CN116844111A

CN116844111A - Method, device, platform and storage medium for distributed control management

Info

Publication number: CN116844111A
Application number: CN202310797595.7A
Authority: CN
Inventors: 王斌克; 杨旭; 张冬; 沈伟平; 白文翔; 彭永飞; 徐井卫; 周祥; 贾云东; 朱争芳
Original assignee: Suzhou Keda Technology Co Ltd
Current assignee: Suzhou Keyuan Software Technology Development Co ltd; Suzhou Keda Technology Co Ltd
Priority date: 2023-06-30
Filing date: 2023-06-30
Publication date: 2023-10-03

Abstract

The invention relates to the technical field of computers, and discloses a method, a device, a platform and a storage medium for distributed control management, wherein the method comprises the following steps: acquiring a target control task and task information thereof, wherein the task information comprises a target control area and a target control object; determining a target control system from the accessed control systems based on the target control area; if the target control area is larger than a preset range, acquiring a large library control task corresponding to the target control task in the target control system, wherein the large library control task is used for controlling a plurality of control objects in the same target control area; writing the target control object into the large library control task to control the target control object to obtain a control result. The problem that when a control task is set for a large control area, the control flow is complex and the control efficiency is low can be solved.

Description

Method, device, platform and storage medium for distributed control management

Technical Field

The invention relates to the technical field of computers, in particular to a method, a device, a platform and a storage medium for management and control.

Background

With the increasing perfection of video monitoring construction, face images can be acquired in corresponding control areas through video monitoring equipment, and then the face images are sent to a background of a control system for recognition and comparison in a network transmission mode, so that control objects can be accurately recognized, and a good monitoring effect is achieved. When a person responsible for controlling is controlling the face of the controlled object, a specific control task needs to be filled in the control management platform, for example: information such as a control name, control behavior, control start time, control end time, control area and the like. However, for the task of large-scale control area, because the area of the control area is large, the information to be filled is large, which results in complex control flow and low control efficiency.

Disclosure of Invention

In view of the above, the present invention provides a method, an apparatus, a platform and a storage medium for controlling and managing, so as to solve the problems of complex control flow and low control efficiency when a control task is set for a large control area.

In a first aspect, the present invention provides a method for controlling and managing, the method including:

acquiring a target control task and task information thereof, wherein the task information comprises a target control area and a target control object;

Determining a target control system from the accessed control systems based on the target control area;

if the target control area is larger than a preset range, acquiring a large library control task corresponding to the target control task in the target control system, wherein the large library control task is used for controlling a plurality of control objects in the same target control area;

writing the target control object into the large library control task to control the target control object to obtain a control result.

In the mode, the large library control task is deployed in the target control system, so that when the target control area of the obtained target control task is larger than a preset range, the large library control task corresponding to the target control task can be obtained, and the target control object is directly written into the large library control task to control the target control object. Therefore, when the range of the control area is larger, the task information filling flow and the control flow can be simplified, so that the control efficiency is improved.

In an alternative embodiment, the method further comprises:

and if the target control area is smaller than or equal to the preset range, issuing the target control task and the target control object thereof to the target control system so as to control the target control object to obtain a control result.

In this way, when the target control area is smaller than or equal to the preset range, the target control task and the target control object thereof are directly issued to the target control system, so that the control efficiency of the target control task in a small range can be ensured.

In an alternative embodiment, the method further comprises:

when receiving a target control result fed back by any accessed control system, acquiring a target task code corresponding to the target control result;

when the target task code is the task code of the large library control task, acquiring a control object corresponding to the target control result;

according to the control object corresponding to the target control result, determining a control task corresponding to the target control result and a task state of the control task from prestored control tasks corresponding to the large library control task;

and when the task state meets a preset effective control condition, sending the target control result to a corresponding service system according to the control task corresponding to the target control result.

In the mode, when the received target control result is a control result fed back by the large library control task, firstly, according to a control object corresponding to the target control result, inquiring the corresponding control task and the task state thereof, and when the task state meets the preset effective control condition, sending the target control result to the corresponding service system. Therefore, the invalid control result can be prevented from being fed back to the corresponding service system, so that unnecessary occupied storage resources and data transmission resources are reduced, and the control efficiency is improved.

In an alternative embodiment, the method further comprises:

and when the target task code is not the task code of the large library control task, determining a control task corresponding to the target control result according to the target task code so as to send the target control result to a corresponding service system.

In the mode, when the received target control result is not the control result of the large-library control task, the target control result is directly sent to the corresponding service system according to the target task code, so that the transmission efficiency of the control result of the small-range control task can be ensured.

In an alternative embodiment, the method further comprises:

acquiring a preset system assessment task, wherein the system assessment task comprises at least one assessment subtask;

the accessed distribution control systems are checked based on the checking subtasks to obtain task checking results of the accessed distribution control systems corresponding to the checking subtasks;

corresponding to each accessed distributed control system, fusing task assessment results of all assessment subtasks to obtain system assessment results of the accessed distributed control systems;

and when the system check result is that the distributed control system does not reach the standard, generating corresponding system alarm information.

In the mode, each accessed control system is checked through a preset system check task, and corresponding system alarm information is generated when the system check result of the control system is that the control system does not reach the standard, so that related personnel can check the control system in time, effective execution of target task object control is ensured, and the control efficiency and control quality are improved.

In an optional implementation manner, the checking the accessed control system based on the checking subtasks to obtain task checking results of each accessed control system corresponding to each checking subtask includes:

analyzing the assessment subtasks to determine task targets;

acquiring a processing result sample and a target processing result corresponding to each accessed control system based on the task target;

and obtaining task assessment results of each accessed control system corresponding to the assessment subtask based on the target processing results of each accessed control system and the corresponding processing result samples.

In the method, before each control system is checked, a task target of a current checking subtask is determined, and then a processing result sample and a target processing result corresponding to each control system are obtained according to the task target so as to obtain a task checking result of the control system in a comparison mode. Therefore, whether each control system accords with the corresponding check requirement or not can be effectively judged, and whether the performance of the control system is stable or not can be effectively judged.

In an optional implementation manner, the acquiring, based on the task targets, a processing result sample and a target processing result corresponding to each accessed control system includes:

if the task target is an operation performance assessment, acquiring operation control samples corresponding to each accessed control system from the assessment subtask; issuing each operation control sample as a control object to a corresponding control system so that the corresponding control system performs control on the operation control samples to obtain processing result samples corresponding to each accessed control system; issuing each processing result sample to a corresponding distribution control system as a distribution control object so that the corresponding distribution control system distributes and controls the processing result sample to obtain a target processing result corresponding to each accessed distribution control system;

or if the task target is to search for the graph performance check by the graph, invoking an external image quality detection interface, and performing quality detection on the captured historical images in the distribution results fed back by the accessed distribution systems to obtain a first quality score of the historical images; screening historical images with the first quality scores in a first preset score range from the historical images corresponding to all the accessed control systems according to the first quality scores of the historical images, and taking the historical images with the first quality scores in a first preset score range as a processing result sample corresponding to each accessed control system; issuing each processing result sample to a corresponding distribution control system to perform graph searching to obtain a target processing result corresponding to each accessed distribution control system;

Or if the task target is the recall rate examination, determining the control object with the track of each accessed control system in the preset recall rate examination days from the control results fed back by each accessed control system; issuing the distributed control objects with the tracks to corresponding distributed control systems to perform graph searching to obtain processing result samples corresponding to each accessed distributed control system; acquiring the track of the control object with the track in the same time and the same equipment from the corresponding control system to serve as a target processing result;

or if the task target is accuracy rate assessment, locally acquiring a control object with a track with a preset accuracy rate assessment number from the preset accuracy rate assessment days to obtain a processing result sample corresponding to each accessed control system; obtaining all tracks of the processing result samples from the corresponding distributed control systems to obtain target processing results corresponding to each accessed distributed control system;

or if the task target is checked by the file searching standard rate, an external image quality detection interface is called, and quality detection is carried out on the file cover photos of the distributed objects corresponding to the local accessed distributed control systems, so that a second quality score of the file cover photos is obtained; screening the file cover photos with the second quality scores in a second preset score range from the file cover photos corresponding to each accessed distribution control system according to the second quality scores of the file cover photos to obtain processing result samples corresponding to each accessed distribution control system; and issuing each processing result sample to a corresponding distribution control system to perform graph searching, so as to obtain a target processing result corresponding to each accessed distribution control system.

