Long-time energy storage planning method based on unit clustering

Abstract

The invention relates to a long-time energy storage planning method based on unit clustering, and the method comprises the steps: building a unit clustering method through machine learning according to the technical parameters of units, and dividing a plurality of units into unit groups; according to the technical parameters of long-time energy storage, constructing a long-time energy storage scheduling model taking a year as an operation cycle; a unit clustering unit combination of a relaxation integer variable is constructed, long-time energy storage scheduling is fused, and long-time energy storage and unit synchronous scheduling is constructed; and long-time energy storage and unit synchronous scheduling are used as operation simulation, and a unit combination-containing long-time energy storage planning model is constructed. Compared with the prior art, the method has the advantages that the problem of unit commitment with the year as the time scale is solved, the final optimization result achieves long-time energy storage configuration scheme solving adaptive to system operation, and therefore accurate correlation between the long-time scale operation characteristics of the system and the long-time energy storage operation mode can be achieved, and energy storage optimal configuration is achieved.

Description

Technical Field

[0001] The present invention relates to the technical field of power system operation planning, and in particular to a long-term energy storage planning method based on unit clustering. Background Art

[0002] At present, the penetration rate of new energy technologies in the power system is constantly increasing. With the expansion of the scale of the power system, the role of unit flexibility in modeling is becoming more and more obvious. Therefore, it is particularly important to consider the unit combination constraints in the long-term power system operation optimization scenario. At the same time, the introduction of long-term energy storage also poses new challenges to the establishment of unit combination models.

[0003] For many years, both at home and abroad, there have been rich research results in the unit combination and energy storage modeling of power systems containing renewable energy. For example, reference 1 "Safety-constrained unit combination considering coordinated optimization of source, grid, load and storage" (Sha Qiangyi, Meng Qingchao, Huang Qingjiang. Electrical Engineering, 2024, (17): 38-44.) In response to the uncertainty of high-penetration wind power and photovoltaic power, a safety-constrained unit combination model including demand response, transmission line transmission switching and public energy storage is proposed. This model copes with the uncertainty of wind power and photovoltaic power output and system component failures by coordinating and optimizing the flexible resources on the source, grid, load and storage side. Reference 2 "Two-layer optimization model of energy storage system under the access of new energy power generation" (Chen Jianguo, Zheng Tuo, Hao Junyi, Dong Youlin, Hu Jingwei, Su Yixin. Journal of South-Central University for Nationalities (Natural Science Edition), 2024, 43(02): 245-251.) In view of the problems caused by the access of renewable distributed power sources such as wind power generation and photovoltaic power generation to the power grid, a two-layer optimization model of energy storage system under the access of wind and solar power is proposed. The upper model considers energy storage planning, and the lower model considers the economy and stability of the power grid when the energy storage system is running. The improved whale algorithm and the YALMIP+CPLEX joint method are used to solve it in MATLAB. Reference 3 "Source-load two-layer unit optimization combination model with wind power grid connection" (Li Yingliang, Dong Zhiwei, Bai Boxu, Guo Feiyan, Liu Mengxi, Li Fei. Acta Energiae Solaris Sinica, 2024, 45(03): 399-407.) proposed a two-layer unit combination model that takes into account the adjustable resources on both the source and load sides to solve the problems of insufficient system peak-shaving capacity and new energy consumption. The upper model establishes a price-type demand response model for the load side, and the lower model establishes a unit combination model for the source side that includes deep peak-shaving auxiliary services in the power market and a reward-and-punishment ladder-type carbon trading mechanism in the carbon market.

[0004] Based on various research results, there are many patent applications involving unit combination modeling and energy storage modeling. Patent application 1 "A large-scale safety-constrained unit combination acceleration method to improve model simplicity" (Chen Xi, Yan Daobo, Zhu Bangxian, Fang Rengcun, Dong Zhe, Zou Yuxin, Li Huihui, Bie Fangmei, Wei Li, Ai Xiaomeng. CN202310252548.4[P]. 2024-09-03.) uses the non-tight maximum line flow constraint identification and the unit state partial fixation method to improve the solution speed of the unit combination problem while ensuring the superiority of the solution. Patent application 2 "A unit combination calculation method based on LSTM" (Ren Guanhua, Jiao Yang, Chen Yanbo, Wu Chao, Zheng Guodong, Gao Yulong, Chen Donghui, Ma Yanhu, Sun Zhixiang. CN202110770004.8 [P]. 2021-09-21) proposes a unit combination calculation method based on LSTM, which uses the MinMaxScaler normalization method and K-means clustering algorithm to process historical data, and obtains the unit combination plan for the next day through two-step deep learning. The LSTM neural network is used to learn a large amount of historical data to obtain the unit combination mapping relationship. The unit combination plan can be directly obtained by calling the trained deep learning model.

