KR20220095313A - Digital Twin Based Distributed Energy Resource and Power Line Management Plan System and Method thereof - Google Patents

Abstract

The present invention relates to an operation planning system and operation planning method for efficient operation and optimization. The system includes: a bigdata system for collecting various driving information from a distributed resource system; a machine learning system performing prediction for a power system and renewable generation through machine learning for the driving information; a digital twin system that can perform simulation like an actual environment by using the driving information and the prediction information; and a data link channel intermediating and managing communication for connection of all of the systems.

Description

Digital Twin Based Distributed Energy Resource and Power Line Management Plan System and Method thereof

The present invention relates to a distributed resource and power system operation planning system and operation planning method for efficient operation and optimization through digital twin-based simulation and linking distributed resources in a microgrid environment.

Digital twin technology refers to a digital virtual object implemented in a digital environment by replicating the exact same environment as the actual industrial site with software. It can mean a solution that collects data in real time and provides it to field operators, and verifies the effectiveness of technical review through simulation of complex design and operation that is difficult to confirm in the real environment.

Site operators can check the field operation status in real time through a digital twin that implements the actual field virtually, review optimization and feasibility for design and operation through simulation, and collect field operation data such as predicted failures or accidents. It has the advantage of being able to operate the site in an optimal state because it can respond immediately after checking through the digital twin.

The need for energy conservation is increasing due to various environmental factors such as social policies, business ethics, and economic feasibility. Therefore, companies, institutions, and factories with high power consumption introduce various energy saving facilities and demand management systems such as ESS (Energy Storage System), PV (Solar Power Generation System), and DR (Demand Management System) for energy saving. are doing

As the demand for energy saving increases, manufacturers and developers for each facility diversify, which increases management costs and makes it impossible to actively respond to system changes.

In order to efficiently integrate and manage the facilities and systems developed and built in various environments, an environment that can flexibly link each independent facility and system through a standardized interface is required. As a result, the need for measures to increase operational efficiency is emerging.

Accordingly, the present invention collects operation information on distributed resources such as energy saving facilities and management devices such as ESS, PV, DR, etc. according to the user's request, and analyzes the operation information of the collected distributed resources to predict the risk of distributed resources in advance. It is to provide a system and method for establishing a strategic operation plan through driving simulation and optimization of a complex environment desired by the user.

The solution means of the present invention is a big data system for collecting various driving information from a distributed resource system; a machine learning system for deriving predictive information on the electric power system and renewable power generation through machine learning on the driving information collected through the big data system; a digital twin system capable of simulating like a real environment through driving information of the distributed resource system and prediction information of the machine learning system; A distributed resource and power system operation planning system including a data interworking channel for the big data system, the machine learning system, and the digital twin system to communicate with each other and exchange information with the distributed resource system may be provided.

In addition, the present invention comprises the steps of collecting various driving information from a distributed resource system in a big data system; deriving prediction information on distributed resources and power systems through a machine learning system for driving information collected through the big data system; A distributed resource and power system operation planning method including the step of simulating like a real environment with a digital twin system through the operation information of the distributed resource system and the prediction information of the machine learning system may be provided.

The present invention is based on data on the operation status of industrial new and renewable energy facilities in a microgrid environment, through a digital twin-based system that is synchronized with actual new and renewable energy power generation facilities and ESS It is possible to provide information for the purpose of efficient maintenance of installed renewable energy generation resources and to increase operation efficiency.

The present invention can ensure high availability of a distributed resource system by loading and analyzing a large amount of driving information collected from various distributed resources in a big data system, and detecting the risk of distributed resources in advance based on the analyzed contents, It is possible to establish a strategic operation plan through driving simulation and optimization of the complex environment desired by the user.

1 is a block diagram showing a configuration for implementing the operation planning system of the present invention.
2 is a block diagram showing the configuration of a simulation unit of the present invention.
3 is a flowchart illustrating a simulation process for implementing the operation planning method of the present invention.

Hereinafter, specific contents for carrying out the present invention will be described in detail based on the accompanying exemplary drawings.

1 is a block diagram showing a configuration for implementing the operation planning system of the present invention. Referring to FIG. 1 , the operation planning system 10 of the present invention provides a big data system 200 for collecting various driving information from the distributed resource system 100 , the driving information collected through the big data system 200 . The machine learning system 300 for deriving prediction information about distributed resources and power system through machine learning about It may be configured to include a digital twin system 400 that can do this.

