CN114757797B - A data model-driven middle-office architecture method for power grid resource business - Google Patents

Abstract

The invention proposes a middle-end architecture method for power grid resource business based on data model driving, which includes: extracting real-time business requirements to obtain core entities and application entities; establishing a unified data model for the core entities; The entity builds a shared service center and determines the service link between the shared service centers; performs historical service portraits according to the historical service indicators of the shared service centers in the service links, and conducts real-time service portraits according to the real-time service indicators of all shared service centers to generate pending service portraits. List of elastic resource allocation services; perform service matching on the list of services to be allocated for elastic resources, update the resource pool, and dynamically optimize the architecture of the grid resource business center based on the elastic allocation strategy. The invention realizes the dynamic elastic resource allocation of the power grid resource business middle station through the service profiling means combining historical and real-time service index characteristics and the service characteristics, and improves the efficiency of providing application services.

Description

A data model-driven middle-office architecture method for power grid resource business

technical field

The invention belongs to the field of big data platforms, and in particular relates to a middle-end architecture method for power grid resource services driven by a data model.

Background technique

The power grid data model is the basis for the realization of power grid informatization and digitization, the core support for changing business models and promoting power grid technology upgrades, and the basic premise for promoting equipment management, equipment operation and maintenance, and equipment maintenance to participate in energy informatization and virtualization. In order to change the mode of grid resource business, promote grid technology upgrade, and promote grid resources, equipment management, equipment operation and maintenance, and equipment maintenance to participate in energy informatization and virtualization, it is urgent to build a unified grid resource business platform to integrate grid resource business. data to eliminate data maintenance silos and break down data access barriers.

The power grid resource business middle platform is an aggregate with standardized model data, standardized service interfaces, and basic common business capabilities. It is a new infrastructure platform that is different from traditional monolithic business systems. In order to give full play to the data benefits of Taichung in the power grid resource business, Taichung in the power grid resource business integrates a large number of services related to power grid resources, equipment, and business, such as management services for power grid resources and equipment, real-time detection services for power grid operations, and power grid operations. Topological dynamic image sharing services, etc., due to the complexity of grid-oriented services, the peak and low peak periods of different services are different and the real-time changes are large, which increases the difficulty of resource allocation in the grid resource business center, and the existing resource allocation strategies are mostly Uniform allocation only based on the current resource situation is difficult to apply to the power grid resource business middle-end with complex changes in service resources and diverse service characteristics.

SUMMARY OF THE INVENTION

In terms of the construction of the power grid business middle platform, at this stage, most of the business departments maintain their own data platforms independently. Due to the inconsistency in modeling methods and data format definitions, it is more difficult to build a power grid business middle platform. In order to To solve the above-mentioned shortcomings and deficiencies, the present invention proposes a data model-driven grid resource business middle-office architecture method, including:

S100: Obtain real-time business requirements, perform entity extraction on the real-time business requirements, and obtain core-type entities and application-type entities in the real-time business requirements;

S200: Establish a unified data model for the power grid operation logic, the full life cycle of power equipment, and the operation, maintenance and repair process of the core entities, where the unified data model includes a power grid resource model, an equipment asset model, and an operation and maintenance model;

S300 : constructing a shared service center in a power grid resource business middle-office based on an application entity, determining a service link between the shared service centers corresponding to the unified data model, and forming a power grid resource business middle-office architecture through the service link;

S400: Perform historical service portraits according to the historical service indicators of the shared service center in the service link, perform real-time service portraits according to the real-time service indicators of all shared service centers in the power grid resource business, and dynamically generate the historical service portraits and real-time service portraits List of services to be allocated for flexible resource allocation;

S500: Based on the service characteristics of the service link, perform service matching on the service list to be allocated for elastic resources, update the resource pool according to the service matching result, and dynamically optimize the architecture of the grid resource service middle station through the resource pool based on the elastic allocation strategy.