In this manner, for different task targets, the processing result sample of the current task target is determined from the data fed back by the local or to-be-checked control system, and then the target processing result corresponding to the processing result sample is obtained from the to-be-checked control system, so that the performance of each accessed control system under the current task target can be effectively evaluated by using the processing result sample and the target processing result corresponding to each accessed control system.

In a second aspect, the present invention provides a management device, including:

the task information comprises a target control area and a target control object;

the distribution control system selection module is used for determining a target distribution control system from the accessed distribution control systems based on the target distribution control area;

the large library task selection module is used for acquiring large library control tasks corresponding to the target control tasks in the target control system if the target control area is larger than a preset range, wherein the large library control tasks are used for controlling a plurality of control objects in the same target control area;

And the control object control module is used for writing the target control object into the large library control task so as to control the target control object to obtain a control result.

In a third aspect, the present invention provides a distributed management platform, including:

the view library is used for issuing a target control task and task information thereof, wherein the task information comprises a target control area and a target control object;

the multi-algorithm management and dispatch engine system is connected with the view library and is used for executing the management method of any implementation mode of the first aspect or the corresponding implementation mode;

and the at least one control system is connected with the multi-algorithm management and scheduling engine system and is used for controlling the target control object under the scheduling of the multi-algorithm management and scheduling engine system to obtain a control result.

In a fourth aspect, the present invention provides a computer readable storage medium having stored thereon computer instructions for causing a computer to execute the above-described first aspect or any one of its corresponding embodiments of the method for managing management of a fabric.

Drawings

In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings that are needed in the description of the embodiments or the prior art will be briefly described, and it is obvious that the drawings in the description below are some embodiments of the present invention, and other drawings can be obtained according to the drawings without inventive effort for a person skilled in the art.

FIG. 1 is a flow chart of a first method of administration and management according to an embodiment of the present invention;

FIG. 2 is a flow chart of a second method of administration and management according to an embodiment of the present invention;

FIG. 3 is a diagram of a multi-engine parsing platform framework in accordance with an embodiment of the present invention

FIG. 4 is a schematic diagram of data interaction logic of a multi-engine parsing platform in a large deployment scenario according to an embodiment of the present invention;

FIG. 5 is a schematic diagram of overall data interaction logic of a multi-engine parsing platform in accordance with an embodiment of the present invention;

FIG. 6 is a flow chart of a third method of administration and management according to an embodiment of the present invention;

FIG. 7 is a schematic flow chart of a control service for controlling objects according to an embodiment of the invention;

FIG. 8 is a schematic flow chart of a file aggregation service according to an embodiment of the present invention;

FIG. 9 is a schematic flow diagram of a graph search service according to an embodiment of the present invention;

fig. 10 is a block diagram of a configuration of a management apparatus according to an embodiment of the present invention.

Detailed Description

For the purpose of making the objects, technical solutions and advantages of the embodiments of the present invention more apparent, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings in the embodiments of the present invention, and it is apparent that the described embodiments are some embodiments of the present invention, but not all embodiments of the present invention. All other embodiments, which can be made by those skilled in the art based on the embodiments of the invention without making any inventive effort, are intended to be within the scope of the invention.

With the increasing perfection of video monitoring construction, face images can be acquired in corresponding control areas through video monitoring equipment, and then the face images are sent to a background of a control system for recognition and comparison in a network transmission mode, so that control objects can be accurately recognized, and a good monitoring effect is achieved. When a person responsible for controlling the face of a controlled object is controlled, a specific control task needs to be filled in the control management platform. However, for the task of large-scale control area, because the area of the control area is large, the information to be filled is large, which results in complex control flow and low control efficiency.

In view of the foregoing, there is provided in accordance with an embodiment of the invention a method of administration, it being noted that the steps illustrated in the flowcharts of the figures may be performed in a computer system, such as a set of computer executable instructions, and that although a logical order is illustrated in the flowcharts, in some cases the steps illustrated or described may be performed in an order other than that illustrated herein.

The embodiment provides a distributed control management method which can be used for accessing distributed control management platforms of different distributed control systems. Fig. 1 is a flowchart of a method of administration management according to an embodiment of the present invention, as shown in fig. 1, the flowchart includes the steps of:

Step S101, acquiring a target control task and task information thereof, wherein the task information comprises a target control area and a target control object.

It should be noted that, according to the actual situation, the task information may further include a task code of the target task, a task type, and an object code of the target task. If a plurality of large Fan Weibu-controlled control tasks exist in each accessed control system, the target control task does not need to be created again, but the target control object is placed in the corresponding control task after being analyzed.

Step S102, determining a target control system from the accessed control systems based on the target control area.

Step S103, if the target control area is larger than a preset range, acquiring a large library control task corresponding to the target control task in the target control system, wherein the large library control task is used for controlling a plurality of control objects in the same target control area.

Before the large library control task corresponding to the target control task in the target control system is obtained, an instruction for creating the large library control task can be initiated to each control system through an offline method, so that each control system creates the large library control task. The attribute requirements of the large library management and control task are as follows: the large library is always effective and only needs to be created once, the task code is fixed, the control area is larger than the preset range, the control threshold is fixed (such as 95%), the control time is always effective, the task control state is always in control, and the task code of the large library is required to inform a task release object (such as a view library). Of course, in actual situations, a plurality of large library management tasks can be created according to actual needs, and the number of the large library management tasks is not limited herein.

It should be noted that the preset range is determined based on the actual requirements of the corresponding large library management and control task.

Step S104, writing the target control object into the large library control task to control the target control object to obtain a control result.

It should be noted that, each control system has a Kafka task type alarm topic queue, and the control result of the control system based on the big library control task can be put into the Kafka task type alarm topic queue for reading.

According to the method for managing and controlling the large library, the large library control task is deployed in the target control system, so that when the target control area of the obtained target control task is larger than the preset range, the large library control task corresponding to the target control task can be obtained, and the target control object is directly written into the large library control task to control the target control object. Therefore, when the range of the control area is larger, the task information filling flow and the control flow can be simplified, so that the control efficiency is improved.

Fig. 2 is a flowchart of a method of administration management according to an embodiment of the present invention, as shown in fig. 2, the flowchart including the steps of:

step S201, a target control task and task information thereof are acquired, wherein the task information comprises a target control area and a target control object. The details of step S101 are not repeated here.

Step S202, determining a target control system from the accessed control systems based on the target control area. The details of step S102 are not repeated here.

Step S203, if the target control area is greater than the preset range, acquiring a large library control task corresponding to the target control task in the target control system, where the large library control task is used to control a plurality of control objects in the same target control area. The details of step S103 are not repeated here.

And step S204, writing the target control object into the large library control task to control the target control object to obtain a control result.

