Multi-scene load real-time control method for real-time regulation and control requirements of power grid

Abstract

The invention relates to the field of load regulation and control, in particular to a multi-scene load real-time control method for a power grid real-time regulation and control demand, which comprises the following steps: acquiring energy consumption data of a user to be evaluated; according to the demand response target, establishing a user online energy consumption behavior model dynamically responding to the load aggregation control request; and inputting the to-be-evaluated user energy consumption data into a trained user online energy consumption behavior model, and outputting a user adjustable load adjustment ratio by the model. According to the invention, the dispatching center layer issues a demand response target to the load aggregation layer according to the power grid operation condition; according to a load aggregation layer, based on a dynamic response load aggregation control request of the user online energy consumption behavior model, an online algorithm is adopted, and the load aggregation control request is sent to the user; and the load terminal layer is used for collecting electric energy data, responding to the regulation and control instruction issued by the load aggregation layer in real time, and realizing real-time sensing and efficient regulation and control effects of the adjustable load.

Description

Technical Field

[0001] The present invention relates to the technical field of load regulation, and in particular to a multi-scenario load real-time control method oriented to real-time regulation requirements of power grids. Background Art

[0002] As the proportion of distributed resources connected to the grid increases, a large number of traditional units are retired, a large number of flexible loads on the demand side are connected in an unordered manner, and the peak load situation is becoming increasingly prominent. It is difficult for distributed power sources to track the load curve, and the phenomenon of wind and solar abandonment is serious. The grid regulation capacity is seriously insufficient, and safe and stable operation faces challenges. It is very easy to cause the grid to operate overloaded and even cause the grid to collapse, and the resulting large-scale blackouts occur frequently. Realizing demand response on the demand side is an important direction to improve the system regulation capacity, promote the consumption of new energy, and ensure the safety and stability of the system.

[0003] In the face of the new power system operation characteristics, it is urgent and necessary to incorporate demand-side adjustable load resources into the normalized power grid operation and control, promote the transformation of the power grid from "source follows load" to "source and load interact", and realize the evaluation of adjustable load regulation capacity. In addition, to realize the adjustable load's participation in power grid frequency regulation and achieve frequency regulation capabilities similar to those of traditional hydropower and thermal power units, it is necessary to build an efficient and real-time technical support system to enable the adjustable load to achieve rapid and accurate power regulation in a short time scale and respond quickly to frequency changes in the power system in seconds.

[0004] Therefore, those skilled in the art provide a multi-scenario load real-time control method oriented to the real-time regulation needs of the power grid to solve the problems raised in the above background technology. Summary of the invention

[0005] In order to solve the above technical problems, the present invention provides a multi-scenario load real-time control method for real-time power grid regulation needs, comprising the following steps:

[0006] Step S1: Obtaining energy consumption data of users to be evaluated, wherein the energy consumption data of users include user energy consumption behavior, uncertain factors and electricity prices;

[0007] Step S2: Establishing a user online energy consumption behavior model that dynamically responds to load aggregation control requests according to the demand response target;

[0008] Step S3: inputting the energy consumption data of the user to be evaluated into a trained user online energy consumption behavior model, and the model outputs a user adjustable load adjustment ratio.

[0009] Preferably, the user online energy usage behavior model is implemented on a combined online learning algorithm of user fatigue.

[0010] Preferably, the user online energy usage behavior model is trained by the following method:

[0011] Step 1: Obtain historical energy consumption data of the source domain and the target domain; construct a source domain adjustable load generalized model and a target domain adjustable load generalized model according to the user energy consumption behavior and the uncertain factors affecting the user energy consumption behavior in the source domain and the target domain data, respectively, and obtain the quadratic regression dynamic relationship between the user energy consumption behavior and multiple uncertain factors in the source domain and the target domain based on the source domain adjustable load generalized model and the target domain adjustable load generalized model respectively;

[0012] Step 2: Cluster the historical energy consumption data of the source domain and the target domain respectively to extract similar energy consumption behaviors among different users. According to the clustering results of the source domain and the target domain and the quadratic regression dynamic relationship between the energy consumption behaviors of the source domain and the target domain users and various uncertain factors, the quadratic regression parameters of the energy consumption behaviors of the source domain and the target domain users and the uncertain factors within the class are obtained respectively.