In addition, the management system 10 is a data interworking channel 500 for the big data system 200 , the machine learning system 300 , and the digital twin system 400 to communicate with the distributed resource system 100 and exchange information with each other. may include

The data interworking channel 500 may be for efficiently integrating and managing various and heterogeneous distributed resources. When a request and response between systems is required as needed, data frequently occurs and needs to be delivered to multiple systems, data is not frequently generated but needs to be delivered to multiple systems, and the like.

In other words, the data interworking channel 500 may be provided to ensure scalability by implementing a simple and pluggable system.

The distributed resource system 100 includes an energy storage management system (ESS) 110 , a photovoltaic power generation management system (PVMS) 120 , a wind power generation management system (WMS) 130 , and a demand response management system as a renewable power generation system. (DRMS) 140 and may include at least one of the power demand management system (EMS) (150). Each distributed resource system 100 may independently operate an individual target facility or system, and may be loaded (stored) as independent data in the big data system 200 .

The energy storage management system 110 may be an energy storage device for charging and discharging power, and may be provided in one or a plurality.

The photovoltaic power generation management system 120 may be a photovoltaic management system that manages a photovoltaic power generation facility that generates electric power using solar energy as a renewable energy power generation module.

The wind power generation management system 130 may be a device for managing wind power.

The demand response management system 140 may be a system for providing and managing a demand response service to a user.

The demand response management system 140 may generate demand response information based on the user's power consumption information, and may provide a demand response service to the consumers based on the generated demand response information. Here, the demand response information may include a power consumption reduction request amount, an actual demand reduction amount, a reduction participation rate, an actual reduction participation rate, and the like.

The power system may be a load of each business site, etc.

The big data system 200 includes a driving information collection unit 210 that collects a large amount of driving information from each distributed resource system 100, a data preprocessing and loading unit 220 that preprocesses and loads the collected driving information data, It may include a data analysis unit 230 for analyzing the pre-processed and loaded driving information data, and a data providing API (Application Programming Interface) 240 for providing the analyzed data to the digital twin system 400 .

The machine learning system 300 learns about the operation information of each distributed resource system 100, and sets a machine learning model through this learning to determine the amount of power consumption, solar power generation, and wind power generation for the power system composed of various loads. It could be predictable.

The machine learning system 300 includes a failure and maintenance time prediction unit 310 for predicting failure and maintenance times for each system constituting the distributed resource system 100 , a power consumption prediction unit 320 for each power system, and solar The solar power generation amount prediction unit 330 for predicting the amount of power generation of the photovoltaic facility may be configured to include at least one or more of the wind power generation amount prediction unit 340 for predicting the generation amount of the wind power generation facility.

The failure and inspection time prediction unit 310 of the machine learning system 300 predicts the failure prediction and inspection time for the facility through the prediction of the expected lifespan of the PCS and battery and the analysis of the number of occurrences of temperature and voltage abnormalities of the battery cells. can be calculated.

The machine learning system 300 may receive and learn pattern classification and recognition information on the amount of power consumption, set the machine learning model, and then predict the amount of power consumption in the future through the power consumption prediction unit 320 .

Looking at the learning contents in the machine learning system 300, the power usage pattern classification is, for example, classifying load patterns having the same load increase/decrease among a plurality of load patterns, extracting a representative pattern from these classified patterns, and representing After determining whether the pattern exceeds the set error rate, if it does not exceed the set error rate, it is calculated as a representative pattern, and if it exceeds the set error rate, the reclassification operation can be repeatedly performed.

Accordingly, the machine learning system 300 learns to set the machine learning model, and then the power consumption prediction unit 320 may predict the future power consumption based on the machine learning model.

The photovoltaic power generation amount prediction unit 330 may predict the photovoltaic power generation amount through data on the photovoltaic power generation information of the photovoltaic power generation management system 120 .

The wind power generation amount prediction unit 340 may predict the wind power generation amount through data on the wind power generation information of the wind power generation management system 130 .

The digital twin system 400 may receive information from the big data system 200 and the machine learning system 300 to perform simulations and derive results.

The digital twin system 400 collects driving information of the simulation setting unit 410 that sets the conditions for the object to be simulated, and the big data system 200 that collects data on driving information of each distributed resource system 100 . The driving information inquiry unit 420 that inquires the driving information data through the unit 210, the prediction information inquiry unit 430 that inquires the prediction information by the machine learning system 300 based on the inquiry driving information, performs a simulation The simulation unit 440 may include a simulation unit 440 and a simulation result transmission unit 450 for transmitting the simulation result by the simulation unit 440 to the user.

2 is a block diagram showing the configuration of a simulation unit of the present invention.

