Distributed energy storage optimization scheduling method and system with minimum peak-valley difference of power distribution network

Abstract

The invention discloses a distributed energy storage optimization scheduling method and system with the minimum peak-valley difference of a power distribution network, and the method comprises the steps: collecting power grid data, including load data and distributed energy storage data; establishing a distributed energy storage mathematical model, and analyzing distributed energy storage state parameters; the distributed energy storage devices are clustered according to the state parameters of distributed energy storage, K distributed energy storage clusters are obtained, and an equivalent centralized energy storage model is established for each cluster; and taking the equivalent centralized energy storage model as a control unit and taking the minimum peak-valley difference of the load of the power distribution network as a target function, and carrying out optimal scheduling on the charging and discharging power of the distributed energy storage. The method can reduce the variable dimension of the distributed energy storage, reduces the solving difficulty, facilitates the reasonable scheduling of the distributed energy storage, reduces the peak-valley difference of the operation of the power distribution network, guarantees the safe and stable operation of the power grid, and improves the utilization rate of the distributed energy storage in the regulation and control operation process.

Description

Technical Field

[0001] The present invention belongs to the technical field of energy storage scheduling, and relates to a distributed energy storage optimization scheduling method and system with minimum peak-to-valley difference in distribution network. Background Art

[0002] Due to the large number of distributed energy storage and small capacity, if it directly participates in the operation of the distribution network, when the number of optimization cycles and the amount of energy storage are large, many variables will appear in the optimization model, which will greatly affect the optimization effect and time, and will lead to problems such as difficulty in algorithm solution and explosion of decision variable dimensions. The K-means algorithm is usually used for distributed energy storage optimization and control.

[0003] Traditional distributed energy storage optimization control based on the K-means algorithm only considers the life of energy storage and the economic efficiency of operating costs, while ignoring the problem of power grid security and stability with the smallest peak-to-valley difference. In addition, the traditional K-means algorithm randomly selects K points in the data set as clustering centers at the beginning. The completely random selection has a great impact on the clustering results and running time, resulting in slow algorithm convergence. Therefore, there is an urgent need for a scheduling solution that can efficiently control distributed energy storage and maximize the security of the power grid. Summary of the invention

[0004] In order to solve the deficiencies in the prior art, the present invention provides a distributed energy storage optimization scheduling method and system with the smallest peak-to-valley difference in the distribution network, which solves the problem that the distributed energy storage is huge in quantity and difficult to regulate efficiently and economically. Since the distributed energy storage has a large amount of energy and a small capacity, if it directly participates in the operation of the distribution network, it will lead to problems such as difficulty in solving the algorithm and explosion of the dimension of the decision variables.

[0005] The present invention adopts the following technical solution.

[0006] A first aspect of the present invention provides a distributed energy storage optimization scheduling method with minimum peak-to-valley difference in a distribution network, comprising:

[0007] Step 1: Collect grid data, including load data and distributed energy storage data;

[0008] Step 2: Establish a distributed energy storage mathematical model based on the data collected in step 1, and analyze the distributed energy storage state parameters;

[0009] Step 3: Cluster the distributed energy storage devices according to the state parameters of the distributed energy storage obtained by the analysis in step 2 to obtain K distributed energy storage clusters, and establish an equivalent centralized energy storage model for each cluster;

[0010] Step 4: Taking the equivalent centralized energy storage model established in step 3 as the control unit and minimizing the peak-to-valley difference of the distribution network load as the objective function for optimization, the charging and discharging power of the distributed energy storage is optimized.

[0011] Preferably, in step 1, the active power of the load in the power grid, the state of charge SOC, the charging and discharging power, the charging and discharging efficiency and the rated capacity of the distributed energy storage are collected.