[0013] Step 3: Calculate the cosine similarity of the quadratic regression parameter characteristics of the target domain and source domain users’ energy consumption behaviors and uncertain factors within the class, and obtain the probability distribution of the adjustable load regulation capacity of the target domain based on the cosine similarity;

[0014] Step 4: Based on the user fatigue algorithm, the probability distribution of the adjustable load regulation capacity of the target domain is obtained, and the energy consumption behavior characteristics of the users in the source domain and the target domain are aligned according to the local maximum mean difference. The loss function is minimized by continuously training the model until the model converges.

[0015] Preferably, the control request for the probability of the target domain adjustable load regulation capability includes the following method:

[0016] Step 1) Inputting the user's fatigue coefficient function;

[0017] Step 2) Send a load aggregation control request to each user, and estimate the response probability of each user based on the observed performance;

[0018] Step 3) Estimate the maximum load power of all users in the area according to the load power size and sort them from large to small;

[0019] Step 4) A distribution of the probability of achieving the target domain adjustable load regulation capability according to the maximum load power.

[0020] The scheme also includes a power grid load aggregation collaborative control system, including a dispatch center layer, a load aggregation layer, and a load terminal layer;

[0021] The dispatch center layer issues demand response targets to the load aggregation layer according to the power grid operation status;

[0022] The load aggregation layer responds to the load aggregation control request dynamically based on the user's online energy consumption behavior model according to the demand response target issued by the dispatch center layer, and uses an online algorithm to send a load aggregation control request to the user;

[0023] The load terminal layer is used to collect electric energy data and respond in real time to the control instructions issued by the load aggregation layer.

[0024] In this solution, the dispatch center, based on the D5000 basic platform, builds a two-level load dispatch control system of sub-center and provincial dispatch; the dispatch center layer includes production control area I and management information area III;

[0025] The production control area I includes a load real-time control unit, which is used to implement functional modules such as load resource access, load control, and market interface, realize multi-level dispatching to collect, aggregate, monitor, and regulate adjustable load resources in real time, receive the clearing results of the spot and auxiliary service markets, and provide real-time and operating data to the load operation management function;

[0026] The management information III area includes a load operation management unit, which is used to implement modules such as load capacity evaluation, model management, information release and comprehensive display, realize model data sharing and comprehensive display of adjustable load resources between multi-level dispatching, and provide model data to the load real-time control function.

[0027] The load aggregation layer is realized by various load aggregation platforms, such as the smart energy service platform within the power grid company, the smart Internet of Vehicles platform, and the social third-party aggregation platform. However, the above platforms are all deployed in Area IV or the Internet. When interacting with the source-grid-load-storage coordinated control system in Area I for cross-area data and control command interaction, there are many transmission links and low interaction efficiency. In terms of adjustment time, it is still at the minute level and cannot support fast response application scenarios such as frequency modulation.

[0028] The load aggregation layer includes a multi-load aggregation energy router and a multi-load aggregation energy controller;

[0029] The multi-load aggregation energy router supports real-time optimization scheduling and control of adjustable loads, collects and aggregates massive discrete load resources in scenarios such as commercial buildings, energy storage, charging piles and industrial enterprises, and connects to industrial DCS, building BA system, energy storage PCS system and multi-load control terminal. It is compatible with multiple device physical interfaces, supports multi-protocol access function, has flexible load adjustable capacity evaluation function, predicts base value load, calculates upper and lower limits of load adjustable amount, and prepares load control plan and formulates load control strategy.

