Scheduling method and device of nano-management algorithm, electronic equipment and storage medium

Abstract

The embodiment of the invention provides a scheduling method and device of a nano-management algorithm, electronic equipment and a storage medium, and the method comprises the steps: determining a labeled data set according to a mirror image algorithm; the mirror image algorithm corresponds to the nanotube algorithm; sending a starting request of the mirror image algorithm to the intelligent platform, and receiving a starting result sent by the intelligent platform according to the starting request; sending an algorithm service request to the intelligent platform according to the starting result and the labeled data set, and receiving an algorithm service result sent by the intelligent platform according to the algorithm service request; and carrying out statistics on the algorithm service result to obtain a statistical result, and issuing a mirror image algorithm according to the statistical result. Therefore, the management algorithm can be managed, operated and maintained through the algorithm enabling system and the intelligent platform.

Description

Technical Field

[0001] The present invention relates to the field of computer technology, and in particular to a scheduling method for a management algorithm, a scheduling device for a management algorithm, an electronic device and a computer-readable storage medium. Background Art

[0002] At present, there are many brands and versions of algorithms used in the actual network environment, lack of unified management, and the empowerment effect is not obvious. In addition, for the many manufacturers and multiple versions of similar algorithms that have been deployed, the characteristics of each algorithm cannot be fully utilized, resulting in problems of difficult management and operation and maintenance. Summary of the invention

[0003] In view of the above problems, embodiments of the present invention are proposed to provide a scheduling method for a managed algorithm, a scheduling device for a managed algorithm, an electronic device and a computer-readable storage medium that overcome the above problems or at least partially solve the above problems.

[0004] In order to solve the above problems, an embodiment of the present invention discloses a scheduling method for a management algorithm, which is applied to an algorithm enabling system and includes:

[0005] Determine the labeled data set according to the mirror algorithm; the mirror algorithm corresponds to the management algorithm;

[0006] Sending a start request of the mirror algorithm to the smart platform, and receiving a start result sent by the smart platform according to the start request;

[0007] Sending an algorithm service request to the intelligent platform according to the startup result and the annotated data set, and receiving an algorithm service result sent by the intelligent platform according to the algorithm service request;

[0008] The algorithm service result is counted to obtain a statistical result, and the mirror algorithm is published according to the statistical result.

[0009] Optionally, determining the labeled data set according to the mirror algorithm includes:

[0010] Get the selection request sent by the server;

[0011] Sending a query request to the intelligent platform according to the selection request;

[0012] Receiving the annotation information sent by the intelligent platform according to the query request;

[0013] A task creation request sent by the server is received, and a labeled data set is determined according to the task creation request and the labeling information.

[0014] Optionally, the receiving the task creation request sent by the server and determining the labeled data set according to the task creation request and the labeling information includes:

[0015] Sending the annotation information to the server;

[0016] receiving a material package sent by the server; the material package is determined by the server according to the marking information;

[0017] Verifying the material package;

[0018] When the material package passes the verification and a task creation request sent by the server is received, a data set pull request is sent to the intelligent platform;

[0019] The annotated data set is received from the intelligent platform according to the data set pulling request, and the annotated data set is stored.

[0020] Optionally, sending a start request of the mirror algorithm to the smart platform, and receiving a start result sent by the smart platform according to the start request, includes:

[0021] Receiving a start prediction request sent by the server, and sending a start request of the mirror algorithm to the intelligent platform according to the start prediction request;

[0022] Receiving a startup result sent by the smart platform according to the startup request;

[0023] According to the startup result, the state of the mirror algorithm is updated to started, and the startup result is sent to the server.

[0024] Optionally, sending an algorithm service request to the intelligent platform according to the startup result and the annotated data set, and receiving an algorithm service result sent by the intelligent platform according to the algorithm service request, includes:

[0025] Obtaining a startup completion notification sent by the smart platform according to the startup request; the startup completion notification is forwarded by the smart platform to the algorithm enabling system through the algorithm scheduling system;

[0026] According to the startup completion notification, updating the state of the mirror algorithm to running, and acquiring the annotated data set corresponding to the mirror algorithm;

[0027] Sending a test interface request to the gateway according to the annotated data set; so that the gateway sends an algorithm interface request to the algorithm scheduling system according to the test interface request, and the algorithm scheduling system sends an algorithm service request to the intelligent platform according to the algorithm interface request;

[0028] Receive the algorithm service result sent by the intelligent platform according to the algorithm service request; the algorithm service result is sent by the intelligent platform to the algorithm scheduling system, the algorithm scheduling system sends the algorithm service result to the gateway, and the gateway sends the algorithm service result to the algorithm empowerment system.

[0029] Optionally, the performing statistics on the algorithm service result to obtain a statistical result, and publishing the mirror algorithm according to the statistical result includes:

[0030] Performing statistics on the algorithm service results to obtain statistical results;

[0031] When the statistical result meets the preset result condition, the mirror algorithm is published, a publishing result is obtained, and the state of the mirror algorithm is updated to published;

[0032] The publishing result is sent to the server.

[0033] Optionally, before the step of collecting statistics on the algorithm service results to obtain statistical results and publishing the mirror algorithm according to the statistical results, the step further includes:

[0034] The algorithm service result is saved, and the state of the prediction result of the mirror algorithm is updated according to the algorithm service result;

[0035] A stop service request is sent to the intelligent platform, and a stop completion notification is received from the intelligent platform according to the stop service request; the stop completion notification is sent by the intelligent platform to the algorithm scheduling system, and the algorithm scheduling system sends the stop completion notification to the algorithm enabling system.