[0012] Preferably, in step 2, the distributed energy storage mathematical model established based on the data collected in step 1 is:

[0013]

[0014] SOC min ≤SOC i ≤SOC max

[0015]

[0016] Among them, SOC0, SOC t , SOC t-1 is the SOC value at the initial energy storage, t, and t-1;

[0017] P ch and P dis They are the charging and discharging power of energy storage from time t to time t-1;

[0018] η ch and η dis are the energy storage charging and discharging efficiency respectively; Δt is a continuous charging and discharging period; S rat is the rated capacity of the energy storage;

[0019] SOC min , SOC max are the minimum and maximum allowed values ​​of energy storage SOC, respectively;

[0020] P dismax and P chmax are the maximum charging power and discharging power of the energy storage respectively.

[0021] Preferably, in step 2, the distributed energy storage state parameters in the distributed energy storage mathematical model include: power, state of charge, and available capacity.

[0022] Preferably, in step 3, the process of clustering the distributed energy storage devices according to the state parameters of the distributed energy storage to obtain K distributed energy storage clusters is:

[0023] Step 3.1: Use the state parameters of each energy storage device to form a feature vector, and use the feature vectors of n energy storage devices to form a sample in is the i-th point in the sample, representing the characteristic vector composed of the state parameters of energy storage device i, and Where V i1 、Vi2 、V i3 for The three aggregated characteristic quantities represent the values ​​of the three state parameters of the energy storage device i;

[0024] From the sample Randomly select a point As the initial aggregation center, where k is 1, 2, .., or n;

[0025] Step 3.2: Calculate the distance D from the nearest initial clustering center for each point in the sample;

[0026]

[0027] Step 3.3: Add the farthest point as the next clustering center;

[0028] Step 3.4: Repeat steps 3.2 and 3.3 until K clustering centers are selected;

[0029] Step 3.5: Calculate the distance from each point in the sample to any aggregation center distance;

[0030] Assign each point to the cluster corresponding to the aggregation center with the smallest distance, and update the aggregation center of each cluster:

[0031]

[0032] in for The updated aggregation center of the cluster, W m 、N m are the number of data points with existing aggregation centers and the number of total cluster centers respectively;

[0033] Step 3.6: Repeat step 3.5 until the average function E converges, end clustering, and output K clusters, i.e., K distributed energy storage clusters, where the average function E is:

[0034]

[0035] Where p is all points in space, is the average value of cluster cm.

[0036] Preferably, in step 3, the equivalent centralized energy storage model established for each cluster is:

[0037]

[0038]

[0039]

[0040]

[0041]

[0042] SOC min ≤SOC all ≤SOC max

[0043] SOC first =SOC last

[0044] Among them, P allrate is the rated power of the equivalent centralized energy storage, P irate is the rated power of energy storage device i, N is the total number of energy storage devices;

[0045] S allrate is the rated capacity of the equivalent centralized energy storage; S irate is the rated capacity of energy storage device i;

[0046] η allch and η alldis are the charging efficiency and discharging efficiency of equivalent centralized energy storage respectively; η ich and η idis are the charging efficiency and discharging efficiency of energy storage device i, respectively;

[0047] SOC all is the SOC value of the equivalent centralized energy storage, SOC min and SOC max are the minimum and maximum values ​​allowed by the equivalent centralized energy storage SOC, SOC first and SOC last They are the initial SOC and final SOC of the equivalent centralized energy storage, respectively.

[0048] Preferably, in step 4, the objective function with the equivalent centralized energy storage model as the control unit and the minimum peak-to-valley difference of the distribution network load as the optimization target is:

[0049]

[0050] Where f Δp is the peak-to-valley difference of the load in the distribution network at each period; K is the number of load power intervals; M j is the number of loads in load power interval j; P javrload is the average power within the load power interval j; P avrload is the average power of the distribution network in all periods; P jess It is the total active output of the equivalent centralized energy storage in the load power interval j. A positive value indicates discharge, and a negative value indicates charging.

