as follows:
[0065] S61. If the value of ratio 1 exceeds 5%, it indicates that a performance bottleneck occurs in the first data channel;
[0066] S62: If the value of ratio 2 exceeds 5%, it indicates that a performance bottleneck occurs in the second data channel when uploading the autonomous driving event;
[0067] S63: If the value of ratio 3 exceeds 5%, it indicates that a performance bottleneck occurs in the second data channel when downloading the autonomous driving event;
[0068] S7. Based on the performance bottleneck located in step S6, a message file corresponding to the performance bottleneck is generated in the on-board computer, wherein the message file records the number of autonomous driving events lost by the first data channel, the number of autonomous driving events lost by the second data channel during upload, and the number of autonomous driving events lost by the second data channel during download; the on-board computer then sends the message file to the cloud platform through the second data channel to assist maintenance personnel in remote positioning and alarming, thereby realizing performance monitoring of the process of transmitting autonomous driving events to the cloud platform.
[0069] Compared with the prior art, the beneficial effects of this embodiment are:
[0070] In the face of the increasing demand for autonomous driving control with huge amounts of data, compared with the problem that existing technologies cannot automatically locate performance bottlenecks, this embodiment compares whether the difference ratios among TPS1, TPS2, TPS3, and TPS4 are within a reasonable range, and combines the errors that may be caused by delays in the TPS transmission process to obtain the specific location of the performance bottleneck in the data channel, thereby realizing automated performance monitoring of the data transmission process of uploading autonomous driving events to the cloud platform, and facilitating maintenance personnel to remotely optimize performance bottlenecks.
[0071] Example 2
[0072] This embodiment proposes a performance monitoring system for an autonomous driving event cloud platform, such as Figure 2 shown.
[0073] The performance monitoring system for the cloud platform of the autonomous driving event of this embodiment includes an autonomous driving controller, a vehicle-mounted computer, and a cloud platform. A first data channel is provided between the autonomous driving controller and the vehicle-mounted computer. A second data channel is provided between the vehicle-mounted computer and the cloud platform.
[0074] The automatic driving controller is provided with a first statistical module. The first statistical module is used to count the automatic driving event processing volume per second TPS1 sent by the automatic driving controller through the first data channel. The first statistical module is connected to the first data channel.
[0075] The on-board computer is provided with a second statistical module, a delay processing module, a comparison and judgment module, and a message file generation module. The comparison and judgment module is used for. The comparison and judgment module includes a first calculation module, a second calculation module, a third calculation module, and a judgment module. The cloud platform is connected to the second data channel, and the second data channel is respectively connected to the judgment module, the message file generation module, and the second statistical module. The second data channel is logically divided into an upload link and a download link, and both the upload link and the download link are used to transmit autonomous driving events, TPS, and message files.
[0076] The second statistical module is connected to the first data channel. The second statistical module is used to count the number of autonomous driving events processed per second TPS2 received by the onboard computer through the first data channel, the number of autonomous driving events processed per second TPS4 received by the onboard computer through the second data channel, and the number of autonomous driving events processed per second TPS3 received by the cloud platform through the second data channel. The second statistical module is connected to the second data channel.
[0077] The second statistical module is connected to the delay processing module. The delay processing module is used to add the first set value to the original size of TPS2, the second set value to the original size of TPS3, and the third set value to the original size of TPS4 before the comparison and judgment module compares the differences among TPS1, TPS2, TPS3, and TPS4.
[0078] The first calculation module, the second calculation module and the third calculation module are respectively connected to the delay processing module.
[0079] The first calculation module is used to calculate the difference between TPS1 and TPS2 after adding the first set value, and then calculate the ratio of the difference to TPS1, which is recorded as ratio 1. The second calculation module is used to calculate the difference between TPS2 and TPS3 after adding the second set value, and then calculate the ratio of the difference to TPS2, which is recorded as ratio 2. The third calculation module calculates the difference between TPS3 and TPS4 after adding the third set value, and then calculates the ratio of the difference to TPS3, which is recorded as ratio 3.