[0036] Optionally, before determining the labeled data set according to the mirror algorithm, the method further includes:

[0037] Get the image selection request and image version information sent by the service;

[0038] A mirror algorithm is determined according to the mirror selection request and the mirror version information, and the version information corresponding to the mirror algorithm is sent to the server, so that the server determines whether the mirror algorithm is correct according to the version information corresponding to the mirror algorithm and the mirror version information.

[0039] The embodiment of the present invention further discloses a scheduling device for a management algorithm, which is applied to an algorithm enabling system, and includes:

[0040] A first determination module is used to determine the labeled data set according to a mirror algorithm; the mirror algorithm corresponds to the management algorithm;

[0041] A first receiving module, used for sending a start request of the mirror algorithm to the smart platform, and receiving a start result sent by the smart platform according to the start request;

[0042] A second receiving module, used to send an algorithm service request to the intelligent platform according to the startup result and the annotated data set, and receive an algorithm service result sent by the intelligent platform according to the algorithm service request;

[0043] The publishing module is used to count the algorithm service results, obtain statistical results, and publish the mirror algorithm according to the statistical results.

[0044] Optionally, the first determining module includes:

[0045] A first acquisition submodule is used to acquire a selection request sent by a server;

[0046] A first sending submodule, used to send a query request to the intelligent platform according to the selection request;

[0047] A first receiving submodule, configured to receive the annotation information sent by the intelligent platform according to the query request;

[0048] The determination submodule is used to receive the task creation request sent by the server, and determine the labeled data set according to the task creation request and the labeling information.

[0049] Optionally, the first receiving submodule includes:

[0050] A first sending unit, configured to send the annotation information to the server;

[0051] A first receiving unit, configured to receive a material package sent by the server; the material package is determined by the server according to the marking information;

[0052] A verification unit, used for verifying the material package;

[0053] A second sending unit is used to send a request to pull a data set to the intelligent platform when the material package passes the verification and receives a create task request sent by the server;

[0054] The second receiving unit is used to receive the annotated data set sent by the intelligent platform according to the data set pulling request, and store the annotated data set.

[0055] Optionally, the first receiving module includes:

[0056] A second sending submodule is used to receive a start prediction request sent by the server, and send a start request of the mirror algorithm to the intelligent platform according to the start prediction request;

[0057] A second receiving submodule, configured to receive a startup result sent by the smart platform according to the startup request;

[0058] The third sending submodule is used to update the state of the mirror algorithm to started according to the startup result, and send the startup result to the server.

[0059] Optionally, the second receiving module includes:

[0060] A second acquisition submodule is used to acquire a startup completion notification sent by the smart platform according to the startup request; the startup completion notification is forwarded by the smart platform to the algorithm enabling system through the algorithm scheduling system;

[0061] A third acquisition submodule, configured to update the state of the mirror algorithm to running according to the startup completion notification, and acquire the annotated data set corresponding to the mirror algorithm;

[0062] A fourth sending submodule is used to send a test interface request to the gateway according to the annotated data set; so that the gateway sends an algorithm interface request to the algorithm scheduling system according to the test interface request, and the algorithm scheduling system sends an algorithm service request to the intelligent platform according to the algorithm interface request;

[0063] The third receiving submodule is used to receive the algorithm service result sent by the intelligent platform according to the algorithm service request; the algorithm service result is sent by the intelligent platform to the algorithm scheduling system, the algorithm scheduling system sends the algorithm service result to the gateway, and the gateway sends the algorithm service result to the algorithm empowerment system.

[0064] Optionally, the publishing module includes:

[0065] A statistical submodule, used to collect statistics on the algorithm service results to obtain statistical results;

[0066] A publishing submodule, used for publishing the mirror algorithm, obtaining a publishing result, and updating the status of the mirror algorithm to published when the statistical result meets a preset result condition;

[0067] The fifth sending submodule is used to send the publishing result to the server.

[0068] Optionally, the device further comprises:

[0069] An updating module, used to save the algorithm service result and update the state of the prediction result of the mirror algorithm according to the algorithm service result;

[0070] A sending module is used to send a stop service request to the intelligent platform and receive a stop completion notification sent by the intelligent platform according to the stop service request; the stop completion notification is sent by the intelligent platform to the algorithm scheduling system, and the algorithm scheduling system sends the stop completion notification to the algorithm enabling system.

[0071] Optionally, the device further comprises:

[0072] The acquisition module is used to obtain the image selection request and image version information sent by the service;

[0073] The second determination module is used to determine the mirror algorithm according to the mirror selection request and the mirror version information, and send the version information corresponding to the mirror algorithm to the server, so that the server determines whether the mirror algorithm is correct according to the version information corresponding to the mirror algorithm and the mirror version information.

[0074] The present invention also discloses a device, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor implements the steps of the scheduling method of the management algorithm as described above when executing the computer program.

[0075] The present invention also discloses a computer-readable storage medium, on which a computer program is stored. When the computer program is executed by a processor, the steps of the scheduling method of the management algorithm are implemented as described above.