[0051] Preferably, the constraint conditions of the objective function include power flow constraint and equivalent centralized energy storage model;

[0052] The power flow constraint is:

[0053]

[0054] Among them, P ji’s is the active power injected by node i' in load power interval j; U j’I’ is the voltage from node j' to node i', U j’l is the voltage from node j' to node l, G i’l and B il are the conductance and susceptance from node i' to node l respectively; θ i’l is the phase angle difference between node i' and node l; P ji’avrload is the active power of the node i' load in the load power interval j; P ji’ess is the active power of the energy storage installed at node i' in load power interval j.

[0055] Preferably, in step 4, the objective function and its constraints are solved to obtain the total active output P of the equivalent centralized energy storage model. jess , for P jess The power of each energy storage device in the equivalent centralized energy storage model is allocated to achieve the optimal scheduling of the charging and discharging power of the distributed energy storage. The specific allocation method is:

[0056] P jess When it is a negative value, the charging power is distributed as follows:

[0057]

[0058] P jess When it is positive, the discharge power distribution is as follows:

[0059]

[0060] P jess When it is 0, no power allocation is performed;

[0061] Where P irate is the rated power of energy storage device i; P Jrate is the rated power of the energy storage device J; P ich and P idis are the charging and discharging power of energy storage device i respectively; P Jch and P Jdis are the charging and discharging power of energy storage device J respectively; P allch and P alldisis the total charging and discharging power required; N is the total number of energy storage devices; f ch (SOC i ) and f dis (SOC i ) are the SOC functions of charging and discharging of energy storage device i; f ch (SOC J ) and f dis (SOC J ) are the SOC functions of the energy storage device J for charging and discharging, respectively.

[0062] A second aspect of the present invention provides a distributed energy storage optimization scheduling system with minimum peak-to-valley difference in a distribution network, comprising:

[0063] Data acquisition module, used to collect power grid data, including load data and distributed energy storage data;

[0064] A mathematical model building module is used to establish a distributed energy storage mathematical model based on the collected data and analyze the distributed energy storage state parameters;

[0065] A clustering grouping module is used to cluster distributed energy storage devices according to state parameters of distributed energy storage to obtain K distributed energy storage clusters, and establish an equivalent centralized energy storage model for each cluster;

[0066] The optimization scheduling module is used to optimize the charging and discharging power of distributed energy storage by taking the equivalent centralized energy storage model as the control unit and minimizing the peak-to-valley difference of the distribution network load as the objective function of the optimization target.

[0067] A terminal comprises a processor and a storage medium; the storage medium is used to store instructions; the processor is used to operate according to the instructions to execute the steps of the method.

[0068] A computer-readable storage medium stores a computer program, which implements the steps of the method when executed by a processor.

[0069] The beneficial effects of the present invention are as follows:

[0070] 1. The present invention proposes a method for large-scale distributed energy storage to participate in the operation of the distribution network, which solves the problem of aggregate optimization of a large number of distributed energy storage resources connected to the power grid in the large-scale distributed energy storage scenario. Most existing studies have established a multi-cycle optimization scheduling model for distributed energy storage. However, when the number of optimization cycles and the amount of energy storage are large, many variables will appear in the optimization model, which greatly affects the optimization effect and time.

[0071] 2. The present invention reduces the variable dimension of distributed energy storage, reduces the difficulty of solving, is conducive to the reasonable scheduling of distributed energy storage, reduces the peak-to-valley difference in distribution network operation, ensures the safe and stable operation of the power grid, and improves the utilization rate of distributed energy storage during the regulation and operation process.