Optionally, the S200 includes:

Establish a power grid resource model according to the power grid operation logic of the core class entity, and the power grid resource model includes a logic model of the primary equipment of the power grid, the network topology of the secondary equipment and the electrical parameters;

Establish an equipment asset model according to the full life cycle of the power equipment of the core entity, and the equipment asset model includes the life stage model of equipment procurement, storage, commissioning, maintenance, overhaul, and decommissioning;

An operation and maintenance maintenance model is established according to the operation and maintenance maintenance process of the core class entities, and the operation and maintenance maintenance model includes the process models of maintenance tool management, maintenance plan management, operation and maintenance plan management, acceptance management and equipment scrap management.

Optionally, the shared service center includes a first-level center for maintaining grid static ledger information, a second-level center corresponding to application entities for dynamically managing middle-office data, and a third-level center for power grid business expansion applications. center, wherein the first-level center includes a power grid asset center, a power grid resource center, a power grid topology center, and an operation resource center, and the second-level center includes a model management center, a measurement point management center, a metering application center, and a power grid environment center. The above-mentioned three-level centers include power grid graphics center, equipment status center, power grid analysis center and operation management center.

Optionally, the S300 includes:

For the power grid resource model, the service links include that the power grid resource center establishes service links with the power grid graphics center, the power grid analysis center, and the job management center, respectively, and the power grid topology center and the The power grid graphics center, the equipment status center and the power grid analysis center establish a service link;

For the equipment asset model, the service link includes that the power grid asset center establishes a service link with the equipment status center and the operation management center respectively;

For the operation and maintenance maintenance model, the service link includes establishing a service link between the operation resource center and the operation management center.

Optionally, the S400 includes:

S410: Input the historical service indicators into different time series prediction models respectively to obtain the peak period prediction interval of the service link, and generate the peak period service portrait and the low peak period service according to the service conditions of the service link in the peak period prediction interval Prediction and recommendation results of portraits;

S420: According to the real-time service indicators, perform real-time detection of peak and low-peak periods on the services of all shared service centers in Taichung in the power grid resource business, and obtain real-time recommendation results of peak-period service profiles and low-peak period service profiles;

S430: Filter and merge the predicted recommendation result and the real-time recommendation result to obtain the recommendation service during the peak period and the recommendation service during the low peak period;

S440: Filter out the services that do not need to be expanded in the peak period recommendation service and the services that do not need to be scaled in the low-peak period recommendation service, and generate a waiting list from the filtered peak period recommendation service and low-peak period recommendation service. A list of elastic resource allocation services.

Optionally, the S410 includes:

Input the historical service indicators into several different time series prediction models, and obtain the prediction results of the median value of several peak periods;

A weighted average process is performed on the prediction result to obtain a reference time value, and the peak period prediction interval of the service link is calculated in combination with the average duration and time offset of the service peak period of the power grid resource service middle station.

Optionally, the peak period prediction interval is [recomm_time - average_time/2 - offset_time/2, recomm_time + average_time/2 + offset_time/2], where recomm_time represents the reference time value, average_time represents the average service peak duration, offset_time Represents a time offset.

Optionally, the S420 includes:

The real-time operation characteristics of each service are acquired based on a preset period, and if the number of times the real-time operation characteristics continuously exceed the preset threshold reaches the preset value, the service is marked as a peak period service, otherwise it is marked as a low-peak period service.

Optionally, the S500 includes:

Distinguish peak recommended services and low peak recommended services in the list of services to be allocated for elastic resources, determine the service characteristics of peak recommended services, and match peak recommended services and low peak recommended services based on service characteristics to obtain the same services. Featured low-peak recommended services;

Recycle the idle resources of the low-peak recommended service obtained by matching to the resource pool, and allocate the idle resources in the resource pool to the peak-time recommended service by generating an elastic allocation strategy;

Wherein, the service characteristics include CPU-intensive, IO-intensive, and GPU-intensive.

Optionally, the historical service indicators and real-time service indicators include the number of IO operations of the service, CPU utilization, network delay, real-time QPS, preset QPS threshold, average request delay, dependent services, and average time-consuming threshold.