[0076] The embodiments of the present invention include the following advantages:

[0077] In an embodiment of the present invention, the algorithm enabling system first determines a labeled data set based on a mirror algorithm, wherein the mirror algorithm competes with the management algorithm, and then sends a start request for the mirror algorithm to the intelligent platform, and receives a start result sent by the intelligent platform according to the start request, and then sends an algorithm service request to the intelligent platform according to the start result and the standard data set, and receives the algorithm service result sent by the intelligent platform according to the algorithm service request, and finally performs statistics on the algorithm service results to obtain statistical results, and publishes the mirror algorithm according to the statistical results, so that the mirror algorithm corresponding to the management algorithm is scheduled and the algorithm service is performed through the algorithm enabling system and the intelligent platform, and finally, the mirror algorithm is published according to the solution corresponding to the algorithm service, and then the management algorithm can be managed and operated through the algorithm enabling system and the intelligent platform. BRIEF DESCRIPTION OF THE DRAWINGS

[0078] Figure 1It is a flowchart of the steps of a scheduling method of a management algorithm provided by an embodiment of the present invention;

[0079] Figure 2 is a flowchart of steps for determining a labeled data set provided by an embodiment of the present invention;

[0080] Figure 3 is a flowchart of the steps of a mirror algorithm prediction provided by an embodiment of the present invention;

[0081] Figure 4 It is a structural block diagram of a concurrent task provided by an embodiment of the present invention;

[0082] Figure 5 It is a structural block diagram of a scheduling device for a management algorithm provided in an embodiment of the present invention. DETAILED DESCRIPTION

[0083] In order to make the above-mentioned objects, features and advantages of the present invention more obvious and easy to understand, the present invention is further described in detail below with reference to the accompanying drawings and specific embodiments.

[0084] At present, there are many algorithm brands and versions, lack of unified management, and the empowerment effect is not obvious. There are certain problems of difficulty in management and operation and maintenance for each algorithm. In order to solve the above technical problems, the present invention provides a scheduling method for the management algorithm. Its core concept is that the algorithm empowerment system first determines the labeled data set according to the mirror algorithm, wherein the mirror algorithm and the management algorithm play against each other, and then sends a start request for the mirror algorithm to the intelligent platform, and receives the start result sent by the intelligent platform according to the start request, and then sends an algorithm service request to the intelligent platform according to the start result and the standard data set, and receives the algorithm service result sent by the intelligent platform according to the algorithm service request, and finally counts the algorithm service results to obtain the statistical results, and publishes the mirror algorithm according to the statistical results, so that the mirror algorithm corresponding to the management algorithm is scheduled and the algorithm service is performed through the algorithm empowerment system and the intelligent platform, and finally the mirror algorithm is released according to the corresponding solution of the algorithm service, and then the management algorithm can be managed and maintained through the algorithm empowerment system and the intelligent platform.

[0085] Reference Figure 1 , shows a flowchart of a scheduling method for a management algorithm provided by an embodiment of the present invention, the method is applied to an algorithm enabling system, and the method may specifically include the following steps:

[0086] Step 101, determine the labeled data set according to the mirror algorithm; the mirror algorithm corresponds to the management algorithm.

[0087] In the embodiment of the present invention, when training or scheduling the management algorithm, it is necessary to first determine the labeled data set according to the mirror algorithm corresponding to the management algorithm.

[0088] In one embodiment, determining a labeled data set according to a mirror algorithm may include: obtaining a selection request sent by a server; sending a query request to an intelligent platform according to the selection request; receiving labeling information sent by the intelligent platform according to the query request; receiving a create task request sent by the server, and determining a labeled data set according to the create task request and the labeling information.

[0089] Specifically, the server can obtain the user's selection request, which is a selection of a data set corresponding to the managed algorithm. The selection request may include the name of the data set. The server then sends the selection request to the algorithm empowerment system. After the algorithm empowerment system obtains the selection request, it can send a query request to the intelligent platform according to the name corresponding to the data set in the selection request. The query request includes the data set. The intelligent platform can determine the annotation information of the annotated data set according to the name corresponding to the data set in the query request, and send the annotation information to the algorithm empowerment system. When the algorithm empowerment system receives the task creation request sent by the server, it can determine the annotated data set according to the task creation request and the annotation information.

[0090] In the present invention, the intelligent platform has stored the labeled information for the data set, and the labeled information may include the algorithm prediction box information and confidence level, etc. The labeled information may be labeled by the intelligent platform and the labeled results may be manually checked.

[0091] In one embodiment, receiving a task creation request sent by a server and determining a labeled data set based on the task creation request and labeling information may include: sending the labeling information to the server; receiving a material package sent by the server; the material package is determined by the server based on the labeling information; verifying the material package; when the material package verification passes and the task creation request sent by the server is received, sending a data set pull request to the intelligent platform; receiving the labeled data set sent by the intelligent platform based on the data set pull request, and storing the labeled data set.

[0092] Specifically, after receiving the annotation information sent by the intelligent platform, the algorithm empowerment system can send the annotation information to the server. After receiving the annotation information, the server can send the compressed material package to the algorithm empowerment system. The algorithm empowerment system can verify the material package to verify whether the content and format in the material package meet the preset requirements. When the material package passes the verification, the verification result can be sent to the server. When the verification fails, the material package can be obtained from the server again and verified. After the server receives the verification pass information, it can send a create task request to the algorithm empowerment system. After receiving the create task request, the algorithm empowerment system can send a data set pull request to the intelligent platform. The data set pull request can include annotation information. The intelligent platform can determine the labeled data set based on the annotation information in the data set pull request, and send the labeled data set to the algorithm empowerment system. The algorithm empowerment system can store the labeled data set in the database, or asynchronously transfer the labeled data set to the distributed storage system.