[0030] The multi-load aggregation energy controller realizes the aggregation of adjustable single-unit resources of different types and characteristics in the regional power grid and the optimized control within the region, responds to the load control instructions and data acquisition requirements of the dispatching center layer, and solves the problems of a large number of resources, diverse types, and small capacity of some single units.

[0031] The load terminal layer supports multi-protocol and multi-scenario applications, realizes the interconnection and information perception of different types of energy-consuming equipment, monitoring meters, and sensor equipment in commercial buildings, energy storage, charging piles, and industrial enterprises, and has the functions of energy data collection, processing, storage, and control. The load terminal layer includes a multi-load control terminal, which adopts a modular design and has an encryption module to block security risks from the terminal acquisition control network, ensure secure interconnection and data interaction with the dispatching access network, and perceive the operation data of the electric pile in real time, including the current operating status, current load, charging time, SOC, upward adjustment margin, downward adjustment margin, etc. At the same time, it can respond to the control instructions issued by the multi-load terminal in real time.

[0032] Technical effects and advantages of the present invention:

[0033] 1. Fast response speed, according to the principle of "direct purchase and direct control", to achieve real-time perception and efficient regulation of adjustable loads.

[0034] 2. Through load aggregation, the distributed adjustable load resources with low voltage level, small single capacity, and wide distribution are aggregated to reach a certain power scale and participate in grid frequency regulation.

[0035] 3. Most adjustable load resources do not have access conditions to the power dispatching data network. They need to communicate data with the dispatching end through the Internet. The load terminal layer is set up to block security risks from the terminal acquisition and control network, ensure secure interconnection and data interaction with the dispatching access network, and achieve data security access. BRIEF DESCRIPTION OF THE DRAWINGS

[0036] Figure 1 It is a system function diagram of the dispatch center layer provided in an embodiment of the present application;

[0037] Figure 2 This is an architectural diagram of a multiple load control terminal provided in an embodiment of the present application. DETAILED DESCRIPTION

[0038] The present invention is further described in detail below in conjunction with the accompanying drawings and specific embodiments. The embodiments of the present invention are provided for the purpose of illustration and description, and are not intended to be exhaustive or to limit the present invention to the disclosed forms. Many modifications and variations will be apparent to those of ordinary skill in the art. The embodiments are selected and described in order to better illustrate the principles and practical applications of the present invention, and to enable those of ordinary skill in the art to understand the present invention and thereby design various embodiments with various modifications suitable for specific uses.

[0039] Example 1

[0040] See also Figures 1-2 , in this embodiment, a power grid load aggregation collaborative control system is provided, including a dispatch center layer, a load aggregation layer, and a load terminal layer;

[0041] The dispatch center layer issues demand response targets to the load aggregation layer according to the power grid operation status;

[0042] The load aggregation layer responds to the load aggregation control request dynamically based on the user's online energy consumption behavior model according to the demand response target issued by the dispatch center layer, and uses an online algorithm to send a load aggregation control request to the user;

[0043] The load terminal layer is used to collect electric energy data and respond in real time to the control instructions issued by the load aggregation layer.

[0044] refer to Figure 1 The dispatch center, based on the D5000 basic platform, builds a two-level load dispatch control system of sub-center and provincial dispatch; the dispatch center layer includes production control area I and management information area III;

[0045] The production control area I includes a load real-time control unit, which is used to implement functional modules such as load resource access, load control, and market interface, realize multi-level dispatching to collect, aggregate, monitor, and regulate adjustable load resources in real time, receive the clearing results of the spot and auxiliary service markets, and provide real-time and operating data to the load operation management function;

[0046] The management information III area includes a load operation management unit, which is used to implement modules such as load capacity evaluation, model management, information release and comprehensive display, realize model data sharing and comprehensive display of adjustable load resources between multi-level dispatching, and provide model data to the load real-time control function.