The beneficial effects brought by the technical scheme provided by the invention are:

This application is designed for the top-level framework of the data model for the power grid resource business center, and integrates the power grid operation logic of the power grid, the full life cycle of power equipment, and the operation and maintenance process in a model-driven form into the power grid resource business center. Service architecture, While breaking the barriers of data access and application, it also takes into account the user-oriented application services of the power grid resource business center, and adds a shared service center to the grid resource business center structure.

In order to solve the problem of service resource allocation caused by the overly complex architecture in the power grid resource business, this application uses a service profiling method that combines historical and real-time service index characteristics, and combines the service characteristics of each shared service center in the power grid resource business. The dynamic and elastic resource allocation of the power grid resource business middle station is realized, and the efficiency of the application service provided by the power grid resource business middle station is improved.

Description of drawings

In order to illustrate the technical solutions of the present invention more clearly, the following briefly introduces the accompanying drawings used in the description of the embodiments. Obviously, the accompanying drawings in the following description are only some embodiments of the present invention, which are of great significance to the art For those of ordinary skill, other drawings can also be obtained from these drawings without any creative effort.

1 is a schematic flowchart of a data model-driven grid resource service middle-end architecture method proposed by an embodiment of the present invention;

Figure 2 is a schematic diagram of a core class entity and an application class entity;

FIG. 3 is a schematic diagram of service links between shared service centers.

Detailed ways

In order to make the purposes, technical solutions and advantages of the embodiments of the present invention clearer, the technical solutions in 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. Obviously, the described embodiments It is only a part of the embodiments of the present invention, but not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative efforts shall fall within the protection scope of the present invention.

It should be understood that, in various embodiments of the present invention, the size of the sequence numbers of each process does not mean the sequence of execution, and the execution sequence of each process should be determined by its functions and internal logic, and should not be used in the embodiments of the present invention. Implementation constitutes any limitation.

The technical solutions of the present invention will be described in detail below with specific examples. The following specific embodiments may be combined with each other, and the same or similar concepts or processes may not be repeated in some embodiments.

As shown in FIG. 1 , this embodiment proposes a data model-driven grid resource service middle-end architecture method, including:

S100: Obtain real-time business requirements, perform entity extraction on the real-time business requirements, and obtain core-type entities and application-type entities in the real-time business requirements;

S200: Establish a unified data model for the power grid operation logic, the full life cycle of power equipment, and the operation, maintenance and repair process of the core entities, where the unified data model includes a power grid resource model, an equipment asset model, and an operation and maintenance model;

S300 : constructing a shared service center in a power grid resource business middle-office based on an application entity, determining a service link between the shared service centers corresponding to the unified data model, and forming a power grid resource business middle-office architecture through the service link;

S400: Obtain historical service indicators of the shared service centers in the service link, perform service portraits based on real-time detection of peak and low-peak periods of all shared service centers in Taizhong in the power grid resource business, and dynamically generate resources to be elasticized according to the results of the service portraits Assign a list of services;

S500: Based on the service characteristics of the service link, perform service matching on the service list to be allocated for elastic resources, update the resource pool according to the service matching result, and dynamically optimize the architecture of the grid resource service middle station through the resource pool based on the elastic allocation strategy.

In this embodiment, the power grid business middle platform is a centralized Internet data processing platform that integrates standardized model data, standardized service interfaces, and basic common business capabilities, and is a new infrastructure platform that is different from traditional monolithic business systems. In the power grid business center, the power grid business can be integrated into a unified and common data platform in the form of shared applications, so as to eliminate the power grid business data island and break the data access barriers.

From the perspective of unified data modeling, this embodiment proposes a unified modeling method for grid resource business middle-end, breaks through the integration barriers of grid resource data, equipment assets, and measurement and collection data, eliminates data islands, and promotes data quality improvement. Take advantage of data efficiency. At the same time, by setting up a series of shared application centers, it promotes comprehensive equipment description specifications, accurate and reasonable grid connection, clear data organization logic, dynamic sharing of graphic services, and promotes multi-professional collaboration, full-service integration, and comprehensively enhances the digital center of the power grid.