[0093] In an embodiment of the present invention, before determining the labeled data set according to the mirror algorithm, the method may also include: obtaining a mirror selection request and mirror version information sent by the service; determining the mirror algorithm according to the mirror selection request and the mirror version information, and sending the version information corresponding to the mirror algorithm to the server, so that the server determines whether the mirror algorithm is correct according to the version information corresponding to the mirror algorithm and the mirror version information.

[0094] Specifically, the server can first send an image selection request and image version information to the algorithm empowerment system. The algorithm empowerment system determines the corresponding image algorithm based on the image selection request and the image version information, and sends the version information corresponding to the image algorithm to the server. The server can determine whether the image algorithm determined by the algorithm empowerment system is correct based on the version information corresponding to the image algorithm and the image version information. If the determined image algorithm is incorrect, the algorithm empowerment system can determine the image algorithm again.

[0095] like Figure 2As shown, a flowchart of the steps of determining a labeled data set provided by an embodiment of the present invention is shown. The server first sends an image selection request and image version information to the algorithm empowerment system. The algorithm empowerment system determines the image algorithm according to the image selection request and the image version information, and sends the version information of the image algorithm to the server. The server then sends a selection request to the algorithm empowerment system. The algorithm empowerment system sends a query request to the intelligent platform according to the selection request, wherein the query request includes the name corresponding to the data set. The intelligent platform determines the labeling information of the labeled data set according to the name corresponding to the data set, and sends the labeling information to the algorithm empowerment system. The algorithm empowerment system sends the labeling information to the server. The server sends a material package to the algorithm empowerment system. The algorithm empowerment system verifies the material package and sends the verification result to the server. After the verification is passed, the server sends a create task request to the algorithm empowerment system. The algorithm empowerment system sends a request to pull a data set to the intelligent platform. The intelligent platform determines the labeled data set according to the pull data set request, and sends the labeled data set to the algorithm empowerment system. The algorithm empowerment system stores the labeled data set in the database and sends the creation result corresponding to the task creation request to the server.

[0096] Step 102: Send a start request for the mirror algorithm to the smart platform, and receive a start result sent by the smart platform according to the start request.

[0097] In an embodiment of the present invention, after determining the labeled data set, the algorithm enabling system can send a start request for the mirror algorithm to the intelligent platform. The intelligent platform can generate a start result based on the mirror algorithm start request and send the start result to the algorithm enabling system.

[0098] In one embodiment, a start request for a mirroring algorithm is sent to an intelligent platform, and a start result sent by the intelligent platform according to the start request is received, including: receiving a start prediction request sent by a server, and sending a start request for the mirroring algorithm to the intelligent platform according to the start prediction request; receiving a start result sent by the intelligent platform according to the start request; and updating the status of the mirroring algorithm to started according to the start result, and sending the start result to the server.

[0099] Specifically, the server can send a car prediction request to the algorithm empowerment system, and the algorithm empowerment system can send a start request for the mirror algorithm to the intelligent platform based on the start prediction request, wherein the start request for the mirror algorithm can include the mirror algorithm. After receiving the start request, the intelligent platform can start the mirror algorithm and send the start result of the mirror algorithm to the algorithm empowerment system. The algorithm empowerment system can update the status of the mirror algorithm to started according to the start result and send the start result to the server.

[0100] Step 103: Send an algorithm service request to the intelligent platform according to the startup result and the annotated data set, and receive an algorithm service result sent by the intelligent platform according to the algorithm service request.

[0101] In an embodiment of the present invention, after the algorithm empowerment system receives the startup result sent by the intelligent platform, it can send an algorithm service request to the intelligent platform based on the startup result and the annotated data set. After the intelligent platform receives the algorithm service request, it can generate an algorithm server result based on the algorithm service request and send the algorithm service result to the algorithm empowerment system.

[0102] In the present invention, when an algorithm service request is sent to the intelligent platform according to the startup result and the labeled data set, an algorithm service request can also be sent to the intelligent platform according to the startup result, the labeled data set and the material package.

[0103] In one embodiment, an algorithm service request is sent to an intelligent platform according to the startup result and the annotated data set, and an algorithm service result sent by the intelligent platform according to the algorithm service request is received, including: obtaining a startup completion notification sent by the intelligent platform according to the startup request; the startup completion notification is forwarded by the intelligent platform to the algorithm empowerment system through the algorithm scheduling system; according to the startup completion notification, the status of the mirror algorithm is updated to running, and the annotated data set corresponding to the mirror algorithm is obtained; according to the annotated data set, a test interface request is sent to the gateway; so that the gateway sends an algorithm interface request to the algorithm scheduling system according to the test interface request, and the algorithm scheduling system sends an algorithm service request to the intelligent platform according to the algorithm interface request; receiving the algorithm service result sent by the intelligent platform according to the algorithm service request; the algorithm service result is sent by the intelligent platform to the algorithm scheduling system, the algorithm scheduling system sends the algorithm service result to the gateway, and the gateway sends the algorithm service result to the algorithm empowerment system.