Referring to FIG. 2 , the simulation unit 440 classifies, for example, a power information request unit 4401 requesting power information from the power demand management system (xEMS) 150 , and a pattern for the obtained power information. Recognized classification recognition unit 442, solar power generation request unit 443, solar power generation prediction information unit 444, charge/discharge amount calculation unit 445, ESS control request unit 446, ESS control result notification unit 447 ), an ESS control result inquiry unit 448 , and an event information notification unit 449 for notifying event information.

The actual power demand management system 150 includes an ESS control command unit that receives a request from the ESS control request unit 446, an ESS information collection unit that collects data information, a usage information unit that receives power information from the power information request unit 441, etc. can be made with

The actual energy storage management system 110 provides the operation information on the ESS operation information sharing unit to the ESS control result notification unit 447 through the ESS control unit that controls the ESS through the ESS control command unit, and the data interworking channel 500 . It may consist of an event sharing unit that stores event information such as a driving information sharing unit and an ESS device abnormality and provides it to the event information notification unit 449 .

The actual photovoltaic power generation management system 120 may include a photovoltaic power generation information unit and an event sharing unit that shares events such as abnormalities in the photovoltaic device.

Real-time information such as charge/discharge amount, power usage, and solar power generation amount generated during simulation by the simulation unit 440 is transmitted through the same path and interface as the real-time interlocking of the actual digital twin system 400, and higher-level control such as PMS The system can transmit minute-by-minute data in seconds.

The results of the simulation unit 440 can be viewed through the screen unit, and it is possible to check the maximum load comparison before and after the simulation, the electricity consumption rate and benefits and economic feasibility analysis, and the change in the power consumption due to ESS charging and discharging and PV power generation.

It is possible to analyze the operation of the facility by using the simulation result, and calculate the inspection time through failure prediction based on the lifespan and condition of major parts.

3 is a flowchart showing a method for performing a simulation by the digital twin system 400 of the present invention.

Referring to FIG. 3 , the method for performing a simulation of the present invention may include setting an object and a range to be simulated through the simulation setting unit 410 and then executing the simulation.

When setting the simulation, initializing the simulation and selecting the simulation period and load, selecting a distributed resource through the user's custom data, determining whether the selected distributed resource is a solar system, a solar system case, setting the solar panel inverter capacity, if not a solar system, determining whether it is a wind power system, if it is a wind system, setting the inverter capacity of the wind turbine, if not a wind power system, whether it is an ESS In the case of ESS, setting the battery and PCS capacity and operation algorithm, in the case of non-ESS, determining whether the system is a demand response system, in the case of a demand response system, setting the reduction implementation rate and target reduction step, it may include the step of completing the simulation setting.

The custom data may be power data modified or generated by the user.

If the user custom data is not used, the method may include determining whether to use the demand response system through the distributed resources of the selected load customer. In the case of using the demand response system, set the reduction fulfillment rate and target reduction, and when not using the demand response system, complete the simulation setting and run the simulation without setting the reduction implementation rate and target reduction as the next setting can do.

The simulation execution step is a step of inquiring information about the simulation setting, inquiring distributed resource operation information and power information, inquiring the predicted power consumption, determining whether it is a solar system, and in the case of a solar system, the predicted solar Inquiring the amount of photovoltaic power generation, failure prediction and inspection timing, determining whether it is a wind power system if it is not a solar system, inquiring the predicted wind power generation amount, failure prediction and inspection time in case of a wind power system, In the case of ESS, determining whether or not the ESS is an ESS, determining whether to perform the algorithm in the case of ESS, and in the case of performing the algorithm, perform the peak cut charge/discharge algorithm after performing the algorithm, calculate the amount of charge and discharge, and calculate the amount of power consumption It may include the step of calculating, then transmitting the simulation result and completing the simulation execution.

If it is not the ESS, calculating the amount of power consumption, transmitting the simulation result, and completing the simulation execution may be included.

In addition, when the algorithm is not performed, the method may include calculating the amount of charge and discharge without performing the peak-cut charge/discharge algorithm, calculate the amount of power consumption, transmit the simulation result, and complete the simulation execution.

The distributed resource and power system management system may be mounted on the management server, and the management server may show the simulation result through the screen unit.