By way of example, the task code of the corresponding large-library control task is assigned to the target control object, and the target control object is issued to the corresponding target control system according to the large-library control task mode.

Further, the method further comprises:

step S205, if the target control area is smaller than or equal to the preset range, issuing the target control task and the target control object thereof to the target control system, so as to perform control on the target control object to obtain a control result.

In actual operation, the target control system directly issuing the target control task to the target control area can be determined according to the range of the target control area, or the target control object of the target control task is written into the corresponding large-scale library control task, or the task type is written into the task information according to the range of the target control area by the task issuing object (such as a view library), and the range of the control area corresponding to the current target control task is judged according to the task type in the task information, so that the target control object is directly issued to the target control system or is issued to the target control system after being written into the large-scale library control task.

According to the control management method provided by the embodiment, when the target control area is smaller than or equal to the preset range, the target control task and the target control object thereof are directly issued to the target control system, so that the control efficiency of the target control task in a small range can be ensured.

Specifically, the method further comprises:

step S206, when receiving a target control result fed back by any accessed control system, obtaining a target task code corresponding to the target control result.

Step S207, when the target task code is the task code of the large library control task, a control object corresponding to the target control result is obtained.

It should be noted that, if the task code of the task associated with the target task is the task code of the big library task, it may be determined that the current target task is based on the feedback of the big library task, but because the task objects associated with the big library task may correspond to different task, further obtaining the task object corresponding to the target task object is needed, so as to query the corresponding task according to the task object corresponding to the target task object, and further send the target task to the service system corresponding to the task.

Step S208, determining the control task corresponding to the target control result and the task state thereof from the pre-stored control tasks corresponding to the large library control tasks according to the control object corresponding to the target control result.

Specifically, the object codes of the control objects associated with the target control result can be obtained, the task codes of the associated control tasks are inquired and obtained according to the object codes of the associated control objects, so that the task states of the associated control tasks are obtained, and whether the target control tasks are in the control states or not is judged according to the task states. Specifically, the task state includes information such as a task control state, a task control area, a bayonet, a task control device, and a task effective time corresponding to the target control result.

Step S209, when the task state meets a preset effective control condition, sending the target control result to a corresponding service system according to a control task corresponding to the target control result.

Specifically, whether the task is still in the control state, the effective time of the task, the control area corresponding to the object code of the control object and the control equipment meet the control task requirement are judged based on the task state, whether the task state meets the effective control condition or not is determined, namely whether the control task of the target control result is effective or not is determined, and whether the target control result is effective or not is further judged. It should be noted that, if the target control task corresponding to the control result does not meet the effective control condition, it is determined that the target control result is invalid, so that the target control result does not need to be sent to the corresponding service system.

According to the control management method provided by the embodiment, when the received target control result is the target control result fed back by the library control task, the corresponding control task and the task state thereof are inquired according to the control object corresponding to the target control result, and when the task state meets the preset effective control condition, the target control result is sent to the corresponding service system. Therefore, the invalid control result can be prevented from being fed back to the corresponding service system, so that unnecessary occupied storage resources and data transmission resources are reduced, and the control efficiency is improved.

Further, the method further comprises:

step S210, when the target task code is not the task code of the big library task, determining a task corresponding to the target task code according to the target task code, so as to send the target task code to a corresponding service system.

It can be appreciated that if the task code of the associated task does not belong to the task code of the big library task, it can be determined that the current target task result is not based on the big library task feedback. Therefore, the target control result can be directly fed back to the business system corresponding to the task code of the control task.

According to the distribution control management method provided by the embodiment, when the received target distribution control result is not the distribution control result of the large-scale distribution control task, the target distribution control result is directly sent to the corresponding service system according to the target task code, so that the transmission efficiency of the distribution control result of the small-scale distribution control task can be ensured.

For example, referring to fig. 3, taking a software architecture "multi-engine parsing platform" applying the management and control method of the embodiment of the present invention as an example, the multi-engine parsing platform is composed of an external system and an internal system. 1. External system: the system comprises a view library, a video image service support platform and various distribution control systems; the control system corresponds to each control area, and the algorithm of different control areas is called based on task requirements. 2. Internal system: a multi-algorithm management and dispatch engine system, the multi-algorithm management and dispatch engine system comprising: the system comprises a message queue module, an algorithm supervision and evaluation module, an audit log module, a data checking module, an operation and maintenance management module, an interface scheduling module and an algorithm scheduling module. Wherein the message queue module: and sending the algorithm analysis result to Kafka, analyzing the consumption storage by the multi-engine module, and forwarding the analysis result to a view library. The algorithm supervision and evaluation module: and (3) checking whether the algorithm of each control system meets the requirements of users, whether the performance meets the standards and is stable, displaying the checking result in an operation and maintenance management module, and judging whether the algorithm of the current control system meets the requirements according to the page display data information. Audit log module: and the method is responsible for processing calling information and detailed information among the modules and displaying the calling information and the detailed information to the operation and maintenance management module according to a certain mode. And the data reconciliation module: and comparing the two pieces of data of the analysis data stored in the view library with the analysis data stored by the algorithm, and displaying the difference to the operation and maintenance management module. And the operation and maintenance management module is used for: and the system is responsible for receiving the data of each module to display the data and managing the online condition of the control system. Interface scheduling module and algorithm scheduling module: and the task is responsible for processing the task issued by the view library and the third-party service, the task is forwarded to each distribution control system according to the rest protocol message, and then the view library is trimmed and recalled according to the return result.

Specifically, the multi-algorithm management and scheduling engine system receives picture processing tasks dispatched by applications, creates intelligent analysis tasks, performs intelligent analysis on various pictures, supports flow load balancing, and pushes analysis tasks and results in a standard protocol mode. The task scheduling and classifying flow is transferred between the application and the algorithms of the plurality of the control systems, so that the algorithms of the plurality of the control systems are shielded. Specifically, the view library may perform data transmission and interaction with the multi-algorithm management and scheduling engine platform through Kafka or interface, and select a transmission mode according to the size of the data volume, for example: the task is a target control task, the target control task is a large-library control task of a large control area, at the moment, the data flow of the control object is large, and the data needs to be issued to a plurality of control areas, and then Kafka is adopted to transmit the data. If only a certain control object is controlled, the data can be transmitted by using an interface mode.

It can be understood that the multi-algorithm management and dispatch engine system is used as a platform for managing each of the distributed control systems, and manages the resource utilization rate, the function availability, the concurrency stability and the abnormal fault tolerance of each of the distributed control systems in a mode of checking, monitoring and alarming. In the multi-algorithm management and dispatch engine system, service registration, configuration center, assessment center, request routing gateway technology (such as service registration, configuration center, request routing gateway and the like deployed on a dolphin management platform) and the like can be adopted to perform one-key starting and deployment, the service registration is used for observing whether a management and control system is stable, the assessment center is used for assessing whether various functions, performances and stability requirements of clients are met within a period of time, the routing gateway technology provides a unique entry for the system, authentication is performed on requests of a requester, each request authority is identified, abnormal requests are intercepted, and the security of background service is guaranteed.