In this embodiment, the core class entities and the application class entities are shown in FIG. 2 , wherein the core class entities are basic business elements of the power grid, including:

(1) Power System Resources, such as power grid unit modules such as generator sets, transmission lines, substations and auxiliary equipment, DC lines of converter stations, and power grid topology;

(2) Equipment assets (Asset), such as power transmission (overhead, cable, pipeline network), substation (primary and auxiliary equipment), power distribution, power consumption and other single power equipment;

(3) Transportation inspection business (Activity Record, Document), such as work plan and scheduling, process and process, results and reports and other grid-specific businesses.

The application entities are various data elements corresponding to the auxiliary services of the power grid, including:

(1) Measurement data (Measurement), such as real-time voltage, real-time current and other electrical measurement values;

(2) Location, such as asset location, weather location, etc.;

(3) Environmental Information, such as lightning, icing, wildfires, typhoons, geological disasters, etc.

In this embodiment, the power grid resources (Power System Resource), equipment assets (Asset), and inspection services (Activity Record, Document) are mainly used as the core, and measurement data (Measurement), location (Location), and environment (Environmental) are used as the core. Information) Introduction to associated core class entities, which are used to describe the influence of the electrical environment and the natural environment on the power grid.

In this embodiment, the unified data model is established respectively corresponding to the grid operation logic, the full life cycle of power equipment, and the operation and maintenance process of the three core entities, that is, the unified data model includes a grid resource model framework and an equipment asset model. Framework and O&M overhaul model framework. Wherein, the power grid operation logic of the core type entity includes the network topology of the primary equipment of the power grid, the network topology of the secondary equipment, and the logical relationship of electrical parameters, and the full life cycle of the power equipment of the core type entity includes equipment procurement, storage, investment The stage model of operation, maintenance, overhaul and decommissioning, the operation and maintenance overhaul process of the core class entity includes overhaul tool management, overhaul plan management, operation and maintenance plan management, acceptance management and equipment scrapping management.

It can be seen that in this embodiment, the unified data model corresponding to power grid resources is a logical relationship model, which is to logically model primary equipment and secondary devices of all voltage levels from the perspective of power grid operation, and focuses on describing power grid equipment. Functions, electrical parameters and network topology, the main contents include the container relationship, connection relationship, equipment electrical parameters, electrical measurement values, etc. of primary equipment in the power grid.

The unified data model corresponding to equipment assets is the equipment asset management stage model, which models assets from the perspective of a single physical entity, and reflects the entire life cycle processes of asset procurement, storage, commissioning, maintenance, overhaul, and decommissioning. The model establishes and removes the relationship through the business process of equipment commissioning and decommissioning and the power grid resource model. The main contents include the container relationship of assets, primary and secondary equipment asset parameters, equipment asset status, environmental conditions, and asset evaluation analysis models.

The unified data model corresponding to the operation and inspection business is the operation and maintenance process model, which is a unified modeling of the business objects involved in equipment management from the perspectives of power grid equipment operation and maintenance, maintenance, evaluation, and scrapping. Planning management, operation and maintenance management, maintenance management, inspection management, acceptance management, evaluation management, condition monitoring, scrap management, technical overhaul management, engineering management, etc.

On the basis of the above unified data model, this embodiment specifically expands the unified data model through a shared application center. For this reason, this embodiment decouples the power grid business functions into 12 shared application centers based on the above application entities. The relational structure of the primary center, secondary center, and tertiary center is shown in Figure 3. The primary center includes a power grid asset center, a power grid resource center, a power grid topology center, and an operation resource center. The secondary center It includes a model management center, a measurement point management center, a metering application center, and a power grid environment center. The three-level centers include a power grid graphics center, an equipment status center, a power grid analysis center, and an operation management center. The power grid asset center, power grid resource center, power grid topology center, and operation resource center mainly maintain the static ledger information of the power grid. The model management center realizes the management functions of various grid assets, power grid resources and operation and inspection business models of the center. The application center and the power grid environment center mainly maintain the dynamic collection data of the power grid. The power grid graphics center, equipment status center, power grid analysis center, and operation management center are mainly close to the power grid business. specific:

The power grid asset center focuses on physical assets such as power grid equipment, channel facilities, overhead line corridors, auxiliary systems, etc., with lean management as the main line, and supports operations, maintenance, maintenance, supplier performance and other business development. It provides functions such as new addition, capital transfer, spare parts management, depreciation, decommissioning and scrapping, inventory, information query and maintenance of power grid assets to the outside world; its core services include asset information query, asset life cycle management, and asset memorabilia management.