10... Operational Planning System 100... Distributed Resource System
110... Energy storage management system 120... Solar power system
130... wind power management system 140... demand response management system
150... power demand management system
200... Big data system 210... Driving information collection unit
220... Data preprocessing and loading section 230... Data analysis section
240... Data providing API part
300... Machine learning system 310... Failure and maintenance time prediction unit
320... Power consumption prediction unit 330... Solar power generation prediction unit
340... Wind power generation forecasting unit 400... Digital twin system
410... Simulation setting unit 420... Driving information inquiry unit
430... Forecast information inquiry unit 440... Simulation unit
441... Power information request unit 442... Power information pattern classification recognition unit
443... Solar power generation information request unit 444... Solar power generation forecast information request unit
445... ESS charge/discharge calculation unit 446... ESS control request unit
447... ESS control result notification unit 448... ESS control result inquiry unit
449... Event information notification unit
450... Simulation result transmission unit 500... Data linkage channel

Claims (8)

Big data system for collecting various driving information from distributed resource system;
a machine learning system for deriving predictive information on the electric power system and renewable power generation through machine learning on the driving information collected through the big data system;
a digital twin system capable of simulating like a real environment through driving information of the distributed resource system and prediction information of the machine learning system;
a data interworking channel through which the big data system, the machine learning system, and the digital twin system communicate with each other to exchange information with the distributed resource system;
Distributed resource and power system operation planning system that includes.
According to claim 1,
The distributed resource system,
including at least one of an energy storage management system (ESS), a photovoltaic power generation management system (PVMS), a wind power generation management system, a demand response management system (DRMS), and an electric power demand management system (EMS),
The big data system is
A driving information collection unit that collects driving information of each distributed resource system, a data preprocessing and loading unit that pre-processes and loads the collected driving information data, a data analysis unit that analyzes the pre-processed and loaded driving information data, and digitally analyzes the analyzed data. A distributed resource and power grid operation planning system with data provisioning APIs to provide to the twin system.
According to claim 1,
The machine learning system is
A failure and inspection time prediction unit for predicting a failure and inspection time for the distributed resource system;
Power consumption prediction unit for each power system,
A solar power generation prediction unit that predicts the generation amount of the solar power generation facility,
A distributed resource and power system operation planning system comprising at least one of a wind power generation amount prediction unit for predicting the generation amount of a wind power generation facility.
According to claim 1,
The digital twin system is
A simulation setting unit that sets conditions for the object to be simulated;
a driving information inquiry unit that inquires about driving information data through the driving information collection unit of the big data system that has collected data on driving information of each distributed resource system;
Prediction information inquiry unit that inquires information predicted based on the inquired driving information;
a simulation unit that performs a simulation;
Distributed resource and power system operation planning system including a simulation result transmission unit for transmitting the simulation result by the simulation unit.
collecting various driving information from a distributed resource system in a big data system;
deriving prediction information on distributed resources and power systems through a machine learning system for driving information collected through the big data system;
simulating like a real environment with a digital twin system through the operation information of the distributed resource system and the prediction information of the machine learning system;
Distributed resource and power system operation planning method comprising a.
6. The method of claim 5,
The simulation step is
Setting the target and scope to be simulated through the simulation setting unit, and then executing the simulation;
Initializing the simulation and selecting the simulation duration and load when setting up the simulation;
selecting a distributed resource through the user's custom data;
determining whether the selected distributed resource is a solar system;
In the case of a solar system, setting the solar panel inverter capacity;
determining whether it is a wind system if not a solar system;
In the case of a wind system, setting the inverter capacity of the wind turbine;
If it is not a wind power system, determining whether it is an ESS;
In case of ESS, setting the battery and PCS capacity, operation algorithm,
If it is not an ESS, determining whether it is a demand response system;
In the case of a demand response system, a distributed resource and power system operation planning method comprising the steps of setting a reduction implementation rate and a target reduction amount, and completing simulation settings.
6. The method of claim 5,
The simulation step is
Setting the target and scope to be simulated through the simulation setting unit, and then executing the simulation;
The simulation execution steps are:
Inquiring information on simulation settings, inquiring distributed resource operation information and power information, inquiring predicted power consumption, determining whether it is a solar system,
In the case of a photovoltaic system, the step of inquiring the predicted solar power generation amount, failure prediction, and inspection time;
determining whether it is a wind system if not a solar system;
In the case of a wind power system, the step of inquiring the predicted wind power generation amount, failure prediction and inspection time;
If it is not a wind power system, determining whether it is an ESS;
In the case of ESS, the step of determining whether to perform the algorithm;
In the case of performing an algorithm, performing a peak cut charging/discharging algorithm after performing the algorithm, calculating the amount of charge and discharge, and calculating the amount of power consumption;
A method for planning distributed resource and power system operation, comprising transmitting simulation results and completing simulation execution.
8. The method of claim 7,
If it is not an ESS, calculating the amount of power consumption, transmitting the simulation result, and completing the simulation execution;
Distributed resource and power system operation plan, comprising the steps of calculating the amount of charge and discharge without performing the peak-cut charge-discharge algorithm when the algorithm is not performed, calculating the amount of power consumption, transmitting the simulation result, and completing the simulation execution Way.

Images