For example, referring to fig. 4 and fig. 5, taking a multi-engine parsing platform in a large control scenario, a target control task is a user control task as an example, the control management method of the present invention is further described:

1. the view library is responsible for storing snapshot data, and can upload images of objects to be controlled and associated information in the view library, wherein the associated information comprises a target control area. As shown in fig. 4, the user may be based on various business needs or on predefined rule needs, such as: the algorithm monitoring task can preset a monitoring period, the algorithm data of each distribution control system are summarized in the monitoring period, and after the specified time is reached, the algorithm data are analyzed to realize the monitoring and evaluation of each distribution control system so as to trigger the task, and the task is issued to a plurality of algorithm and scheduling engine systems. Further, if the task is a big library control task, the view library identifies the task code (i.e., task ID) and the task type of the target control task as the big library control task and issues the task code to the multi-algorithm and scheduling engine system. It should be noted that, the view library may generate a user control task and issue the user control task to the multi-algorithm and scheduling engine system, where the user distributes corresponding control objects based on the issued task, and the user divides the plurality of control objects into a plurality of user control tasks or a pre-built large library control task based on the control requirement. For example: 100 control objects need to be subjected to large Fan Weibu control, and a plurality of control tasks controlled by large Fan Weibu are available, so that the task does not need to be created again, and the 100 control objects can be placed in the corresponding control tasks after being analyzed.

2. The multi-algorithm management and scheduling engine system receives a target control task (labeled as a user control task in fig. 4) issued by the view graph library, and if the user control task is a small control task for controlling a single control object, the target control area in the task information can be issued to a corresponding target control area to call the corresponding control system for processing. If the user control task is a user control task aiming at a large area, recording information of the user control task, converting a target control object of the user control task into a control object of a large library control task, assigning a corresponding task code (a large library task ID in the figure) of the large library control task to the target control object, and issuing the target control object to a control system according to a large library control task mode. In addition, the multi-algorithm management and scheduling engine system can establish a large library control task table and a control object memory table of the large library control task based on the received information of the control object in the large library control task and the task code of the user control task, so as to cache the information of the user control task and the information of the target control object into the memory, thereby realizing subsequent rapid calculation and comparison. And initiating an instruction for creating the large library task to each distribution control system in an offline mode, so that each distribution control system creates the large library distribution control task.

3. And the cloth control system comprises: each control system is responsible for analyzing snapshot data, comparing the snapshot data, receiving the scheduling of the multi-engine platform, and returning the result to the view library for display. Specifically, the control system is provided with a Kafka task type alarm subject queue, the control system adds tasks with task codes of large-library control tasks to the large-library control library for control based on the received control tasks, and performs control on task codes of small-library control tasks based on requirements. And if the alarm information is generated based on the control result, the alarm information is put into a Kafka task type alarm subject queue.

4. The view library reads the alarm information of the small distributed tasks in the Kafka task type alarm topic queue, namely: the task code of the task in the message queue is not the alarm information of the task code of the large library task, and the corresponding alarm information is sent to the corresponding service system based on the task code of the task in the read alarm information.

5. The multi-algorithm management and scheduling engine system reads the alarm information that the task code of the task in the Kafka task type alarm subject queue is the task code of the large library task, obtains the object code (in the figure, the ID of the object) of the associated task, finds the actual task code of the user task associated with the object code of the object by querying the memory table of the large library object, and then queries the memory table of the large library task to obtain the information of the task-corresponding task-distributing area, the bayonet, the equipment, the effective time of the task, the task state and the like. Judging whether the task is still in a control state, the effective time, a control area of the object code of the alarm face and control equipment meet the control task requirement or not based on the information, and further judging whether the alarm message is effective or not; and writing the effective alarm data into a large library report alarm subject queue, and reading and consuming by the view library.

Fig. 6 is a flowchart of a method of administration management according to an embodiment of the present invention, as shown in fig. 6, the flowchart including the steps of:

step S301, a target task and task information thereof are acquired, where the task information includes a target task area and a target task object. The details of step S101 are not repeated here.

Step S302, determining a target control system from the accessed control systems based on the target control area. The details of step S102 are not repeated here.

Step S303, if the target control area is greater than the preset range, acquiring a large library control task corresponding to the target control task in the target control system, wherein the large library control task is used for controlling a plurality of control objects in the same target control area. The details of step S103 are not repeated here.

And step S304, writing the target control object into the large library control task to control the target control object to obtain a control result. The details of step S104 are not repeated here.

In some alternative embodiments, the method further comprises:

step S305, a preset system assessment task is obtained, wherein the system assessment task comprises at least one assessment subtask.

Step S306, the accessed control system is checked based on the checking subtasks, and task checking results of each accessed control system corresponding to each checking subtask are obtained.

Specifically, the step S306 includes:

and step 3061, analyzing the assessment subtasks to determine task targets.

Specifically, the task targets include at least one of operation performance assessment, graph search performance assessment, recall assessment, accuracy assessment, diffusivity assessment, graph search standard rate assessment, track search standard rate assessment, data accounting assessment, file accounting consistency rate assessment and file track accounting consistency rate assessment.

Step S3062, based on the task targets, obtaining processing result samples and target processing results corresponding to each accessed control system.

Step 3063, obtaining task assessment results of each accessed control system corresponding to the assessment subtask based on the target processing results of each accessed control system and the corresponding processing result samples.

Step S307, corresponding to each accessed control system, fusing task assessment results of all assessment subtasks to obtain system assessment results of the accessed control system.

Step S308, when the system check result is that the distributed control system does not reach the standard, corresponding system alarm information is generated.

According to the control management method provided by the embodiment, each accessed control system is checked through the preset system check task, and when the system check result of the control system is that the control system does not reach the standard, corresponding system alarm information is generated, so that related personnel can check the control system in time, effective execution of target task object control is ensured, and the control efficiency and control quality are improved.

It should be noted that, with the continuous increase of domestic and foreign market demands and the development of intelligent, integrated and standardized algorithm model technologies, a distribution control system cannot completely meet the demands of clients for analysis of massive video and image resource data. Therefore, the multi-engine analysis platform is introduced to decouple the application platform from the algorithm services of a plurality of distributed control systems, unify the protocols, and transmit the protocols between the application and the algorithms to each required algorithm so that the services are processed by a unique standard protocol. The multi-algorithm management and scheduling engine system is used for issuing target control tasks and target control objects based on the target control tasks and target control objects issued by the view library, issuing the target control tasks to a target control area, and processing the control objects by the algorithm library corresponding to the control system of the target control area. However, the conventional parsing engine application and algorithm are single, for example, an algorithm of a distributed control system appears again, and the former needs to be re-connected with a new algorithm, and even reconstruction can appear to ensure normal service, so that the efficiency is low. Therefore, if a plurality of distributed control systems exist, each distributed control system needs to be managed and monitored, and the assessment is carried out regularly based on the management and monitoring results, so that the distributed control systems used for subsequent task scheduling are adjusted based on the assessment of each distributed control system, the optimal scheduling of the distributed control systems is realized, and the task processing efficiency and accuracy are improved.

Further, if the task objective is the operation performance assessment, step S3062 includes:

and a step a1, acquiring operation control samples corresponding to each accessed control system from the assessment subtask.

Specifically, in actual operation, the task target may be started to be an assessment subtask for assessing the computing performance based on a preset computing performance assessment period or user operation. The operation control sample comprises a file cover photo and a face photo corresponding to the file. Specifically, the operation control sample may be obtained by: and randomly taking 20 real-name files with more than or equal to 30 tracks newly added to the three days to the distribution system A, and providing 10 file cover photos to the distribution system B and the distribution system C respectively as operation distribution samples. And randomly taking 20 real-name files with more than or equal to 30 tracks newly added to the distribution system B in three days, and providing 10 file cover photos for the distribution system A and the distribution system C respectively as operation distribution samples. The control system C takes 20 real-name files with more than or equal to 30 tracks from the control area a in three days and provides the files to the control area b as operation control samples. The general principle of operation and control sample selection is as follows: the operation control samples need to come from a control area which is not in charge of the corresponding control system so as to avoid interference. In addition, the method also supports the user to directly provide operation control samples.