[0005] The existing methods have the following inherent defects:

[0006] (1) The systems involved are either small in scale or face the problem of increased computing time due to increased scale, which limits the application of unit combination and related applications in large systems.

[0007] (2) The time scale of the unit combination and energy storage modeling involved is limited to days or a few typical days, which cannot adapt to the longer time scale fluctuations in months or years caused by seasonal changes in renewable energy output. This makes the regulation capacity of energy storage on a long time scale underestimated. Summary of the invention

[0008] The purpose of the present invention is to overcome the defects of the above-mentioned prior art and to provide a long-term energy storage planning method based on unit clustering to realize long-term energy storage planning of power system.

[0009] The purpose of the present invention can be achieved by the following technical solutions:

[0010] One aspect of the present invention provides a long-term energy storage planning method based on unit clustering, comprising the following steps:

[0011] Step S1, according to the technical parameters of the units, a unit clustering method is established by using machine learning to divide multiple units into unit groups;

[0012] Step S2, constructing a long-term energy storage scheduling model with a year as the operating cycle according to the technical parameters of the long-term energy storage;

[0013] Step S3, constructing a clustered unit combination of relaxed integer variables, integrating long-term energy storage scheduling, and constructing long-term energy storage and unit synchronous scheduling;

[0014] Step S4, taking long-term energy storage and unit synchronous scheduling as operation simulation, constructing a long-term energy storage planning model including unit combination.

[0015] As a preferred technical solution, in the step S1, a unit clustering method is established using machine learning based on the unit technical parameters, and multiple units are divided into unit groups, which is characterized in that: the unit clustering center is selected using a machine learning method, and the same type of units are represented as unit groups.

[0016] As a preferred technical solution, the technical parameters of the unit include one or more of minimum power generation, maximum power generation, climb limit, climb limit, minimum online time, minimum offline time, startup cost coefficient, fixed operating cost coefficient, and variable operating cost coefficient.

[0017] As a preferred technical solution, in the step S2, a long-term energy storage scheduling model with an operating cycle of one year is constructed according to the technical parameters of the long-term energy storage, which is characterized in that the technical parameters of the long-term energy storage include one or more of the charging and discharging power, charging and discharging efficiency, and operating cycle of the long-term energy storage.

[0018] As a preferred technical solution, in the step S3, a unit clustering unit combination of relaxed integer variables is constructed, integrated into the long-term energy storage scheduling, and the long-term energy storage and unit synchronous scheduling is constructed, which is characterized in that: the unit clustering unit combination of relaxed integer variables is based on the clustered unit group, with the goal of minimizing the cost function, including one or more of unit combination constraints, power balance constraints, and long-term energy storage operation mode constraints.

[0019] As a preferred technical solution, in step S3, the objective function of the unit clustering unit combination of relaxed integer variables is to minimize the system operating cost, which is characterized by: the system operating cost includes one or more of the unit start-up and shutdown cost, the unit variable operating cost, and the unit fixed operating cost.

[0020] As a preferred technical solution, in step S4, the long-term energy storage planning model containing the unit combination has an objective function as follows:

[0021]

[0022] Among them, T is the total number of time periods, C is the total number of units, They represent the startup cost, fixed power generation cost and variable power generation cost of class C units respectively. with U c，tThey represent the number of started units and the number of units in operation at time t of unit group c respectively. with U c，t In the model, all of them were relaxed into continuous variables, P c，t represents the power generation of the generator group c at time t, represents the investment cost of long-term energy storage s, inv s Represents the planned quantity of long-term energy storage s.