[0104] Specifically, after the intelligent platform completes the startup of the mirror algorithm, the intelligent platform can send a startup completion notification to the scheduling algorithm system, and the scheduling algorithm system then sends the startup completion notification to the algorithm empowerment system. After receiving the startup completion notification, the algorithm empowerment system can update the status of the mirror algorithm to running, and then obtain the annotated data set corresponding to the mirror algorithm, which may include image information, or obtain the annotated data set while also obtaining the material package, which may also include image information. The algorithm empowerment system then sends a test interface request to the gateway, which may include an annotated data set or a material package. After the gateway receives the test interface request, it can send an algorithm interface request to the algorithm scheduling system, which may include an annotated data set or a material package. After the algorithm scheduling system receives the algorithm interface request, it can send an algorithm service request to the intelligent platform, which may include an annotated data set or a material package. Then the intelligent platform generates the corresponding algorithm service result based on the algorithm service request and the mirror algorithm, and sends the algorithm server result to the algorithm scheduling system. The algorithm scheduling system sends the algorithm service result to the gateway, and the gateway finally sends the algorithm service result to the algorithm empowerment system.

[0105] Step 104, statistics are collected on the algorithm service results to obtain statistical results, and the mirror algorithm is published according to the statistical results.

[0106] In an embodiment of the present invention, after receiving the algorithm service results, the algorithm enabling system can count the algorithm service results to obtain statistical results, and then publish the mirror algorithm based on the statistical results.

[0107] In an embodiment of the present invention, statistics are collected on the algorithm service results to obtain statistical results, and before the mirror algorithm is released based on the statistical results, the following may also be included: saving the algorithm service results, and updating the status of the prediction results of the mirror algorithm based on the algorithm service results; sending a stop service request to the intelligent platform, and receiving a stop completion notification sent by the intelligent platform based on the stop service request; the stop completion notification is sent by the intelligent platform to the algorithm scheduling system, and the algorithm scheduling system sends the stop completion notification to the algorithm enabling system.

[0108] Specifically, after receiving the algorithm service results, the algorithm empowerment system can save the algorithm service results, and based on the algorithm service results, update the status of the prediction results of the mirror algorithm, and then send a stop service request to the intelligent platform. After receiving the stop service request, the intelligent platform can stop the prediction of the mirror algorithm, and generate a stop completion notification, and send the stop completion notification to the algorithm scheduling system, and the algorithm scheduling system will then send the stop completion notification to the algorithm empowerment system.

[0109] In one embodiment, statistics are performed on the algorithm service results to obtain statistical results, and a mirror algorithm is published based on the statistical results, including: statistics are performed on the algorithm service results to obtain statistical results; when the statistical results meet preset result conditions, the mirror algorithm is published to obtain a publishing result, and the status of the mirror algorithm is updated to published; and the publishing result is sent to a server.

[0110] In order to better understand the present invention, Figure 3 The present invention will be described. Figure 3 The invention shows a flowchart of the steps of a mirror algorithm prediction provided by an embodiment of the present invention. The server sends a start prediction request to the algorithm enabling system, the algorithm enabling system sends a start request of the mirror algorithm to the intelligent platform, the intelligent platform responds to the start request, generates a start result, sends the start result to the algorithm enabling system, the algorithm enabling system updates the state of the mirror algorithm according to the start result, and sends the start result to the server. After the intelligent platform completes the mirror algorithm startup, a start completion notification can be generated, and the start completion notification can be sent to the algorithm scheduling system, the algorithm scheduling system then sends the start completion notification to the algorithm enabling system, after the algorithm enabling system receives the start completion notification, the state of the mirror algorithm can be updated to be in operation, and then the annotated data set is obtained, and according to the annotated data set, a test interface request is sent to the gateway, the gateway sends an algorithm interface request to the algorithm scheduling system according to the test interface request, the algorithm scheduling system sends an algorithm service request to the intelligent platform according to the algorithm interface request, the intelligent platform responds to the algorithm service request to generate an algorithm service result, and sends the algorithm service result to the algorithm scheduling system, the algorithm scheduling system then sends the algorithm service result to the gateway, the gateway sends the algorithm service result to the algorithm enabling system, the algorithm enabling system saves the algorithm service result, and updates the progress of the mirror algorithm. The algorithm empowerment system updates the status of the prediction results according to the algorithm service results, and then sends a stop service request to the intelligent platform. The intelligent platform responds to the stop service request, generates a stop result, and then sends the stop result to the algorithm empowerment system. After the intelligent platform stops the service, it can generate a stop completion notification, and then send the stop completion notification to the algorithm scheduling system. The algorithm scheduling system sends the stop completion notification to the algorithm empowerment system, and the algorithm empowerment system updates the status of the mirror algorithm to stop service. The server sends a mirror publishing request to the algorithm empowerment system, and the algorithm empowerment system updates the status of the mirror algorithm to published and sends the publishing result to the server.

[0111] In the present invention, based on the scheduling steps of the mirror algorithm of the present invention, the prediction tasks can be isolated, and the algorithm scheduling system regularly schedules the workflow, and when a certain work item is hit, a new instance is initiated to concurrently implement the task. Figure 4As shown, it is a structural block diagram of the first concurrent task provided by an embodiment of the present invention. The present invention can trigger task scheduling for multiple prediction types and multiple process conversions. The prediction type can be accuracy prediction, interface prediction, stability prediction and performance prediction. At the same time, when managing task scheduling, scheduling can be performed based on scheduling type, scheduling time, work definition and scheduling result. The scheduling process of the mirror algorithm of the present invention includes data set acquisition, algorithm service startup, execution algorithm prediction, algorithm service shutdown and algorithm service result statistics. Data set acquisition, algorithm service startup, execution algorithm prediction, algorithm service shutdown and algorithm service result statistics all include workflow implementation, successful flow and failed retry. Workflow implementation, successful flow and failed retry all correspond to related consumer threads.