And a2, issuing each operation control sample serving as a control object to a corresponding control system so that the corresponding control system performs control on the operation control sample to obtain a processing result sample corresponding to each accessed control system.

For example, the mapping algorithm is performed on the covers of 10 files to generate the corresponding files, so that 10 pictures with the similarity of the belonging control system A exceeding 95%, the similarity of the belonging control system B exceeding 96% and the similarity of the belonging control system C exceeding 93% are found, 100 pictures are taken as the samples of the processing results, and the samples of the processing results are renumbered according to the input control area/control system.

And a3, issuing each processing result sample to a corresponding control system as a control object, so that the corresponding control system controls the processing result sample to obtain a target processing result corresponding to each accessed control system.

The method includes the steps of sending a control task corresponding to the processing result sample to a corresponding control system, sending the processing result sample to the corresponding control system as a control object, and collecting alarm information generated by the control system to obtain target processing results corresponding to each accessed control system. And then, sending the face photo corresponding to the processing result sample to a view library.

Specifically, when the task objective is the calculation performance assessment, step S3063 includes:

according to a preset alarm leakage rate assessment rule, calculating alarm leakage rates of processing result samples and target processing results corresponding to each accessed control system;

when the alarm omission rate is smaller than or equal to a preset alarm omission rate threshold value, judging that a task assessment result of the corresponding control system corresponding to the assessment subtask is that the alarm omission rate is qualified;

calculating the alarm accuracy of a processing result sample and a target processing result corresponding to each accessed control system according to a preset alarm accuracy assessment rule;

and when the alarm accuracy is greater than or equal to a preset alarm accuracy threshold, judging that the task assessment result of the corresponding control system corresponding to the assessment subtask is qualified in alarm accuracy.

Optionally, the preset alarm leakage rate threshold is 2%, and the preset alarm accuracy rate threshold is 96%. In actual operation, the alarm omission rate allows an error of 3%.

It can be understood that the operation performance assessment comprises two assessments of alarm omission rate and alarm accuracy rate.

According to the distribution management method provided by the embodiment, when a task target is an operation performance assessment, an operation distribution sample is firstly obtained from an assessment subtask, the operation distribution sample is issued to a corresponding distribution system to obtain a processing result sample, and then the processing result sample is issued to the corresponding distribution system in a reverse direction to obtain a target processing result. Therefore, the processing result samples and target processing results corresponding to the control systems can be effectively utilized, the control result conditions fed back by the control systems can be checked, and the corresponding operation performance check results can be accurately obtained.

Further, if the task objective is to search for the graph performance check with the graph, step S3062 further includes:

and b1, invoking an external image quality detection interface, and performing quality detection on the captured historical images in the control results fed back by each accessed control system to obtain a first quality score of the historical images.

And b2, screening the historical images with the first quality scores in a first preset score range from the historical images corresponding to the accessed control systems according to the first quality scores of the historical images, and taking the historical images with the first quality scores in the first preset score range as a processing result sample corresponding to each accessed control system.

The first preset score range needs to be determined according to the actual situation, and is not limited herein.

For example, in actual operation, the graphical search performance check may include the following four check items: high-quality image retrieval standard-reaching rate assessment (1 month hot data), high-quality image retrieval standard-reaching rate assessment (2 month cold data), low-quality image retrieval standard-reaching rate assessment (1 month hot data), and low-quality image retrieval standard-reaching rate assessment (2 month cold data). It should be noted that, the hot data and the cold data refer to different shooting environments of the images, and in actual operation, the processing result samples may be obtained according to a fixed graph searching performance assessment period. The graph searching performance checking period can be set according to requirements, for example: once daily, starting 5-point daily to search the graph to perform performance examination, and sequentially executing examination items to be performed daily. Further, the corresponding task may be triggered based on the user operation. Specifically, the processing result samples are different based on the processing result samples selected by the different examination items, and the processing result samples are specifically as follows: 1. the examination samples for high-quality image retrieval standard-reaching rate examination (1 month hot data) are as follows: and randomly selecting 50 high-quality images within nearly one month according to a distribution control area in charge of a distribution control system, and confirming that only one face exists in the extracted high-quality images, wherein the high-quality images refer to images with image quality scores in a first preset quality range, such as historical images with the first quality scores in the first preset quality range, and the first preset quality range is more than 80 minutes. 2. The examination samples for high-quality image retrieval standard-reaching rate examination (2 months cold data) are as follows: the high-quality images of 25 sheets in the past 30 to 60 days and 60 to 90 days are randomly selected according to the control area in charge of the control system, a rider needs to select images after nine points at night, the images after nine points are selected based on the consideration of the shooting environment of the images, and in actual operation, the images can be changed according to actual requirements without limitation. 3. The examination samples for low-quality image retrieval standard-reaching rate examination (1 month hot data) are as follows: and randomly selecting 50 low-quality images within a month according to a distribution area in charge of a distribution system, wherein the images after nine night are selected by a driver, the low-quality images refer to images with image quality scores in a second preset quality range, such as historical images with the first quality scores in the second preset quality range, and the second preset quality range is 50 to 60 minutes. 4. The examination samples for low-quality image retrieval standard-reaching rate examination (2 months cold data) are as follows: the control area in charge of the control system is randomly chosen for 30 to 60 days, 25 images with low quality are respectively chosen for 60 to 90 days (the situation that no control area in charge is chosen) is not excluded, and a rider needs to choose images after nine points at night.

And b3, issuing each processing result sample to a corresponding distribution control system to perform graph searching to obtain a target processing result corresponding to each accessed distribution control system.

Specifically, the acquired processing result samples are sequentially used for calling the graph searching function of the corresponding control system, and the conditions of time spent by interface calling and whether the original graph can be hit before the similarity are checked and used as the target processing result of the current control system.

Specifically, when the task objective is to search for the graph performance assessment in the graph, step S3063 described above includes:

calculating the image retrieval standard rate of the processing result samples and the target processing results corresponding to each accessed control system according to a preset graph searching performance assessment rule;

when the image retrieval standard reaching rate is larger than or equal to a preset image retrieval standard reaching rate threshold value, judging that the task checking result of the corresponding control system corresponding to the checking subtask is that the image retrieval standard reaching rate is qualified.

Optionally, the preset image retrieval achievement rate threshold of the high-quality image is 95%, and the preset image retrieval achievement rate threshold of the low-quality image is 90%.

According to the distribution management method provided by the embodiment, when the task target is the graph searching performance examination, the corresponding processing result sample is extracted from the distribution results fed back by the distribution system, and then the processing result sample is issued to the corresponding distribution system to perform graph searching, so that the target processing result is obtained, and the processing result sample can be searched in the corresponding distribution system. Therefore, the processing result samples and the target processing results corresponding to the control systems can be effectively utilized, and the corresponding graph searching performance assessment results can be accurately obtained.

It should be noted that, the file track performance assessment of the distributed control system is mainly divided into the following five assessment items: recall rate assessment, accuracy rate assessment, diffusivity assessment, file search standard rate assessment by a picture, file track retrieval standard rate assessment.

Further, if the task target is a recall check, the step S3062 further includes:

step c1, determining the control objects with tracks in the preset recall rate examination days of each accessed control system from the control results fed back by each accessed control system.

Optionally, the preset recall rate is checked for 10 days.

Specifically, the recall rate assessment can be performed according to a preset recall rate assessment period, and the recall rate assessment period can be set according to requirements, for example, an assessment task is started at 20 points per week. And then storing Kafka consumption data into a local database, and randomly drawing unregistered valid files of 20 control objects with tracks in the last 10 days from the local database according to a control area which is responsible for a control system.