[0023] As an optimal technical solution, the long-term energy storage planning model takes into account one or more of the unit combination constraints, power balance constraints, long-term energy storage equipment charging power constraints, long-term energy storage equipment discharging power constraints, long-term energy storage equipment charge state timing constraints, and long-term energy storage equipment capacity constraints.

[0024] Another aspect of the present invention provides an electronic device, comprising: one or more processors and a memory, wherein the memory stores one or more programs, and the one or more programs include instructions for executing the long-term energy storage planning method based on unit clustering as described above.

[0025] Compared with the prior art, the present invention has at least one of the following beneficial effects:

[0026] (1) Realizing the long-term energy storage planning of the power system: The present invention first divides a plurality of units into unit groups by clustering according to the technical parameters of the units, and then uses a pre-constructed long-term energy storage planning model of unit clustering unit combination with relaxed integer variables to obtain the long-term energy storage planning result of the power system, thereby realizing the long-term energy storage planning of the power system.

[0027] (2) Clustering fully considers the differences between unit parameters: The present invention uses machine learning methods to determine the unit clustering center based on the maximum power generation, minimum power generation, ramp-up and ramp-down limits, online time, offline time, startup cost, fixed operating cost, and variable operating cost parameters, and implements unit clustering according to the similarity of technical parameters, fully considering the differences between unit parameters, thereby ensuring the accuracy of clustering.

[0028] (3) High system flexibility: The present invention realizes the solution of the unit combination problem with long-term energy storage scheduling on a yearly time scale. The final optimization result retains the constraints on non-continuous variables such as unit start and stop at a high time resolution, thereby being able to describe the flexibility of the system more strictly and accurately.

[0029] (4) Adapting to long-term scale fluctuations: In order to address the problem that the regulatory capacity of energy storage on a long-term scale is underestimated due to the inability to adapt to longer-term scale fluctuations on a monthly or annual basis caused by seasonal changes in the output of new energy, the present invention solves the unit combination problem on a yearly time scale. The final optimization result solves the long-term energy storage configuration scheme that is compatible with the system operation, thereby achieving an accurate correlation between the long-term operating characteristics of the system and the long-term energy storage operating mode, thereby achieving optimal energy storage configuration. BRIEF DESCRIPTION OF THE DRAWINGS

[0030] Figure 1 It is a schematic diagram of a long-term energy storage planning method based on unit clustering in an embodiment;

[0031] Figure 2 Schematic diagram of an electronic device in an embodiment. DETAILED DESCRIPTION

[0032] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work should fall within the scope of protection of the present invention.

[0033] The terms "first" and "second" are used for descriptive purposes only and should not be understood as indicating or implying relative importance or implicitly indicating the number of the indicated technical features. Thus, a feature defined as "first" or "second" may explicitly or implicitly include one or more of the features. In the description of this application, the meaning of "plurality" is two or more, unless otherwise clearly and specifically defined.

[0034] Example 1

[0035] In view of the problems existing in the above-mentioned prior art, this embodiment provides a long-term energy storage planning method based on unit clustering, see Figure 1 , including the following steps S1 to S4:

[0036] S1, based on the technical parameters of the units, a unit clustering method is established using machine learning to divide multiple units into unit groups;

[0037] S2, based on the technical parameters of long-term energy storage, build a long-term energy storage scheduling model with a year as the operating cycle;

[0038] S3, construct the unit clustering unit combination of relaxed integer variables, integrate long-term energy storage scheduling, and construct long-term energy storage and unit synchronous scheduling;

[0039] S4, taking long-term energy storage and unit synchronous scheduling as operation simulation, construct a long-term energy storage planning model including unit combination.

[0040] The following describes in detail steps S1 to S4 of the long-term energy storage planning method based on unit clustering described in this embodiment.

[0041] Table 1 System parameters required by the model

[0042]

[0043] Step S1, based on the technical parameters of the units, a unit clustering method is established by using machine learning to divide multiple units into unit groups.

[0044] Specifically, in this implementation example, a method for clustering units based on unit technical parameters using a machine learning solution is provided. The K-means clustering method clusters the units according to k clustering centers, replacing the technical parameters of a single unit with technical parameters related to the unit group.