[0112] In an embodiment of the present invention, according to the mirror algorithm, the labeled data set is determined; the mirror algorithm corresponds to the management algorithm; a start request of the mirror algorithm is sent to the intelligent platform, and the start result sent by the intelligent platform according to the start request is received; according to the start result and the labeled data set, an algorithm service request is sent to the intelligent platform, and the algorithm service result sent by the intelligent platform according to the algorithm service request is received; the algorithm service result is counted to obtain the statistical result, and the mirror algorithm is released according to the statistical result. Thus, through the algorithm empowerment system and the intelligent platform, the mirror algorithm corresponding to the management algorithm is scheduled and the algorithm service is performed, and finally according to the solution corresponding to the algorithm service, the mirror algorithm is released, and then the management algorithm can be managed and operated through the algorithm empowerment system and the intelligent platform.

[0113] It should be noted that, for the sake of simplicity, the method embodiments are described as a series of action combinations, but those skilled in the art should be aware that the embodiments of the present invention are not limited by the order of the actions described, because according to the embodiments of the present invention, certain steps can be performed in other orders or simultaneously. Secondly, those skilled in the art should also be aware that the embodiments described in the specification are all preferred embodiments, and the actions involved are not necessarily required by the embodiments of the present invention.

[0114] Reference Figure 5 , shows a structural block diagram of a scheduling device for a management algorithm provided by an embodiment of the present invention, the device is applied to an algorithm enabling system, and specifically may include the following modules:

[0115] A first determination module 501 is used to determine the labeled data set according to a mirror algorithm; the mirror algorithm corresponds to the management algorithm;

[0116] The first receiving module 502 is used to send a start request of the mirror algorithm to the smart platform, and receive a start result sent by the smart platform according to the start request;

[0117] The second receiving module 503 is used to send an algorithm service request to the intelligent platform according to the startup result and the annotated data set, and receive an algorithm service result sent by the intelligent platform according to the algorithm service request;

[0118] The publishing module 504 is used to collect statistics on the algorithm service results, obtain statistical results, and publish the mirror algorithm according to the statistical results.

[0119] Optionally, the first determining module 501 includes:

[0120] A first acquisition submodule is used to acquire a selection request sent by a server;

[0121] A first sending submodule, used to send a query request to the intelligent platform according to the selection request;

[0122] A first receiving submodule, configured to receive the annotation information sent by the intelligent platform according to the query request;

[0123] The determination submodule is used to receive the task creation request sent by the server, and determine the labeled data set according to the task creation request and the labeling information.

[0124] Optionally, the first receiving submodule includes:

[0125] A first sending unit, configured to send the annotation information to the server;

[0126] A first receiving unit, configured to receive a material package sent by the server; the material package is determined by the server according to the marking information;

[0127] A verification unit, used for verifying the material package;

[0128] A second sending unit is used to send a request to pull a data set to the intelligent platform when the material package passes the verification and receives a create task request sent by the server;

[0129] The second receiving unit is used to receive the annotated data set sent by the intelligent platform according to the data set pulling request, and store the annotated data set.

[0130] Optionally, the first receiving module 502 includes:

[0131] A second sending submodule is used to receive a start prediction request sent by the server, and send a start request of the mirror algorithm to the intelligent platform according to the start prediction request;

[0132] A second receiving submodule, configured to receive a startup result sent by the smart platform according to the startup request;

[0133] The third sending submodule is used to update the state of the mirror algorithm to started according to the startup result, and send the startup result to the server.

[0134] Optionally, the second receiving module 503 includes:

[0135] A second acquisition submodule is used to acquire a startup completion notification sent by the smart platform according to the startup request; the startup completion notification is forwarded by the smart platform to the algorithm enabling system through the algorithm scheduling system;

[0136] A third acquisition submodule, configured to update the state of the mirror algorithm to running according to the startup completion notification, and acquire the annotated data set corresponding to the mirror algorithm;

[0137] A fourth sending submodule is used to send a test interface request to the gateway according to the annotated data set; so that the gateway sends an algorithm interface request to the algorithm scheduling system according to the test interface request, and the algorithm scheduling system sends an algorithm service request to the intelligent platform according to the algorithm interface request;

[0138] The third receiving submodule is used to receive the algorithm service result sent by the intelligent platform according to the algorithm service request; the algorithm service result is sent by the intelligent platform to the algorithm scheduling system, the algorithm scheduling system sends the algorithm service result to the gateway, and the gateway sends the algorithm service result to the algorithm empowerment system.

[0139] Optionally, the publishing module 504 includes:

[0140] A statistical submodule, used to collect statistics on the algorithm service results to obtain statistical results;

[0141] A publishing submodule, used for publishing the mirror algorithm, obtaining a publishing result, and updating the status of the mirror algorithm to published when the statistical result meets a preset result condition;

[0142] The fifth sending submodule is used to send the publishing result to the server.