And c2, issuing the control objects with the tracks to corresponding control systems to perform graph searching to obtain processing result samples corresponding to each accessed control system.

Specifically, the file cover of the control object with the track is used for carrying out graph searching on the control system, and the similarity threshold is set to be 90%.

And c3, acquiring the track of the control object with the track at the same time and under the same equipment from the corresponding control system, and taking the track as a target processing result.

It should be noted that, the examination target of the recall rate examination is: the control system searches the pictures to find out whether the face can be blocked or not, and specifically checks the face in a sampling mode.

Specifically, when the task target is a recall rate assessment, step S3063 described above includes:

according to a preset recall rate checking rule, comparing a processing result sample corresponding to each accessed control system with a target processing result, and calculating to obtain a recall rate;

and when the recall rate is greater than or equal to a preset recall rate threshold, judging that the task assessment result of the corresponding control system corresponding to the assessment subtask is qualified in recall rate.

Optionally, the preset recall threshold is 80%.

It should be noted that, the comparison of the processing result samples corresponding to each accessed control system with the target processing result is to determine whether the track corresponding to the processing result sample can be found from the target processing result.

In the method for managing and controlling tasks provided in this embodiment, when a task target is a recall rate assessment, firstly, a control object corresponding to a track meeting a recall rate assessment requirement is extracted from a control result fed back by a control system, and is issued to a corresponding control system to perform graph searching, so as to obtain a corresponding processing result sample, and a track of the control object under the same time and the same equipment is obtained from the corresponding control system to be used as a target processing result. Therefore, whether the target processing result can be aggregated or not can be accurately judged by utilizing the processing result samples and the target processing results corresponding to the control systems so as to obtain the corresponding recall rate assessment result.

Further, if the task target is accuracy assessment, step S3062 further includes:

step d1, locally acquiring the control objects with the tracks of the preset accuracy rate checking number within the preset accuracy rate checking days, so as to obtain the processing result samples corresponding to each accessed control system.

Optionally, the number of days for checking the preset accuracy is 10 days, and the number of checks for the preset accuracy is 10.

Specifically, from Kafka consumption data stored in a local database, 20 valid files with more than 10 tracks in the last 10 days and not logged off are randomly selected from the local database and used as a processing result sample.

And d2, acquiring all tracks of the processing result samples from the corresponding control system to obtain target processing results corresponding to each accessed control system.

Specifically, all tracks of the processing result sample are obtained from the control system, the image quality of the tracks is scored, the tracks with the image quality score higher than 65 scores and the processing result sample are subjected to similarity judgment through an external interface, and the similarity is higher than 90% and is considered to be qualified.

Specifically, when the task target is accuracy assessment, step S3063 includes:

calculating the accuracy of a processing result sample and a target processing result corresponding to each accessed control system according to a preset accuracy assessment rule;

and when the accuracy rate is greater than or equal to a preset accuracy rate threshold value, judging that the task assessment result of the corresponding control system corresponding to the assessment subtask is qualified in accuracy rate.

Optionally, the preset accuracy threshold is 95%. In practice, an error of 5% may be allowed.

According to the control management method provided by the embodiment, when the task target is accuracy assessment, the control objects corresponding to the control systems meeting the accuracy assessment requirements are obtained locally, the control objects are taken as the processing result samples, and then all tracks corresponding to the processing result samples are obtained again from the corresponding control systems, so that the target processing result is obtained. Therefore, the processing result sample and the target processing result can be effectively utilized, whether the control result fed back by the control system based on the control object is accurate or not can be determined, and the accuracy rate assessment result can be obtained.

Further, if the task objective is to check the achievement rate of the search file according to the graph, the step S3062 further includes:

and e1, invoking an external image quality detection interface, and performing quality detection on the file cover photos of the control objects corresponding to the local accessed control systems to obtain a second quality score of the file cover photos.

And e2, screening the file cover photos with the second quality scores in a second preset score range from the file cover photos corresponding to each accessed distribution control system according to the second quality scores of the file cover photos, so as to obtain a processing result sample corresponding to each accessed distribution control system.

Optionally, the second preset score range is greater than 80 scores.

It should be noted that, in actual operation, 100 valid files that are not logged out may be randomly extracted from the distribution area that the distribution system is responsible for, and the file cover photo with the second quality score of more than 80 minutes is used as the processing result sample.

And e3, issuing each processing result sample to a corresponding distribution control system to perform graph searching, and obtaining target processing results corresponding to each accessed distribution control system.

Specifically, the corresponding control system is searched by using the processing result sample, and the calling time and the condition of the searched first image hit the processing result sample are checked to obtain the target processing result of the corresponding control system.

Specifically, when the task objective is to check the achievement rate of the search file in the graph, the step S3063 includes:

calculating the picture search standard rate of the processing result samples and the target processing results corresponding to each accessed control system according to a preset picture search standard rate checking rule;

and when the standard rate of the picture searching is larger than or equal to a preset standard rate threshold value of the picture searching, judging that the task checking result of the corresponding control system corresponding to the checking subtask is that the standard rate of the picture searching is qualified.

Optionally, the map search standard rate threshold is 95%. In practice, an error of 3% may be allowed.

According to the distribution management method provided by the embodiment, when a task target is to check the picture searching standard rate, file cover photos of distribution objects corresponding to each distribution system meeting the picture searching standard rate check requirement are obtained locally to serve as processing result samples, and then the processing result samples are issued to the corresponding distribution system to perform picture searching, so that corresponding target processing results are obtained. Therefore, the graphic file searching performance of the control system can be effectively evaluated by utilizing the processing result sample and the target processing result.

Further, if the task target is the diffusivity assessment, the step S3062 further includes:

and f1, locally acquiring the number of real-name control objects corresponding to each accessed control system as a processing result sample corresponding to each accessed control system.

It should be noted that, the real-name control object includes a control object that has logged out of the archive, i.e. has failed.

And f2, locally acquiring the number of the identification documents corresponding to each accessed control system as a target processing result corresponding to each accessed control system.

Specifically, when the task target is the diffusivity assessment, step S3063 includes:

calculating the ratio of a processing result sample corresponding to each accessed control system to a target processing result according to a preset diffusivity assessment rule to obtain diffusivity;

and when the diffusivity is smaller than or equal to a preset diffusivity threshold, judging that the task assessment result of the corresponding control system corresponding to the assessment subtask is that the diffusivity is qualified.

Optionally, the preset diffusivity threshold is 150%. In practice, an error of 3% may be allowed.

It can be understood that the number of real-name control objects divided by the number of identification documents is the diffusivity.

It should be noted that, the goal of the diffusivity assessment is to determine the degree of multiple steps of one person of the real-name control object.

Further, if the task target is the track retrieval achievement rate assessment, the step S3062 further includes:

step g1, locally acquiring the distribution control objects of the tracks with the preset track retrieval and examination number in the preset track retrieval and examination days, so as to obtain track retrieval and examination objects corresponding to each accessed distribution control system.

Optionally, the preset track search and check day is one month, and the number of the preset track search and check is 5.

In an exemplary embodiment, a batch of files of the control objects are obtained locally, and according to a group of each month in the last half year, more than 5 control objects with tracks in one month are required to be used as tracks of a corresponding control system to search the examination objects.

Step g2, obtaining the track of the track retrieval assessment object from the task release object to obtain a processing result sample corresponding to each accessed control system, wherein the task release object is used for releasing a target control task.

Illustratively, 5 tracks are obtained from a task publishing object (e.g., a gallery).