[0045] Among them, the technical parameters (i.e., the parameters describing the unit group) include one or more of the maximum output, minimum output, maximum ramp power, minimum online / offline time, output cost, startup cost, and the total number of generator sets in the unit group. The clustering data used by the machine learning method is a vector composed of the combination of technical parameters of each unit. Considering the above multiple technical parameters, the vector form of the combination of the unit technical parameters is

[0046] v = [maximum output, minimum output, maximum climbing power, minimum online power, etc.].

[0047] For unit g (g=1,2,…,G), calculate its technical parameter vector v g and the center vector v of the group c (c=1,2,…,C) c The distance d(v g ,v c )=||v k -v c ||2, minimum d(v g ,v c ) value is the group c to which group g should be classified. After completing the group classification, the cluster center is updated to the centroid of each group member. The calculation method is the mean value of the technical parameter vector of each member unit in the group.

[0048]

[0049] Among them, CN c Indicates the total number of units in unit group c. Use centroid As the new cluster center, repeat the above clustering process until each cluster center is no longer updated or reaches the preset maximum number of iterations, and the unit division is completed.

[0050] Step S2, constructing a long-term energy storage scheduling model with a year as the operating cycle according to the technical parameters of the long-term energy storage.

[0051] Specifically, the long-term energy storage technical parameters involved in this step include: one or more of the charging and discharging power, charging and discharging efficiency, and operation cycle of the long-term energy storage.

[0052] In this implementation example, the long-term energy storage scheduling model with a year as the operating cycle includes:

[0053] Charging power constraints for long-term energy storage equipment:

[0054] Discharge power constraints of long-term energy storage equipment:

[0055] Long-term energy storage equipment charge state timing constraints:

[0056] Capacity constraints of long-term energy storage equipment:

[0057] Long-term energy storage equipment operation cycle constraints: sl s,t =sl s,0 ,t=T l ,2T l ,…

[0058] in, sp s , Respectively represent the maximum charge and discharge power of long-term energy storage device s, spc s，t 、spd s，t They represent the charging and discharging power of the long-term energy storage device s at time t, η c , η d They represent the charging and discharging efficiency of long-term energy storage devices, Indicates the upper limit of the state of charge of the long-term energy storage device s, sl s，t represents the charge state of the long-term energy storage device s at time t, T l Indicates the cycle time of long-term energy storage equipment (unit: hours).

[0059] Step S3, constructing a clustered unit combination of relaxed integer variables, integrating long-term energy storage scheduling, and constructing long-term energy storage and unit synchronous scheduling.

[0060] Specifically, in this embodiment, a unit clustering unit commitment model with a relaxed integer variable containing long-term energy storage is constructed, and its objective function is:

[0061]

[0062] Among them, T is the total number of time periods, C is the total number of units, They represent the startup cost, fixed power generation cost and variable power generation cost of class C units respectively. with U c，t They represent the number of starting units and the number of units in operation at time t of unit group c, respectively. In this relaxation model, they are relaxed to continuous variables. c，t Represents the power generation of unit group c at time t.

[0063] In this implementation example, the constraints of the long-term energy storage planning model of unit clustering and unit combination based on relaxed integer variables include: unit combination constraints, power balance constraints, long-term energy storage equipment charging power constraints, long-term energy storage equipment discharging power constraints, long-term energy storage equipment charge state timing constraints, and long-term energy storage equipment capacity constraints. One or more.

[0064] Unit commitment constraints include:

[0065] 1) Unit output constraints:

[0066] 2) Unit climbing constraints:

[0067] 3) Logical constraints on unit switch action and online status:

[0068] 4) Minimum online and offline time constraints for units:

[0069] in, P c They represent the upper and lower limits of the power generation of the representative units in the unit group c, They represent the upper and lower limits of the ramp power of the representative units in the unit group c, TU c , TD c They represent the minimum online and offline time of the representative unit in the unit group c, P c，t , U c，t , They represent the power generation of the unit group c at time t, the number of units in operation, the number of units started, and the number of units shut down, respectively. c Indicates the total number of units in unit group c.

[0070] Step S4, taking long-term energy storage and unit synchronous scheduling as operation simulation, constructing a long-term energy storage planning model including unit combination.