[0047] The load aggregation layer is realized by various load aggregation platforms, such as the smart energy service platform within the power grid company, the smart Internet of Vehicles platform, and the social third-party aggregation platform. However, the above platforms are all deployed in Area IV or the Internet. When interacting with the source-grid-load-storage coordinated control system in Area I for cross-area data and control command interaction, there are many transmission links and low interaction efficiency. In terms of adjustment time, it is still at the minute level and cannot support fast response application scenarios such as frequency modulation.

[0048] In order to achieve the purpose of fast transmission, the load aggregation layer includes a multi-load aggregation energy router and a multi-load aggregation energy controller;

[0049] The multi-load aggregation energy router supports real-time optimization, scheduling and control of adjustable loads, collects and aggregates massive discrete load resources in scenarios such as commercial buildings, energy storage, charging piles and industrial enterprises, and connects to industrial DCS, building BA system, energy storage PCS system and multi-load control terminal. It is compatible with multiple device physical interfaces, supports multi-protocol access function, has flexible load adjustable capacity evaluation function, predicts base value load, calculates upper and lower limits of load adjustable amount, and prepares load control plan and formulates load control strategy.

[0050] The multi-load aggregation energy controller realizes the aggregation of adjustable single-unit resources of different types and characteristics in the regional power grid and the optimized control within the region, responds to the load control instructions and data acquisition requirements of the dispatching center layer, and solves the problems of a large number of resources, diverse types, and small capacity of some single units.

[0051] The load terminal layer supports multi-protocol and multi-scenario applications, realizes the interconnection and information perception of different types of energy-consuming equipment, monitoring meters, and sensor equipment in commercial buildings, energy storage, charging piles, and industrial enterprises, and has the functions of energy data collection, processing, storage, and control. The load terminal layer includes multiple load control terminals, which adopt a modular design and have an encryption module to block security risks from the terminal acquisition and control network, ensure secure interconnection and data interaction with the dispatching access network, and perceive the operation data of the electric pile in real time, including the current operating status, current load, charging time, SOC, upward adjustment margin, downward adjustment margin, etc. At the same time, it can respond to the control instructions issued by the multiple load terminals in real time.

[0052] This embodiment also provides a multi-scenario load real-time control method for real-time power grid regulation requirements, including the following steps:

[0053] Step S1: Obtaining energy consumption data of users to be evaluated, wherein the energy consumption data of users include user energy consumption behavior, uncertain factors and electricity prices;

[0054] Step S2: Establishing a user online energy consumption behavior model that dynamically responds to load aggregation control requests according to the demand response target;

[0055] Step S3: inputting the energy consumption data of the user to be evaluated into a trained user online energy consumption behavior model, and the model outputs a user adjustable load adjustment ratio.

[0056] The user online energy usage behavior model is implemented on the basis of a combined online learning algorithm of user fatigue. The user online energy usage behavior model is trained by the following method:

[0057] Step 1: Obtain historical energy consumption data of the source domain and the target domain; construct a source domain adjustable load generalized model and a target domain adjustable load generalized model according to the user energy consumption behavior and the uncertain factors affecting the user energy consumption behavior in the source domain and the target domain data, respectively, and obtain the quadratic regression dynamic relationship between the user energy consumption behavior and multiple uncertain factors in the source domain and the target domain based on the source domain adjustable load generalized model and the target domain adjustable load generalized model respectively;

[0058] Step 2: Cluster the historical energy consumption data of the source domain and the target domain respectively to extract similar energy consumption behaviors among different users. According to the clustering results of the source domain and the target domain and the quadratic regression dynamic relationship between the energy consumption behaviors of the source domain and the target domain users and various uncertain factors, the quadratic regression parameters of the energy consumption behaviors of the source domain and the target domain users and the uncertain factors within the class are obtained respectively.

[0059] Step 3: Calculate the cosine similarity of the quadratic regression parameter characteristics of the target domain and source domain users’ energy consumption behaviors and uncertain factors within the class, and obtain the probability distribution of the adjustable load regulation capacity of the target domain based on the cosine similarity;

[0060] Step 4: Based on the user fatigue algorithm, the probability distribution of the adjustable load regulation capacity of the target domain is obtained, and the energy consumption behavior characteristics of the users in the source domain and the target domain are aligned according to the local maximum mean difference. The loss function is minimized by continuously training the model until the model converges.