The power grid resource center aims to describe the objective power grid, by abstracting and maintaining the objects and attributes of primary equipment, auxiliary measurement equipment, terminal equipment, DC equipment, pumped storage, conventional hydropower, power generation equipment, and distributed power sources in the power grid. and relationship, to achieve unified maintenance of power grid resources, and orderly management of power grid resources at system level, substation level, feeder level, and station level. Its core services include grid resource query, format conversion, model mapping and maintenance functions. Its core services include resource query and resource management.

The power grid topology center realizes the management of the power grid topology network relationship under the multi-state power grid planning, construction, and operation by maintaining, analyzing and storing the power grid topology; , power supply radius analysis, large feeder and low-voltage station area topology analysis, connection line analysis, equipment tree analysis, inter-station connection analysis, station room diagram analysis and other functions.

The operation resource center realizes resource allocation, allocation, and ledger management based on operation resources such as vehicles, tools, instruments, operators, drones, and robots. Its core services include job resource query, statistics, additions, deletions, and changes, and positioning.

The model management center can meet the needs of power grid business and calculation analysis in real time by maintaining, updating and publishing metadata. Provide normalized certification in terms of model compliance, relevance, credibility, availability, connectivity, graph model consistency, and information interaction to ensure the standardization of power grid models. Its core services include metadata management, model checking, and version management.

The measuring point management center conducts operation and maintenance management on sensing measuring points such as electrical quantity information, equipment operating status information, operating environment information, topology identification information, etc. It can realize the access management of measuring point equipment, analysis and management of measuring point, monitoring and configuration management of measuring point equipment. Provide external measurement equipment, sensing equipment, remote sensing equipment and other sensing point ledger, measurement data, query and maintenance of abnormal events and other functions; its core services include measuring point ledger management, measurement data query, event subscription and Inquire.

The metering application center focuses on the development of metering related services, and realizes access management, configuration management of metering equipment, maintenance of the relationship between metering points and superior equipment, and alarm management of metering abnormal events. Its core services include real-time collection information of metering equipment, frozen collection information, and query of event information.

The power grid environment center integrates the forecast (alarm) data of environmental information such as lightning, icing, mountain fire, typhoon, and geological disasters that affect the safe operation of the power grid, so as to realize the analysis of natural disasters on power grid equipment and line channels. Its core services include grid environmental information query, monitoring and early warning, law analysis, risk assessment, and medium and long-term forecast analysis.

The power grid graphics center provides integrated, one-stop graphics services for shared application centers, cross-professional applications, and polymorphic business applications, and realizes flexible front-end display of graphics and applications, friendly human-computer interaction, and multi-dimensional analysis. Its core services include spatial query, GIS mapping, thematic mapping, and thematic mapping.

The equipment status center takes power grid equipment as the main body, and realizes multi-dimensional analysis and evaluation of equipment operation status based on asset information, resource information, topology information, equipment operation information, equipment ontology perception information, and environmental information. The equipment status center provides functions such as query of equipment operating status, multi-dimensional evaluation, early warning, and abnormal equipment list query.

The power grid analysis center realizes the whole network analysis from the power supply, the power grid to the users through the power grid diagram, measurement, operation, meteorological environment and other information. Its core services include grid-related security analysis, economic analysis, reliability analysis, operating state analysis, state estimation, and power flow analysis.

The operation management center focuses on power grid equipment, according to the operation standards, combined with the actual needs of the site, accurately controls the operation workflow and business logic, realizes the standardized management of various operations, and supports the execution of inspection, detection, fault repair, defect elimination, and maintenance plan execution , testing, live work and other grid operations. Its core services include defect analysis and inquiry, maintenance plan inquiry and maintenance, fault inquiry, inspection plan inquiry and maintenance, hidden danger analysis and inquiry, live work plan inquiry and maintenance, and test record inquiry and maintenance.