[0071] Specifically, in this embodiment, the objective function of the long-term energy storage planning model including unit combination is:

[0072]

[0073] Among them, T is the total number of time periods, C is the total number of units, They represent the startup cost, fixed power generation cost and variable power generation cost of class C units respectively. with U c，t They represent the number of started units and the number of units in operation at time t of unit group c respectively. with U c，t In the model, all of them were relaxed into continuous variables, P c，t represents the power generation of the generator group c at time t, represents the investment cost of long-term energy storage s, inv s Represents the planned quantity of long-term energy storage s.

[0074] In this embodiment, the long-term energy storage planning model considers one or more of unit combination constraints, power balance constraints, long-term energy storage device charging power constraints, long-term energy storage device discharging power constraints, long-term energy storage device charge state timing constraints, and long-term energy storage device capacity constraints.

[0075] In this embodiment, the long-term energy storage scheduling constraints used for planning include:

[0076] Charging power constraints for long-term energy storage equipment:

[0077] Discharge power constraints of long-term energy storage equipment:

[0078] Long-term energy storage equipment charge state timing constraints:

[0079] Capacity constraints of long-term energy storage equipment:

[0080] Among them, the parameters represents the single unit investment and construction cost of the alternative long-term energy storage device s, inv s is an integer variable representing the number of candidate long-term energy storage devices s to be built. sp s , Respectively represent the maximum charge and discharge power of long-term energy storage device s, spc s，t 、spd s，t They represent the charging and discharging power of the long-term energy storage device s at time t, η c , η d They represent the charging and discharging efficiency of the long-term energy storage device s at time t, Indicates the upper limit of the state of charge of the long-term energy storage device s, sl s，t Represents the charge state of the long-term energy storage device s at time t.

[0081] In this embodiment, the power balance constraint is:

[0082] Among them, p w，t represents the output of fan w at time t, D t Represents the total system load at time t.

[0083] In summary, this method solves the unit combination problem with long-term energy storage scheduling on a yearly time scale and realizes the planning of long-term energy storage in the system. Its final optimization result retains the constraints on non-continuous variables such as unit start and stop at a high time resolution, realizes the reasonable scheduling of long-cycle energy storage equipment, and thus can more strictly and accurately describe the flexibility of the system.

[0084] Example 2

[0085] Based on Example 1, this embodiment provides an electronic device, including: one or more processors and a memory, wherein the memory stores one or more programs, and the one or more programs include instructions for executing the long-term energy storage planning method based on unit clustering as described above.

[0086] like Figure 2 As mentioned above, at the hardware level, the electronic device includes a processor, an internal bus, a network interface, a memory, and a non-volatile memory, and may also include other hardware required for the business. The processor reads the corresponding computer program from the non-volatile memory into the memory and then runs it to achieve the above Figure 1 The long-term energy storage planning method described above. Of course, in addition to the software implementation, the present invention does not exclude other implementation methods, such as logic devices or a combination of software and hardware, etc., that is, the execution subject of the following processing flow is not limited to each logic unit, but can also be hardware or logic devices.

[0087] The memory may include non-permanent storage in a computer-readable medium, random access memory (RAM) and / or non-volatile memory in the form of read-only memory (ROM) or flash RAM. The memory is an example of a computer-readable medium.

[0088] Computer readable media include permanent and non-permanent, removable and non-removable media that can be implemented by any method or technology to store information. Information can be computer readable instructions, data structures, program modules or other data. Examples of computer storage media include, but are not limited to, phase change memory (PRAM), static random access memory (SRAM), dynamic random access memory (DRAM), other types of random access memory (RAM), read-only memory (ROM), electrically erasable programmable read-only memory (EEPROM), flash memory or other memory technology, compact disk read-only memory (CD-ROM), digital versatile disk (DVD) or other optical storage, magnetic cassettes, magnetic disk storage or other magnetic storage devices or any other non-transmission media that can be used to store information that can be accessed by a computing device. As defined herein, computer readable media does not include temporary computer readable media (transitory media), such as modulated data signals and carrier waves.