And g3, acquiring the track corresponding to the track retrieval check object from the corresponding control system to obtain target processing results corresponding to each accessed control system.

Specifically, when the task target is the track retrieval achievement rate assessment, the step S3063 includes:

checking whether the processing result samples corresponding to each accessed control system exist in the corresponding target processing result, and checking the calling time consumption and track hit condition of an interface corresponding to the target processing result to obtain a checking result, wherein the track hit condition is used for indicating the condition that the target processing result hits the corresponding processing result sample;

obtaining track retrieval standard reaching rate corresponding to each accessed distribution control system according to a preset track retrieval standard reaching rate checking rule and the checking result;

and when the track retrieval standard reaching rate is larger than or equal to a preset track standard reaching rate threshold value, judging that the task checking result of the corresponding control system corresponding to the checking subtask is that the track retrieval standard reaching rate is qualified.

Optionally, the preset track standard reaching rate threshold is 99%. In actual operation, the track search achievement rate may allow for an error of 4%.

It should be noted that, the objective of the track search standard rate check is to check the track search performance of the control system and whether the track is lost.

Further, if the task objective is data checking, the step S3062 further includes:

and step h1, acquiring the number of face pictures of the responsible control area from each accessed control system to obtain a processing result sample of each accessed control system.

And step h2, locally acquiring the number of face pictures of the distribution areas in charge of each accessed distribution system to obtain target processing results of each accessed distribution system.

Specifically, when the task target is data checking, step S3063 includes:

calculating the data checking consistency rate of the processing result samples corresponding to each accessed control system and the target processing result according to a preset data checking rule;

when the data reconciliation consistency ratio is larger than or equal to a preset data reconciliation consistency ratio threshold, judging that the task assessment result of the corresponding distribution control system corresponding to the assessment subtask is qualified in data reconciliation assessment.

Optionally, the preset data reconciliation rate threshold is 99.5%. In actual operation, a 1% error in data reconciliation rate may be allowed.

It should be noted that, the data checking and checking target is whether the total amount of face pictures received by each control system every day is consistent with the total amount of face pictures counted by the multi-algorithm engine.

Further, if the task target is the file track reconciliation rate assessment, the step S3062 further includes:

step I1, locally acquiring the distributed objects of the tracks with the preset file track consistency rate and the check number in the check days of the preset file track consistency rate, so as to obtain the consistency rate checked objects corresponding to each accessed distributed system.

Optionally, the number of days for checking the preset file track consistency rate is 3 days, and the number of checking the preset file track consistency rate is 3.

The multi-control system engine stores Kafka consumption data into a local database, the assessment sample randomly extracts files with new track more than or equal to 3 pieces in 3 days from the local database, the number of the files is 4 times of the number of areas responsible for control by the control system, and finally half of files with best results are selected to calculate assessment results. For example: and the control system A is responsible for 6 control areas, extracts files of 24 control objects as input samples, and selects 12 files with better results to calculate the assessment results.

And step I2, obtaining the track number corresponding to the consistency rate assessment objects from the local to obtain a processing result sample corresponding to each accessed control system.

And step I3, obtaining the track number corresponding to the consistency rate assessment objects from the corresponding distributed control systems so as to obtain target processing results corresponding to each accessed distributed control system.

Specifically, when the task target is the file track reconciliation rate assessment, step S3063 includes:

according to a preset file track reconciliation rate checking rule, track reconciliation rates of processing result samples and target processing results corresponding to each accessed distribution control system are calculated;

and when the track reconciliation consistency ratio is greater than or equal to a preset track reconciliation consistency ratio threshold, judging that the task check result of the corresponding distribution control system corresponding to the check subtask is that the file track reconciliation consistency ratio is qualified.

Optionally, the preset track reconciliation rate threshold is 99.5%. In practice, a 3% error in the uniformity rate may be allowed.

It should be noted that, the checking target of checking the file track accounting consistency rate is whether the track number of the files in the control system is consistent with the track number of the files stored locally, and specifically, checking is performed in a sampling mode.

Further, if the task target is the file reconciliation rate assessment, step S3062 further includes:

and step J1, obtaining the number of the effective files of the responsible control areas from each accessed control system to obtain a processing result sample of each accessed control system.

And step J2, locally acquiring the effective file quantity of the control area responsible for each accessed control system to obtain the target processing result of each accessed control system.

Specifically, when the task target is the archive reconciliation rate assessment, step S3063 includes:

according to a preset file reconciliation rate checking rule, calculating file reconciliation rates of the processing result samples and the target processing results corresponding to each accessed distribution control system;

and when the file reconciliation consistency ratio is greater than or equal to a preset file reconciliation consistency ratio threshold, judging that the task check result of the corresponding distribution control system corresponding to the check subtask is qualified by the file reconciliation consistency ratio check.

Optionally, the preset archive reconciliation rate threshold is 99.9%. In practice, the archive reconciliation rate may be allowed to have a 5% error.

It should be noted that, the checking target of checking the file checking consistency rate is whether the number of the effective files stored in each of the distributed control systems is consistent with the number of the effective files stored locally.

Taking the above multi-engine analysis platform as an example, the following describes the method for managing and controlling the present invention with three specific embodiments:

in embodiment 1, as shown in fig. 7, each system of the control service for controlling the object mainly works as follows:

view library:

1. and synchronizing the control object information of the 4-class control systems issued by the control user platform, and constructing 4 control information tables of the control objects aiming at different control systems, wherein the corresponding control systems are allowed to read but cannot write.

2. And providing corresponding face real-time acquisition data for a corresponding distribution control system for analysis.

3. And receiving alarm information (namely, a control result) generated after the control by the control system through an alarm Kafka theme queue of the corresponding large library control task and carrying out subsequent processing.

4. The control state information table of the control object of the corresponding control system can be read to confirm the actual control state of the control object.

And the cloth control system comprises:

1. and (3) periodically (for example, synchronizing once every 12 hours) synchronizing the data of the 'control information table of the control object' of the preset gallery, carrying out local control, writing actual control information into the 'control state information table of the control object' corresponding to the local control system after the control is successful, and confirming that the corresponding control object is successfully controlled.

2. And acquiring face real-time acquisition data corresponding to the control system provided by the view library, and performing control comparison analysis.

3. And if the comparison triggers the alarm, sending alarm information to the view library through an alarm Kafka theme queue of the corresponding large library management and control task.

A multi-algorithm management and dispatch engine system:

1. and (5) monitoring and checking the view library synchronous 'the control information table of the control object'.

2. And supervising and checking the control situation of the control object and a corresponding control system.

3. And supervising and checking the generation and receiving conditions of the alarm information of the control object.

In embodiment 2, as shown in fig. 8, each system of the gear gathering service mainly works as follows:

view library:

1. and providing a file gathering target information table, and updating and maintaining according to the service condition.

2. And providing corresponding face real-time acquisition data for a corresponding distribution control system to analyze and gather files.

3. And receiving file information generated by the distribution control system and track information of the file.

And the cloth control system comprises:

1. the data of the file-gathering target information table of the view library is synchronized periodically (e.g. every 12 hours).

2. And acquiring face real-time acquisition data corresponding to the control system provided by the view library, and carrying out local analysis, file gathering and real-name comparison processing.

3. Registering the new archive with the view library and transmitting new archive track data.

4. If the file closing operation exists, a corresponding operation of withdrawing files to be withdrawn is initiated to the view library.

A multi-algorithm management and dispatch engine system:

1. and (5) supervising and checking the data of the view library synchronous file gathering target information table.

2. And supervising and checking the data acquisition condition of the file gathering target information table and the corresponding distribution control system.