[0061] In step 4, the control request of the target domain adjustable load regulation capability probability includes the following methods:

[0062] Step 1) Inputting the user's fatigue coefficient function;

[0063] Step 2) Send a load aggregation control request to each user, and estimate the response probability of each user based on the observed performance;

[0064] Step 3) Estimate the maximum load power of all users in the area according to the load power size and sort them from large to small;

[0065] Step 4) A distribution of the probability of achieving the target domain adjustable load regulation capability according to the maximum load power.

[0066] According to the above system and method, the requirements for adjustable loads to participate in grid frequency regulation can be met. The principle of "direct purchase and direct control" can be adopted to achieve real-time perception and efficient regulation of adjustable loads, thereby achieving a response capability in seconds. Through load aggregation, the dispersed single load resources must be aggregated to a certain power scale before they can participate in grid frequency regulation.

[0067] In addition, in the embodiment, the multi-load control terminal realizes two load control modes: "pile control" and "station control", supporting adjustable loads to participate in multi-level frequency regulation, peak regulation and other application requirements.

[0068] Among them, under the "pile control" mode, the multi-load control terminal transparently forwards the control instructions, and the charging pile directly responds to the dispatcher's control instructions to the charging pile; under the "station control" mode, the multi-load control terminal perceives the status of the charging pile in real time, decomposes the control instructions, and selects the locally pre-established optimal strategy to control the charging load of the charging pile.

[0069] Obviously, the described embodiments are only some embodiments of the present invention, not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field and related fields without creative work should fall within the scope of protection of the present invention. The structures, devices and operating methods not specifically described and explained in the present invention are implemented according to the conventional means in the field unless otherwise specified and limited.

Claims

1. A multi-scenario load real-time control method for real-time power grid regulation needs, characterized in that ,, including the following steps: Step S1: Obtaining energy consumption data of users to be evaluated, wherein the energy consumption data of users include user energy consumption behavior, uncertain factors and electricity prices; Step S2: Establishing a user online energy consumption behavior model that dynamically responds to load aggregation control requests according to the demand response target; Step S3: inputting the energy consumption data of the user to be evaluated into a trained user online energy consumption behavior model, and the model outputs a user adjustable load adjustment ratio.

2. According to claim 1, a multi-scenario load real-time control method for real-time power grid regulation needs is characterized in that: The user online energy usage behavior model is implemented on the basis of a combined online learning algorithm of user fatigue.

3. A multi-scenario load real-time control method for real-time power grid regulation needs according to claim 1, characterized in that: The user online energy consumption behavior model is trained by the following method: Step 1: Obtain historical energy consumption data of the source domain and the target domain; construct a source domain adjustable load generalized model and a target domain adjustable load generalized model according to the user energy consumption behavior and the uncertain factors affecting the user energy consumption behavior in the source domain and the target domain data, respectively, and obtain the quadratic regression dynamic relationship between the user energy consumption behavior and multiple uncertain factors in the source domain and the target domain based on the source domain adjustable load generalized model and the target domain adjustable load generalized model respectively; Step 2: Cluster the historical energy consumption data of the source domain and the target domain respectively to extract similar energy consumption behaviors among different users. According to the clustering results of the source domain and the target domain and the quadratic regression dynamic relationship between the energy consumption behaviors of the source domain and the target domain users and various uncertain factors, the quadratic regression parameters of the energy consumption behaviors of the source domain and the target domain users and the uncertain factors within the class are obtained respectively. Step 3: Calculate the cosine similarity of the quadratic regression parameter characteristics of the target domain and source domain users’ energy consumption behaviors and uncertain factors within the class, and obtain the probability distribution of the adjustable load regulation capacity of the target domain based on the cosine similarity; Step 4: Based on the user fatigue algorithm, the probability distribution of the adjustable load regulation capacity of the target domain is obtained, and the energy consumption behavior characteristics of the users in the source domain and the target domain are aligned according to the local maximum mean difference. The loss function is minimized by continuously training the model until the model converges.