In this embodiment, the basic architecture of the grid resource service middle station is constructed through the above process, and based on the basic architecture, service links between shared service centers are formed according to business requirements, including:

For the power grid resource model, the service links include that the power grid resource center establishes service links with the power grid graphics center, the power grid analysis center, and the job management center, respectively, and the power grid topology center and the The power grid graphics center, the equipment status center and the power grid analysis center establish a service link;

For the equipment asset model, the service link includes that the power grid asset center establishes a service link with the equipment status center and the operation management center respectively;

For the operation and maintenance maintenance model, the service link includes establishing a service link between the operation resource center and the operation management center.

It should be noted that the interaction relationship between the shared application centers described in this embodiment is not limited to the relationship shown in FIG. 3 , and FIG. 3 only shows a definition of the interaction relationship in this embodiment. In other embodiments , as the unified data model is different, the definition of its interaction relationship is also different.

Based on the data model, this embodiment quickly obtains the grid resource service middle-end architecture for different service requirements, and abandons the drawbacks of repeated development, decentralized storage, and multiple interactions in the traditional monolithic application system. The data of various power grid resources such as low-voltage station area and users, distributed power supply and micro-grid can realize unified standards and homologous maintenance, and at the same time integrate common common services to the middle station to form basic and stable service capabilities, which can support various business applications Build flexibly and iterate quickly.

In this embodiment, in order to solve the problem of complex and changeable service resource allocation in the power grid resource business, the elastic allocation of resources is realized by performing a service portrait on the service link. In this embodiment, the main purpose of constructing the service profile is to detect and predict the peak period and the low peak period of the service. Since the services have different peak periods and low peak periods, the redundant resources of the services in the low peak period can be recovered and allocated to the services in other peak periods.

Specifically, the S400 includes:

S410: Input the historical service indicators into different time series prediction models respectively, obtain the peak period prediction interval of the service link, and generate the peak period service and the low peak period service according to the service situation of the service link in the peak period prediction interval Predict the recommendation result;

S420: Perform real-time detection of peak and low-peak periods on all services in the power grid resource business center, and obtain real-time recommendation results of peak-period services and low-peak services;

S430: Filter and merge the predicted recommendation result and the real-time recommendation result to obtain the recommendation service during the peak period and the recommendation service during the low peak period;

S440: Filter out the services that do not need to be expanded in the peak period recommended services and the services that do not need to be scaled in the low-peak period recommended services, and dynamically generate a service list to be elasticized resource allocation according to the filtering situation.

Wherein, the S410 includes:

Input the historical service indicators into several different time series prediction models, respectively, to obtain the prediction results of the intermediate values of several peak periods; in this embodiment, the time series prediction models include models such as Auto-ARIMA, Neural_Prophet, and Exponential Smoothing;

Perform weighted average processing on the prediction results to obtain the reference time value, and calculate the peak period prediction interval of the service link in combination with the average service peak period duration and time offset of the power grid resource service middle station, and the peak period prediction interval is: [recomm_time - average_time/2 - offset_time/2, recomm_time + average_time/2 + offset_time/2], where recomm_time represents the reference time value, average_time represents the average duration of the service peak period, and offset_time represents the time offset.

The S420 includes:

The real-time operation characteristics of each service are acquired based on a preset period, and if the number of times the real-time operation characteristics continuously exceed the preset threshold reaches the preset value, the service is marked as a peak period service, otherwise it is marked as a low-peak period service.

In this embodiment, in order to prevent service jitter, S420 is executed once every 20s, and the cycle is repeated 2-4 times. If the real-time operation characteristic exceeds the preset threshold for 3 consecutive times, it is determined to be a peak period service.