3. And supervising and checking the generation and receiving conditions of the archives and the track data.

In embodiment 3, as shown in fig. 9, the systems of the graph searching service mainly work as follows:

view library:

1. and initiating conventional business instructions such as graph searching and the like to a multi-algorithm management and dispatch engine system through an interface.

And the cloth control system comprises:

1. and receiving conventional business instructions such as graph searching and the like issued by the multi-algorithm management and scheduling engine system, and returning an execution result to the multi-algorithm management and scheduling engine system.

A multi-algorithm management and dispatch engine system:

1. receiving conventional business instructions such as graph searching and the like issued by a view graph library, and forwarding the instructions to a corresponding distribution control system according to the requirements of the instructions.

2. Receiving a result returned after the control system executes the instruction, and returning the result to the view library after integrating the result

3. And (3) supervising and checking the instruction receiving and transmitting process of the view library and the distribution control system and the transmitting and receiving conditions of the result data.

The invention can normalize a plurality of distributed control systems through a multi-algorithm management and dispatch engine system effectively through standard specification, shield the difference of the distributed control systems and effectively solve the complexity at the application level. Secondly, the invention can effectively manage the display of the current service real-time condition of each control system, can effectively display the current service condition in front of the user through the management platform, and can directly check the current service condition through page analysis. Thirdly, the invention can effectively check whether each control system meets the standard, and give an alarm in time under the condition that the algorithm effect and performance of the control system do not meet the standard, and arrange related personnel to check and solve the problem. Fourth, the invention realizes the task arrangement of the control system, analyzes the content according to the protocol specification, judges which control system meets the requirement at present, and accurately transmits the information to the control system meeting the requirement, thereby solving the problems of resource waste, low execution efficiency of each control system and the like, and improving the real-time transmission of data streams.

In this embodiment, a device for controlling and managing is further provided, and the device is used for implementing the foregoing embodiments and preferred embodiments, and is not described in detail. As used below, the term "module" may be a combination of software and/or hardware that implements a predetermined function. While the means described in the following embodiments are preferably implemented in software, implementation in hardware, or a combination of software and hardware, is also possible and contemplated.

The present embodiment provides a distributed management apparatus, as shown in fig. 10, including:

the task information comprises a target task area and a target task object;

a distribution control system selection module 402, configured to determine a target distribution control system from the accessed distribution control systems based on the target distribution control area;

the big library task selection module 403 is configured to obtain a big library task corresponding to the target control task in the target control system if the target control area is greater than a preset range, where the big library control task is configured to control a plurality of control objects in the same target control area;

And the control object control module 404 is configured to write the target control object into the large library control task, so as to perform control on the target control object to obtain a control result.

In some alternative embodiments, the apparatus further comprises:

and the small library task selection module is used for issuing the target control task and the target control object thereof to the target control system if the target control area is smaller than or equal to the preset range so as to control the target control object to obtain a control result.

In some alternative embodiments, the apparatus further comprises:

the control result receiving module is used for acquiring a target task code corresponding to a target control result when receiving the target control result fed back by any accessed control system;

the control object determining module is used for obtaining a control object corresponding to the target control result when the target task code is the task code of the large library control task;

the control task query module is used for determining a control task corresponding to the target control result and a task state of the control task corresponding to the target control result from prestored control tasks corresponding to the large library control task according to the control object corresponding to the target control result;

And the first result feedback module is used for sending the target control result to a corresponding service system according to the control task corresponding to the target control result when the task state meets the preset effective control condition.

In some alternative embodiments, the apparatus further comprises:

and the second result feedback module is used for determining a task to be controlled corresponding to the target task to be controlled according to the target task code when the target task code is not the task code of the large library task to be controlled so as to send the target task to the corresponding service system.

In some alternative embodiments, the apparatus further comprises: the system checking module is used for checking the distribution control system; wherein, the control system examination module includes:

the assessment task acquisition unit is used for acquiring a preset system assessment task, wherein the system assessment task comprises at least one assessment subtask;

the system comprises a distribution control system assessment unit, a task assessment unit and a task assessment unit, wherein the distribution control system assessment unit is used for assessing the accessed distribution control systems based on the assessment subtasks to obtain task assessment results of the accessed distribution control systems corresponding to the assessment subtasks;

the assessment result fusion unit is used for fusing task assessment results of all the assessment subtasks corresponding to each accessed distributed control system to obtain system assessment results of the accessed distributed control systems;

And the system checking and alarming unit is used for generating corresponding system alarming information when the system checking result is that the distributed control system does not reach the standard.

In some alternative embodiments, the control system assessment unit includes:

the task analysis subunit is used for analyzing the assessment subtask and determining a task target;

the task processing subunit is used for acquiring a processing result sample and a target processing result corresponding to each accessed control system based on the task target;

and the assessment sub-unit is used for obtaining task assessment results of each accessed control system corresponding to the assessment sub-task based on the target processing results of each accessed control system and the corresponding processing result samples.

In some alternative embodiments, the task processing subunit is specifically configured to:

Further functional descriptions of the above respective modules and units are the same as those of the above corresponding embodiments, and are not repeated here.

The administration device in this embodiment is presented in the form of functional units, where the units are ASIC (Application Specific Integrated Circuit ) circuits, processors and memories executing one or more software or fixed programs, and/or other devices that can provide the above described functions.

The embodiment of the invention also provides a management platform with the management device shown in the figure 10. The distributed control management platform comprises:

the multi-algorithm management and dispatch engine system is connected with the view library and is used for executing the control management method of any embodiment;

The embodiments of the present invention also provide a computer readable storage medium, and the method according to the embodiments of the present invention described above may be implemented in hardware, firmware, or as a computer code which may be recorded on a storage medium, or as original stored in a remote storage medium or a non-transitory machine readable storage medium downloaded through a network and to be stored in a local storage medium, so that the method described herein may be stored on such software process on a storage medium using a general purpose computer, a special purpose processor, or programmable or special purpose hardware. The storage medium can be a magnetic disk, an optical disk, a read-only memory, a random access memory, a flash memory, a hard disk, a solid state disk or the like; further, the storage medium may also comprise a combination of memories of the kind described above. It will be appreciated that a computer, processor, microprocessor controller or programmable hardware includes a storage element that can store or receive software or computer code that, when accessed and executed by the computer, processor or hardware, implements the methods illustrated by the above embodiments.

Although embodiments of the present invention have been described in connection with the accompanying drawings, various modifications and variations may be made by those skilled in the art without departing from the spirit and scope of the invention, and such modifications and variations fall within the scope of the invention as defined by the appended claims.

Claims

1. A method of administration, the method comprising:

2. The method according to claim 1, wherein the method further comprises:

3. The method according to claim 2, wherein the method further comprises:

4. A method according to claim 3, characterized in that the method further comprises:

5. The method according to claim 1, wherein the method further comprises:

6. The method of claim 5, wherein the evaluating the accessed distributed control systems based on the evaluating subtasks to obtain task evaluating results of each accessed distributed control system corresponding to each evaluating subtask, comprises:

analyzing the assessment subtasks to determine task targets;

7. The method of claim 6, wherein the obtaining, based on the task targets, the processing result samples and the target processing results corresponding to each of the accessed distributed systems includes:

8. A cloth control management apparatus, the apparatus comprising:

9. A distributed management platform, comprising:

a multi-algorithm management and dispatch engine system coupled to the view library for performing the administration management method of any one of claims 1 to 7;

10. A computer-readable storage medium having stored thereon computer instructions for causing a computer to execute the method of the management and control of any one of claims 1 to 7.