[0072] 3. The clustering algorithm of the present invention optimizes the selection of initial values. First, a seed data is randomly selected, and then the distance from other data points to this data point is calculated. Data points with a larger distance are selected to replace the randomly selected data points as new data points. The above steps are iterated until K data points are selected. This algorithm process solves the problem of slow convergence caused by random selection of initial data points, and better realizes efficient regulation of distributed energy storage. BRIEF DESCRIPTION OF THE DRAWINGS

[0073] Figure 1 This is a flow chart of a distributed energy storage optimization scheduling method with minimal peak-to-valley difference in a distribution network according to the present invention. DETAILED DESCRIPTION

[0074] In order to make the purpose, technical scheme and advantages of the present invention clearer, the technical scheme of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. The embodiments described in this application are only embodiments of a part of the present invention, rather than all embodiments. Based on the spirit of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative work belong to the protection scope of the present invention.

[0075] like Figure 1 As shown, the present invention provides a distributed energy storage optimization scheduling method with the minimum peak-to-valley difference of the distribution network, comprising the following steps:

[0076] Step 1: Collect grid data, including load data and distributed energy storage data;

[0077] Further preferably, the collected grid data include the active power of the load in the grid, the state of charge SOC, the charge and discharge power, the charge and discharge efficiency and the rated capacity of the distributed energy storage.

[0078] Step 2: Establish a distributed energy storage mathematical model based on the data collected in step 1 and analyze its state parameters;

[0079] Further preferably, the distributed energy storage mathematical model established based on the data collected in step 1 is:

[0080] The calculation formula of SOCt at time t is as follows:

[0081]

[0082] Energy storage constraints include SOC constraints and charge and discharge constraints.

[0083] The SOC limits are as follows:

[0084] SOC min ≤SOC i ≤SOC max

[0085] The charging and discharging limits are as follows:

[0086]

[0087] Among them, SOC0, SOC t , SOC t-1 is the SOC value at the initial energy storage, t, and t-1;

[0088] P ch and P dis They are the charge and discharge power of energy storage from time t to time t-1, and at most one of them cannot be 0;

[0089] η ch and η dis They are energy storage charging and discharging efficiency;

[0090] Δt is a continuous charge and discharge period;

[0091] S rate is the rated capacity of the energy storage;

[0092] SOC min , SOC max are the minimum and maximum allowed values ​​of energy storage SOC, respectively;

[0093] P dismax and P chmax are the maximum charging power and discharging power of the energy storage respectively.

[0094] The distributed energy storage state parameters in the distributed energy storage mathematical model include: power, state of charge, and available capacity.

[0095] Step 3: Cluster the distributed energy storage devices according to the state parameters of the distributed energy storage obtained by the analysis in step 2 to obtain K distributed energy storage clusters, and establish an equivalent centralized energy storage model for each cluster;

[0096] Further preferably, the parameters of the distributed energy storage are input, the initial cluster center is randomly determined, the distributed energy storage is clustered and grouped using a clustering algorithm, a large number of distributed energy storage resources are divided into K clusters, and an equivalent centralized energy storage model is established for each cluster.

[0097] In actual operation, allowing energy storage to charge and discharge freely will not only lead to unnecessary loss of energy storage life, but also reduce its dispatch potential; and controlling them one by one is costly and labor-intensive. Therefore, it is considered to cluster the distributed energy storage involved in the operation and use the aggregation group as a unit to participate in the regulation.

[0098] The fundamental purpose of clustering a large number of distributed energy storage is to cluster distributed energy storage with similar regulation potential and operational space into one group, so as to give full play to the demand response regulation capability of energy storage in the dispatch process. The operating state of energy storage participating in power system dispatch is a multidimensional variable, including power, state of charge, available capacity, etc.

[0099] Based on the aggregation algorithm, a large number of distributed energy storage resources are divided into K clusters. An equivalent centralized energy storage model is established for each cluster. This aggregation process reduces the number of decision variables and improves the feasibility of distributed energy storage participating in the interactive power system.

[0100] During regulation, the distributed energy storage needs to be clustered into an equivalent centralized energy storage unit. In the scheduling optimization calculation, only one charging and discharging variable of the centralized energy storage is required. In order to reasonably allocate the power of each energy storage unit during the scheduling process and maintain the relative balance of SOC, the rated power and SOC of each energy storage unit are used to jointly determine its charging and discharging power.