[0143] Optionally, the device further comprises:

[0144] An updating module, used to save the algorithm service result and update the state of the prediction result of the mirror algorithm according to the algorithm service result;

[0145] A sending module is used to send a stop service request to the intelligent platform and receive a stop completion notification sent by the intelligent platform according to the stop service request; the stop completion notification is sent by the intelligent platform to the algorithm scheduling system, and the algorithm scheduling system sends the stop completion notification to the algorithm enabling system.

[0146] Optionally, the device further comprises:

[0147] The acquisition module is used to obtain the image selection request and image version information sent by the service;

[0148] The second determination module is used to determine the mirror algorithm according to the mirror selection request and the mirror version information, and send the version information corresponding to the mirror algorithm to the server, so that the server determines whether the mirror algorithm is correct according to the version information corresponding to the mirror algorithm and the mirror version information.

[0149] In the embodiment of the present invention, the first determination module is used to determine the labeled data set according to the mirror algorithm; the mirror algorithm corresponds to the management algorithm; the first receiving module is used to send a start request of the mirror algorithm to the intelligent platform, and receive the start result sent by the intelligent platform according to the start request; the second receiving module is used to send an algorithm service request to the intelligent platform according to the start result and the labeled data set, and receive the algorithm service result sent by the intelligent platform according to the algorithm service request; the publishing module is used to count the algorithm service results, obtain the statistical results, and publish the mirror algorithm according to the statistical results. Thus, through the algorithm empowerment system and the intelligent platform, the mirror algorithm corresponding to the management algorithm is scheduled and the algorithm service is performed, and finally according to the solution corresponding to the algorithm service, the mirror algorithm is published, and then the management algorithm can be managed and operated through the algorithm empowerment system and the intelligent platform.

[0150] As for the device embodiment, since it is basically similar to the method embodiment, the description is relatively simple, and the relevant parts can be referred to the partial description of the method embodiment.

[0151] An embodiment of the present invention further provides an electronic device, including:

[0152] It includes a processor, a memory, and a computer program stored in the memory and capable of running on the processor. When the computer program is executed by the processor, the various processes of the scheduling method embodiment of the above-mentioned management algorithm are implemented, and the same technical effect can be achieved. To avoid repetition, it will not be repeated here.

[0153] An embodiment of the present invention also provides a computer-readable storage medium, on which a computer program is stored. When the computer program is executed by a processor, the various processes of the scheduling method embodiment of the above-mentioned management algorithm are implemented, and the same technical effect can be achieved. To avoid repetition, it will not be repeated here.

[0154] The various embodiments in this specification are described in a progressive manner, and each embodiment focuses on the differences from other embodiments. The same or similar parts between the various embodiments can be referenced to each other.

[0155] It will be appreciated by those skilled in the art that the embodiments of the present invention may be provided as methods, devices, or computer program products. Therefore, the embodiments of the present invention may take the form of a complete hardware embodiment, a complete software embodiment, or an embodiment combining software and hardware. Moreover, the embodiments of the present invention may take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program codes.

[0156] The embodiments of the present invention are described with reference to the flowcharts and / or block diagrams of the methods, terminal devices (systems), and computer program products according to the embodiments of the present invention. It should be understood that each process and / or block in the flowchart and / or block diagram, as well as the combination of the processes and / or blocks in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing terminal device to generate a machine, so that the instructions executed by the processor of the computer or other programmable data processing terminal device generate instructions for implementing the processes in the flowchart and / or block diagram. Figure 1 A process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.

[0157] These computer program instructions may also be stored in a computer-readable memory capable of directing a computer or other programmable data processing terminal device to operate in a specific manner, so that the instructions stored in the computer-readable memory produce a manufactured product including an instruction device, which implements the process Figure 1 A process or multiple processes and / or boxes Figure 1 A function specified in one or more boxes.

[0158] These computer program instructions can also be loaded onto a computer or other programmable data processing terminal device so that a series of operating steps are executed on the computer or other programmable terminal device to produce a computer-implemented process, thereby providing instructions for executing on the computer or other programmable terminal device to implement the process. Figure 1A process or multiple processes and / or boxes Figure 1 The steps for the functions specified in one or more boxes.

[0159] Although the preferred embodiments of the present invention have been described, those skilled in the art may make additional changes and modifications to these embodiments once they have learned the basic creative concept. Therefore, the appended claims are intended to be interpreted as including the preferred embodiments and all changes and modifications that fall within the scope of the embodiments of the present invention.

[0160] Finally, it should be noted that, in this article, relational terms such as first and second, etc. are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Moreover, the terms "include", "comprise" or any other variants thereof are intended to cover non-exclusive inclusion, so that a process, method, article or terminal device including a series of elements includes not only those elements, but also other elements not explicitly listed, or also includes elements inherent to such process, method, article or terminal device. In the absence of further restrictions, the elements defined by the sentence "comprise a ..." do not exclude the existence of other identical elements in the process, method, article or terminal device including the elements.

[0161] The above is a detailed introduction to a scheduling method for a management algorithm, a scheduling device for a management algorithm, an electronic device and a computer-readable storage medium provided by the present invention. Specific examples are used in this article to illustrate the principles and implementation methods of the present invention. The description of the above embodiments is only used to help understand the method of the present invention and its core idea; at the same time, for general technical personnel in this field, according to the idea of ​​the present invention, there will be changes in the specific implementation method and application scope. In summary, the content of this specification should not be understood as a limitation on the present invention.