[0089] The above is only a specific embodiment of the present invention, but the protection scope of the present invention is not limited thereto. Any technician familiar with the technical field can easily think of various equivalent modifications or replacements within the technical scope disclosed by the present invention, and these modifications or replacements should be included in the protection scope of the present invention. Therefore, the protection scope of the present invention shall be based on the protection scope of the claims.

Claims

1. A long-term energy storage planning method based on unit clustering, characterized in that: The steps include: Step S1, according to the technical parameters of the units, a unit clustering method is established by using machine learning to divide multiple units into unit groups; Step S2, constructing a long-term energy storage scheduling model with a year as the operating cycle according to the technical parameters of the long-term energy storage; Step S3, constructing a clustered unit combination of relaxed integer variables, integrating long-term energy storage scheduling, and constructing long-term energy storage and unit synchronous scheduling; Step S4, taking long-term energy storage and unit synchronous scheduling as operation simulation, constructing a long-term energy storage planning model including unit combination.

2. According to claim 1, a long-term energy storage planning method based on unit clustering is characterized in that: In the step S1, a unit clustering method is established based on the unit technical parameters using machine learning, and multiple units are divided into unit groups, which is characterized in that the unit clustering center is selected using a machine learning method, and the same type of units are represented by unit groups.

3. According to claim 2, a long-term energy storage planning method based on unit clustering is characterized in that: The technical parameters of the unit include one or more of minimum power generation, maximum power generation, climbing limit, climbing limit, minimum online time, minimum offline time, startup cost coefficient, fixed operating cost coefficient, and variable operating cost coefficient.

4. The long-term energy storage planning method based on unit clustering according to claim 1 is characterized in that: In the step S2, a long-term energy storage scheduling model with an operating cycle of one year is constructed according to the technical parameters of the long-term energy storage, which is characterized in that the technical parameters of the long-term energy storage include one or more of the charging and discharging power, charging and discharging efficiency, and operating cycle of the long-term energy storage.

5. The long-term energy storage planning method based on unit clustering according to claim 1 is characterized in that: In the step S3, a clustered unit combination of relaxed integer variables is constructed, integrated into long-term energy storage scheduling, and long-term energy storage and unit synchronous scheduling is constructed, which is characterized in that: the clustered unit combination of relaxed integer variables is based on the clustered unit group, with the goal of minimizing the cost function, including one or more of unit combination constraints, power balance constraints, and long-term energy storage operation mode constraints.

6. A long-term energy storage planning method based on unit clustering according to claim 5, characterized in that: The objective function of the unit clustering unit combination of the relaxed integer variables is to minimize the system operation cost, which is characterized by: the system operation cost includes one or more of the unit start-up and shutdown cost, the unit variable operation cost, and the unit fixed operation cost.

7. The long-term energy storage planning method based on unit clustering according to claim 1 is characterized in that: In step S4, the long-term energy storage planning model including unit combination has an objective function as follows: Among them, T is the total number of time periods, C is the total number of units, They represent the startup cost, fixed power generation cost and variable power generation cost of class C units respectively. with U c，t They represent the number of started units and the number of units in operation at time t of unit group c respectively. with U c，t In the model, all of them were relaxed into continuous variables, P c，t represents the power generation of the generator group c at time t, represents the investment cost of long-term energy storage s, inv s Represents the planned quantity of long-term energy storage s.

8. A long-term energy storage planning method based on unit clustering according to claim 7, characterized in that: The long-term energy storage planning model considers one or more of unit combination constraints, power balance constraints, long-term energy storage equipment charging power constraints, long-term energy storage equipment discharging power constraints, long-term energy storage equipment charge state timing constraints, and long-term energy storage equipment capacity constraints.

9. The long-term energy storage planning method based on unit clustering according to claim 1 is characterized in that: In the process of establishing the unit clustering method using machine learning, the centroid is calculated using the following formula: Among them, CN c Indicates the total number of units in unit group c, using the centroid As the new cluster center, the clustering process is repeated until each cluster center is no longer updated or the preset maximum number of iterations is reached, and the unit division is completed.

10. An electronic device, characterized in that: include: One or more processors and a memory, wherein the memory stores one or more programs, and the one or more programs include instructions for executing the long-term energy storage planning method based on unit clustering as described in any one of claims 1-9.

Images