[0101] Based on the energy storage regulation potential and parameter differences, the present invention uses a clustering algorithm to cluster distributed energy storage. The process is as follows:

[0102] Step 3.1: Use n distributed energy storage feature vectors as samples: The feature vector consists of the aggregated feature quantities of each distributed energy storage:

[0103] A distributed energy storage is randomly selected as the initial aggregation center;

[0104] Step 3.2: Calculate the Euclidean distance D from the nearest initial cluster center for each point in the sample;

[0105]

[0106] Step 3.3: Add the farthest point as the next aggregation center;

[0107] Step 3.4: Repeat steps 3.2 and 3.3 until K cluster centers are selected.

[0108] Step 3.5: Calculate the Euclidean distance from each eigenvector to the cluster center:

[0109]

[0110] For each feature vector, it is assigned to the cluster corresponding to the aggregation center with the smallest distance. In each cluster, the aggregation center is updated:

[0111]

[0112] W m 、N m are the number of data points with existing aggregation centers and the number of total cluster centers respectively;

[0113] Step 3.6: Repeat the above step 3.5 until the average function converges, then the clustering is completed, and K clusters are output, that is, K distributed energy storage clusters, where the average function E is:

[0114]

[0115] Where p is all points in space, is the average value of cluster cm. The average function is used to measure whether the clustering is finished. When the function converges, it indicates that the clustering is finished.

[0116] The equivalent centralized energy storage model established for each cluster is:

[0117] Charging and discharging power constraints of distributed energy storage systems:

[0118]

[0119] in,

[0120] P allrate is the rated power of the equivalent centralized energy storage; P irate is the rated power of energy storage device i, N is the total number of energy storage devices;

[0121] The equivalent centralized rated capacity of distributed energy storage is:

[0122]

[0123] Where S allrate is the rated capacity of the equivalent centralized energy storage device; S irate is the rated capacity of energy storage device i;

[0124] Constraints on the equivalent centralized charging efficiency and equivalent centralized discharging efficiency of distributed energy storage:

[0125]

[0126]

[0127] Among them, η allch and ηalldis are the charging efficiency and discharging efficiency of the equivalent centralized energy storage, respectively; ηich and ηidis are the charging efficiency and discharging efficiency of energy storage device i, respectively.

[0128] SOC constraints of distributed energy storage:

[0129] SOC min ≤SOC all ≤SOC max

[0130] SOC first =SOC last

[0131] Among them, SOC all is the SOC value of the equivalent centralized energy storage; SOC min and SOC max are the minimum and maximum values ​​allowed for the equivalent centralized energy storage SOC; SOC first and SOC last They are the initial SOC and final SOC of the equivalent centralized energy storage, respectively.

[0132] Step 4: Taking the equivalent centralized energy storage model established in step 3 as the control unit and minimizing the peak-to-valley difference of the distribution network load as the objective function for optimization, the charging and discharging power of the distributed energy storage is optimized.

[0133] Further preferably, the objective function is to minimize the load fluctuation of the distribution network, the clustered groups are used as control units, the charging and discharging power of the distributed energy storage is optimized, and the charging and discharging instructions of the distributed energy storage are issued.

[0134] (1) Based on the peak-shaving and valley-filling function of distributed energy storage in the distribution network, load fluctuation is used as the objective function of optimal scheduling. The objective function is to ensure the minimum load fluctuation:

[0135]

[0136] where f Δp is the load fluctuation of the distribution network in each period; K is the number of load power intervals; M j is the number of loads in load power interval j; P javrload is the average power within the load power interval j; P avrload It is the average power of the distribution network in all periods (0 o'clock-24 o'clock); P jess It is the total active output of the equivalent centralized energy storage in the load power interval j (this parameter is mainly affected by the capacity, efficiency and SOC of the distributed energy storage system). A positive value indicates discharge and a negative value indicates charging.