Claims

1. A scheduling method for a management algorithm, characterized in that: The method is applied to an algorithm enabling system, and the method includes: Determine the labeled data set according to the mirror algorithm; the mirror algorithm corresponds to the management algorithm; Sending a start request of the mirror algorithm to the smart platform, and receiving a start result sent by the smart platform according to the start request; Sending an algorithm service request to the intelligent platform according to the startup result and the annotated data set, and receiving an algorithm service result sent by the intelligent platform according to the algorithm service request; The algorithm service result is counted to obtain a statistical result, and the mirror algorithm is published according to the statistical result.

2. The scheduling method of the management algorithm according to claim 1, characterized in that: Determining the labeled data set according to the mirror algorithm includes: Get the selection request sent by the server; Sending a query request to the intelligent platform according to the selection request; Receiving the annotation information sent by the intelligent platform according to the query request; A task creation request sent by the server is received, and a labeled data set is determined according to the task creation request and the labeling information.

3. The scheduling method of the management algorithm according to claim 2, characterized in that: The receiving the task creation request sent by the server and determining the labeled data set according to the task creation request and the labeling information includes: Sending the annotation information to the server; receiving a material package sent by the server; the material package is determined by the server according to the marking information; Verifying the material package; When the material package passes the verification and a task creation request sent by the server is received, a data set pull request is sent to the intelligent platform; The annotated data set is received from the intelligent platform according to the data set pulling request, and the annotated data set is stored.

4. The scheduling method of the management algorithm according to claim 1, characterized in that: The sending a start request of the mirror algorithm to the smart platform, and receiving a start result sent by the smart platform according to the start request, includes: Receiving a start prediction request sent by the server, and sending a start request of the mirror algorithm to the intelligent platform according to the start prediction request; Receiving a startup result sent by the smart platform according to the startup request; According to the startup result, the state of the mirror algorithm is updated to started, and the startup result is sent to the server.

5. The scheduling method of the management algorithm according to claim 1, characterized in that: The sending of an algorithm service request to the intelligent platform according to the startup result and the annotated data set, and receiving an algorithm service result sent by the intelligent platform according to the algorithm service request, includes: Obtaining a startup completion notification sent by the smart platform according to the startup request; the startup completion notification is forwarded by the smart platform to the algorithm enabling system through the algorithm scheduling system; According to the startup completion notification, updating the state of the mirror algorithm to running, and acquiring the annotated data set corresponding to the mirror algorithm; According to the annotated data set, a test interface request is sent to the gateway; so that the gateway sends an algorithm interface request to the algorithm scheduling system according to the test interface request, and the algorithm scheduling system sends an algorithm service request to the intelligent platform according to the algorithm interface request; Receive the algorithm service result sent by the intelligent platform according to the algorithm service request; the algorithm service result is sent by the intelligent platform to the algorithm scheduling system, the algorithm scheduling system sends the algorithm service result to the gateway, and the gateway sends the algorithm service result to the algorithm empowerment system.

6. The scheduling method of the management algorithm according to claim 1, characterized in that: The method of performing statistics on the algorithm service results to obtain statistical results, and publishing the mirror algorithm according to the statistical results, includes: Performing statistics on the algorithm service results to obtain statistical results; When the statistical result meets the preset result condition, the mirror algorithm is published, a publishing result is obtained, and the state of the mirror algorithm is updated to published; The publishing result is sent to the server.

7. The scheduling method of the management algorithm according to claim 1, characterized in that: Before the algorithm service result is counted to obtain the statistical result and the mirror algorithm is released according to the statistical result, the method further includes: The algorithm service result is saved, and the state of the prediction result of the mirror algorithm is updated according to the algorithm service result; A stop service request is sent to the intelligent platform, and a stop completion notification is received from the intelligent platform according to the stop service request; the stop completion notification is sent by the intelligent platform to the algorithm scheduling system, and the algorithm scheduling system sends the stop completion notification to the algorithm enabling system.

8. The scheduling method of the management algorithm according to claim 1, characterized in that: Before determining the labeled data set according to the mirror algorithm, the method further includes: Get the image selection request and image version information sent by the service; A mirror algorithm is determined according to the mirror selection request and the mirror version information, and the version information corresponding to the mirror algorithm is sent to the server, so that the server determines whether the mirror algorithm is correct according to the version information corresponding to the mirror algorithm and the mirror version information.

9. A scheduling device for a management algorithm, characterized in that: The device is applied to an algorithm enabling system, and the device includes: A determination module, used to determine the labeled data set according to a mirror algorithm; the mirror algorithm corresponds to the management algorithm; A first receiving module, used for sending a start request of the mirror algorithm to the smart platform, and receiving a start result sent by the smart platform according to the start request; A second receiving module, used to send an algorithm service request to the intelligent platform according to the startup result and the annotated data set, and receive an algorithm service result sent by the intelligent platform according to the algorithm service request; The publishing module is used to count the algorithm service results, obtain statistical results, and publish the mirror algorithm according to the statistical results.

10. An electronic device, characterized in that: include: A processor, a memory, and a computer program stored in the memory and capable of running on the processor, wherein when the computer program is executed by the processor, the steps of the scheduling method of the management algorithm as described in any one of claims 1 to 8 are implemented.

11. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the steps of the scheduling method of the management algorithm described in any one of claims 1 to 8 are implemented.

Images