[0080] The first calculation module, the second calculation module, and the third calculation module are connected to the judgment module respectively. The judgment module is used to judge whether the numerical values of ratio 1, ratio 2, and ratio 3 exceed 5%. If ratio 1 is greater than 5%, it means that the first data channel has a performance bottleneck. If ratio 2 exceeds 5%, it means that the upload link of the second data channel has a performance bottleneck. If ratio 3 is greater than 5%, it means that the download link of the second data channel has a performance bottleneck.
[0081] The judgment module is connected to the second data channel and the message file generation module respectively. The message file generation module is connected to the second data channel. The message file generation module is used to generate a message file corresponding to the performance bottleneck, and the message file records the number of autonomous driving events lost by the first data channel, the number of autonomous driving events lost by the second data channel when uploading, and the number of autonomous driving events lost by the second data channel when downloading. The message file is sent to the cloud platform through the upload link of the second data channel for positioning and alarming, assisting maintenance personnel to remotely locate and optimize performance bottlenecks.
[0082] Compared with the prior art, the beneficial effects of this embodiment are:
[0083] In the process of transmitting the autonomous driving events to the cloud platform, the key of this embodiment is to combine the delay processing module and the comparison and judgment module. Taking into account the errors that may be caused by the delay in the transmission process of the autonomous driving events, by comparing whether the difference ratios among TPS1, TPS2, TPS3, and TPS4 are within a reasonable range, the specific location of the performance bottleneck in the data channel can be automatically located, thereby achieving the function of automatic monitoring, which is convenient for maintenance personnel to remotely optimize the performance bottleneck.
[0084] The above embodiments are only preferred embodiments for fully illustrating the present invention, and the protection scope of the present invention is not limited thereto. Any equivalent substitution or change made by a person skilled in the art on the basis of the present invention is within the protection scope of the present invention.
Claims
1. A method for monitoring the performance of autonomous driving events on a cloud platform, executed on a system consisting of an autonomous driving controller, an onboard computer, and a cloud platform; a first data channel is provided between the autonomous driving controller and the onboard computer, and a second data channel is provided between the onboard computer and the cloud platform; characterized in that: Here are the steps: The autonomous driving controller counts the number of autonomous driving events processed per second sent through the first data channel, recorded as TPS1, and then sends TPS1 to the on-board computer through the first data channel; The onboard computer counts the autonomous driving event processing volume TPS2 received through the first data channel; The cloud platform counts the TPS3 of the autonomous driving events received through the second data channel, and then sends the TPS3 to the on-board computer through the second data channel; The onboard computer counts the autonomous driving event processing volume TPS4 received through the second data channel; The onboard computer compares the differences between TPS1, TPS2, TPS3, and TPS4, and then determines whether performance bottlenecks occur in the first data channel and the second data channel based on the comparison results; The specific process of comparing differences is: The onboard computer compares the differences between TPS1 and TPS2, TPS2 and TPS3, and TPS3 and TPS4 to see if they are within the set range; The specific process of determining whether a performance bottleneck occurs is as follows: If TPS1 is greater than TPS2 and the difference is outside the set range, it indicates that a performance bottleneck has occurred in the first data channel; If TPS2 is greater than TPS3 and the difference is outside the set range, it means that the upload performance of the second data channel has reached a bottleneck. If TPS3 is greater than TPS4 and the difference exceeds the set range, it means that a performance bottleneck has occurred in the download of the second data channel.
2. The method for monitoring the performance of an autonomous driving event cloud platform according to claim 1, characterized in that: When comparing TPS1 and TPS2, the setting range is that the difference between TPS1 and TPS2 accounts for less than 5% of the value of TPS1; When comparing TPS2 and TPS3, the setting range is such that the difference between TPS2 and TPS3 accounts for less than 5% of the value of TPS2; When comparing TPS3 and TPS4, the setting range is such that the difference between TPS3 and TPS4 accounts for less than 5 percent of the value of TPS3.