4. A multi-scenario load real-time control method for real-time power grid regulation needs according to claim 3, characterized in that: The control request for the target domain adjustable load regulation capability probability includes the following methods: Step 1) Input the user's fatigue coefficient function; Step 2) Send a load aggregation control request to each user and estimate the response probability of each user based on the observed performance; Step 3) Estimate the maximum load power of all users in the area according to the load power size and sort them from large to small; Step 4) A distribution of the probability of achieving the target domain adjustable load regulation capability according to the load maximum power.

5. A multi-scenario load real-time control method for power grid real-time regulation needs according to any one of claims 1-4, characterized in that: The control method is implemented based on a power grid load aggregation collaborative control system, which includes: a dispatch center layer, a load aggregation layer, and a load terminal layer; The dispatch center layer issues demand response targets to the load aggregation layer according to the power grid operation status; The load aggregation layer responds to the load aggregation control request dynamically based on the user's online energy consumption behavior model according to the demand response target issued by the dispatch center layer, and uses an online algorithm to send a load aggregation control request to the user; The load terminal layer is used to collect electric energy data and respond in real time to the control instructions issued by the load aggregation layer.

6. A power grid load aggregation and coordinated control system according to claim 5, characterized in that: The dispatch center layer, based on the D5000 basic platform, builds a two-level load dispatch control system of sub-center and provincial dispatch; the dispatch center layer includes production control area I and management information area III; The production control area I includes a load real-time control unit, which is used to implement functional modules such as load resource access, load control, and market interface, realize multi-level dispatching to collect, aggregate, monitor, and regulate adjustable load resources in real time, receive the clearing results of the spot and auxiliary service markets, and provide real-time and operating data to the load operation management function; The management information III area includes a load operation management unit, which is used to implement modules such as load capacity evaluation, model management, information release and comprehensive display, realize model data sharing and comprehensive display of adjustable load resources between multi-level dispatching, and provide model data to the load real-time control function.

7. A power grid load aggregation and coordinated control system according to claim 6, characterized in that: The load aggregation layer is realized by various load aggregation platform, including: the smart energy service platform within the power grid company, the smart Internet of Vehicles platform and the social third-party aggregation platform; The load aggregation layer also includes a multi-load aggregation energy router and a multi-load aggregation energy controller; The multi-load aggregation energy router supports real-time optimization scheduling and control of adjustable loads, collects and aggregates massive discrete load resources in commercial buildings, energy storage, charging piles and industrial enterprise scenarios, and connects to industrial DCS, building BA system, energy storage PCS system and multi-load control terminal downstream. It is compatible with multiple device physical interfaces, supports multi-protocol access function, has flexible load adjustable capacity evaluation function, predicts base value load, calculates upper and lower limits of load adjustable amount, and prepares load control plan and formulates load control strategy. The multi-load aggregation energy controller realizes the aggregation of adjustable single resources of different types and characteristics in the regional power grid and regional optimization control, and responds to the load control instructions and data acquisition requirements of the dispatching center layer.

8. A power grid load aggregation and coordinated control system according to claim 7, characterized in that: The load terminal layer supports multi-protocol and multi-scenario applications, realizes the interconnection and information perception of different types of energy-consuming equipment, monitoring meters, and sensor equipment in commercial buildings, energy storage, charging piles, and industrial enterprises, and has the functions of energy data collection, processing, storage, and control; the load terminal layer includes multiple load control terminals, which adopt a modular design and can sense the operation data of the charging pile in real time, including the current operating status, current load, charging time, SOC, upward margin, downward margin, etc., and can respond to the control instructions issued by the multiple load terminals in real time.

Images