In this embodiment, in order to accurately filter out the services that clearly require flexible resource allocation, the specific process of filtering and merging the predicted recommendation result and the real-time recommendation result in S430 is as follows:

Determine the peak period service H1 and the low peak period service L1 in the predicted recommendation result, and then determine the peak period service H2 and the low peak period service L2 in the real-time recommendation result. When filtering and merging, the intersection of H1 and H2 is used as the recommended service during the peak period. Since the services in the intersection have a high probability of being in the peak period both at present and in the future, it means that the resource shortage of these services is more serious and can be used as the peak period. Periodic recommendation service; the union of L1 and L2 is used as the low-peak period recommendation service, and the centralized service ensures that it is capable of scheduling idle resources. Through the above process, the peak period services that need to be allocated resources and the low-peak period services that can allocate resources can be accurately screened out in the service link. Compared with the conventional elastic resource allocation method, the present embodiment uses S400 to precisely limit the scope of services participating in the elastic allocation, which can be more suitable for the middle station of the power grid resource service whose service conditions change frequently.

In this embodiment, the list of services to be allocated elastic resources is periodically synchronized to Redis through a data synchronization service, and then Redis periodically sends tasks to trigger the elastic resource allocation and recovery execution engine of the grid resource service middle station to execute S500.

The S500 includes:

Distinguish the peak period recommendation service and the low peak period recommendation service in the service list to be elastic resource allocation, determine the service characteristics of the peak period recommendation service, and filter out the low peak period recommendation service with the same service characteristics;

Recycle the selected idle resources of recommended services during low-peak periods into the resource pool, and allocate the idle resources in the resource pool to recommended services during peak periods by generating an elastic allocation strategy;

Wherein, the service characteristics include CPU-intensive, IO-intensive, and GPU-intensive.

In this embodiment, the elastic allocation policy is configured based on the service characteristics of the service link. For example, the current service link is determined to be GPU-intensive because it involves the power grid graphics center in the third-level application. When performing elastic resource allocation, the elastic allocation strategy will preferentially select GPU-intensive low-peak recommended service services in the list of services to be elasticized resource allocation, and among these low-peak recommended service services, preferentially select services belonging to this service link. The idle resources of the service can not only ensure that the idle resources meet the service characteristics, but also maximize the efficiency of resource allocation.

In this embodiment, the elastic allocation strategy is configured in Json format, and the elastic allocation strategy is synchronized to Redis through a data synchronization service.

The serial numbers in the above-mentioned embodiments are only for description, and do not represent the order in which the components are assembled or used.

The above descriptions are only the embodiments of the present invention and are not intended to limit the present invention. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present invention shall be included in the protection scope of the present invention. Inside.

Claims (6)

1. A data model-driven power grid resource service central station architecture method is characterized by comprising the following steps:

s100: acquiring a real-time service demand, and performing entity extraction on the real-time service demand to obtain a core entity and an application entity in the real-time service demand;

s200: respectively establishing a unified data model aiming at the power grid operation logic, the full life cycle of the power equipment and the operation and maintenance overhaul process of the core entity, wherein the unified data model comprises a power grid resource model, an equipment asset model and an operation and maintenance overhaul model;

s300: establishing a shared service center in a power grid resource service central station based on the application entity, determining service links among the shared service centers corresponding to the unified data model, and forming a power grid resource service central station framework through the service links;

s400: performing historical service portrayal according to historical service indexes of the shared service centers in the service link, performing real-time service portrayal according to real-time service indexes of all the shared service centers in the power grid resource service middling, and dynamically generating a to-be-elastic resource distribution service list by combining the historical service portrayal and the real-time service portrayal;

s500: performing service matching on the service list to be elastically distributed based on the service characteristics of the service link, updating the resource pool according to the service matching result, and dynamically optimizing the architecture of the power grid resource service center platform through the resource pool based on the elastic distribution strategy;

the core class entity is a basic service element of the power grid and comprises the following steps: power grid resources, equipment assets, and operation and inspection services; the application entity is various data elements corresponding to auxiliary services of the power grid, and comprises the following steps: measuring data, position and environment;

the S200 includes:

establishing a power grid resource model according to the power grid operation logic of the core class entity, wherein the power grid resource model comprises a logic model of the network topology and the electrical parameters of primary equipment and secondary equipment of a power grid;