3. The method for monitoring the performance of an autonomous driving event on a cloud platform according to claim 2, characterized in that: After the on-board computer compares and finds the performance bottleneck, the on-board computer generates a message file corresponding to the performance bottleneck and sends it to the cloud platform through the second data channel for positioning and alarming.
4. The method for monitoring the performance of an autonomous driving event on a cloud platform according to any one of claims 1 to 3, characterized in that: Before the comparison between TPS1, TPS2, TPS3, and TPS4, a delay process is performed. The specific process is as follows: Before comparing TPS1 and TPS2, add the first set value to the original size of TPS2; Before comparing TPS2 and TPS3, add the second set value to the original size of TPS3; Before comparing TPS3 and TPS4, add the third set value to the original size of TPS4.
5. A performance monitoring system for uploading autonomous driving events to a cloud platform, comprising an autonomous driving controller, an onboard computer, and a cloud platform; a first data channel is provided between the autonomous driving controller and the onboard computer, and a second data channel is provided between the onboard computer and the cloud platform; and characterized in that: The automatic driving controller is provided with a first statistical module; the onboard computer is provided with a second statistical module and a comparison and judgment module; The first statistical module is used to count the autonomous driving event processing volume TPS1 sent by the autonomous driving controller through the first data channel; The second statistics module is used to count the autonomous driving event processing volume TPS2 received by the on-board computer through the first data channel, the autonomous driving event processing volume TPS4 received by the on-board computer through the second data channel, and the autonomous driving event processing volume TPS3 received by the cloud platform through the second data channel; The comparison and judgment module is used to compare the differences between TPS1, TPS2, TPS3, and TPS4, and determine whether performance bottlenecks occur in the first data channel and the second data channel based on the comparison results; The specific process of comparing the difference in the comparison and judgment module is as follows: Compare the differences between TPS1 and TPS2, TPS2 and TPS3, and TPS3 and TPS4 to see if they are within the set range; The comparison and judgment module determines whether a performance bottleneck occurs in the following process: If TPS1 is greater than TPS2 and the difference is outside the set range, it indicates that a performance bottleneck has occurred in the first data channel; If TPS2 is greater than TPS3 and the difference is outside the set range, it means that the upload performance of the second data channel has reached a bottleneck. If TPS3 is greater than TPS4 and the difference exceeds the set range, it means that a performance bottleneck has occurred in the download of the second data channel.
6. The performance monitoring system for the cloud platform for autonomous driving events according to claim 5, characterized in that: The comparison and judgment module includes a first calculation module, a second calculation module, a third calculation module, and a judgment module; The first calculation module is used to compare the ratio of the difference between TPS1 and TPS2 to the size of TPS1; The second calculation module is used to compare whether the ratio of the difference between TPS2 and TPS3 to the value of TPS1 exceeds a set range; The third calculation module is used to compare whether the ratio of the difference between TPS3 and TPS4 to the value of TPS1 exceeds a set range; The judgment module is used to judge whether the calculation results of the first comparison module, the second comparison module and the third comparison module exceed the set range respectively, so as to locate whether the first data channel and the second data channel have performance bottlenecks.
7. The performance monitoring system for the cloud platform for autonomous driving events according to claim 6, characterized in that: The onboard computer also has a delay processing module; The delay processing module is used to add a first set value to the original size of TPS2, a second set value to the original size of TPS3, and a third set value to the original size of TPS4 before the comparison and judgment module compares the sizes of TPS1, TPS2, TPS3, and TPS4.
8. The performance monitoring system for the cloud platform for autonomous driving events according to claim 7, characterized in that: The onboard computer also has a message file generation module; The message file generating module is used to generate a message file corresponding to the performance bottleneck.