establishing an equipment asset model according to the whole life cycle of the power equipment of the core entity, wherein the equipment asset model comprises life stage models of equipment purchasing, warehousing, commissioning, maintenance, overhauling and decommissioning;

establishing an operation maintenance model according to the operation maintenance process of the core entity, wherein the operation maintenance model comprises a process model of maintenance tool management, maintenance plan management, operation maintenance plan management, acceptance check management and equipment scrapping management;

the S300 includes:

for the power grid resource model, the service links comprise that a power grid resource center respectively establishes service links with a power grid graphic center, a power grid analysis center and an operation management center, and a power grid topology center respectively establishes service links with the power grid graphic center, an equipment state center and the power grid analysis center;

for the equipment asset model, the service link comprises a power grid asset center, and the service link is respectively established with an equipment state center and an operation management center;

for the operation maintenance and overhaul model, the service link comprises a service link established by the operation resource center and the operation management center;

the S400 includes:

s410: respectively inputting historical service indexes into different time series prediction models to obtain a peak period prediction interval of a service link, and generating prediction recommendation results of a peak period service portrait and a low peak period service portrait according to service conditions of the service link in the peak period prediction interval;

s420: according to the real-time service index, performing peak-to-peak and low-to-peak real-time detection on the services of all the shared service centers in the power grid resource service center to obtain real-time recommendation results of a peak-to-peak service portrait and a low-to-peak service portrait;

s430: filtering and combining the predicted recommendation result and the real-time recommendation result to obtain a peak period recommendation service and a low peak period recommendation service;

s440: filtering services which do not need capacity expansion in the peak period recommendation services and services which do not need capacity reduction in the low peak period recommendation services respectively, and generating a list of services to be distributed with elastic resources by the filtered peak period recommendation services and the filtered low peak period recommendation services;

the S500 includes:

distinguishing a peak period recommendation service and a low peak period recommendation service in a to-be-elastic resource allocation service list, determining service characteristics of the peak period recommendation service, and matching the peak period recommendation service and the low peak period recommendation service based on service characteristics to obtain a low peak period recommendation service with the same service characteristics;

the idle resources of the low peak period recommended service obtained by matching are recycled into the resource pool, and the idle resources in the resource pool are allocated to the high peak period recommended service by generating an elastic allocation strategy;

wherein the service characteristics include CPU intensive, IO intensive, and GPU intensive.

2. The method according to claim 1, wherein the shared service center includes a primary center for maintaining static grid ledger information, a secondary center corresponding to the application entity for dynamically managing the substation data, and a tertiary center for grid service development application, wherein the primary center includes a grid asset center, a grid resource center, a grid topology center, and a work resource center, the secondary center includes a model management center, a measurement point management center, a metering application center, and a grid environment center, and the tertiary center includes a grid pattern center, an equipment state center, a grid analysis center, and a work management center.

3. The data model-driven grid resource service staging architecture method according to claim 1, wherein the step S410 includes:

respectively inputting the historical service indexes into a plurality of different time sequence prediction models to obtain prediction results of a plurality of peak period intermediate values;

and carrying out weighted average processing on the prediction result to obtain a reference time value, and calculating a peak period prediction interval of a service link by combining the average time length of the service peak period of the power grid resource service middlebox and the time offset.

4. The data model-driven power grid resource service center architecture method according to claim 3, wherein the peak prediction interval is [ recomm _ time-average _ time/2-offset _ time/2, recomm _ time + average _ time/2 + offset _ time/2], where recomm _ time represents a reference time value, average _ time represents an average time duration of the peak service period, and offset _ time represents a time offset.

5. The method according to claim 1, wherein the S420 includes:

and acquiring real-time service indexes of each service based on a preset period, if the times that the real-time service indexes continuously exceed a preset threshold reach a preset value, marking the service as a peak service, and otherwise, marking the service as a low peak service.

6. The data model-driven power grid resource service middleware framework method according to claim 1, wherein the historical service index and the real-time service index comprise IO operation times of service, CPU utilization rate, network delay, real-time QPS, QPS preset threshold, request average delay, dependent service and average time consumption threshold.

Images