[0137] The constraints of the objective function include power flow constraints and equivalent centralized energy storage model, specifically:

[0138] 1) The power flow constraint is:

[0139]

[0140] Among them, P ji’s is the active power injected by node i' in load power interval j; U j’I’ is the voltage from node j' to node i', U j’l is the voltage from node j' to node l, G i’l and B il are the conductance and susceptance from node i' to node l respectively; θ i’l is the phase angle difference between node i' and node l; P ji’avrload is the active power of the node i' load in the load power interval j; P ji’ess is the active power of the energy storage installed at node i' in load power interval j.

[0141] 2) Charging and discharging power constraints of distributed energy storage systems:

[0142]

[0143] in,

[0144] P allrate is the rated power of the equivalent centralized energy storage; P irate is the rated power of energy storage device i, and N is the number of energy storage devices.

[0145] 3) The equivalent centralized rated capacity of distributed energy storage is:

[0146]

[0147] Where S allrate is the rated capacity of the equivalent centralized energy storage device; S irate is the rated capacity of energy storage device i.

[0148] 4) Constraints on the equivalent centralized charging efficiency and equivalent centralized discharging efficiency of distributed energy storage:

[0149]

[0150]

[0151] Among them, η allch and η alldis are the charging efficiency and discharging efficiency of equivalent centralized energy storage respectively; η ich and η idisare the charging efficiency and discharging efficiency of energy storage device i respectively.

[0152] 5) SOC constraints of distributed energy storage:

[0153] SOC min ≤SOC all ≤SOC max

[0154] SOC first =SOC last

[0155] Among them, SOC all is the SOC value of the equivalent centralized energy storage; SOC min and SOC max are the minimum and maximum values ​​allowed for the equivalent centralized energy storage SOC; SOC first and SOC last They are the initial SOC and final SOC of the equivalent centralized energy storage, respectively.

[0156] (2) Power allocation of distributed energy storage in aggregated energy storage systems:

[0157] Solving the objective function and its constraints, we can obtain the total active output P of the equivalent centralized energy storage model. jess , for P jess Carry out power allocation of each energy storage device in the equivalent centralized energy storage model to achieve optimal scheduling of charging and discharging power of distributed energy storage.

[0158] In the present invention, multiple energy storage units are aggregated into an equivalent centralized energy storage unit for use by the dispatch center. Therefore, in the dispatch optimization calculation, only one charge and discharge variable of the centralized energy storage is required. In order to reasonably allocate the power of each energy storage unit during the dispatch process and maintain the relative balance of SOC, the rated power and SOC of each energy storage unit are used to jointly determine its charge and discharge power. The specific determination method is as follows:

[0159] Charge:

[0160]

[0161] Discharge:

[0162]

[0163] Where P irate is the rated power of energy storage device i; P Jrate is the rated power of the energy storage device J; P ich and P idis are the charging and discharging power of energy storage device i respectively; P Jch and P Jdisare the charging and discharging power of energy storage device J respectively; P allch and P alldis is the total charging and discharging power required; N is the total number of energy storage devices; f ch (SOC i ) and f dis (SOC i ) are the SOC functions of charging and discharging of energy storage device i; f ch (SOC J ) and f dis (SOC J ) are the SOC functions of the energy storage device J for charging and discharging. ch (x) and f dis (x) are uniformly represented as the SOC functions of energy storage charging and discharging, where x is SOC i or SOC J .

[0164] P jess It is a value solved according to the objective function. With this value, the energy storage will take corresponding actions. If it is a negative value, the energy storage will be charged. If it is a positive value, the energy storage will be discharged. This value provides a reference for the action of the energy storage. When it is a negative value, the absolute value is equal to P allch , when it is positive, it is equal to P alldis .

[0165] In order to make the energy storage with lower SOC charge more and discharge less (or the energy storage with higher SOC charge less and discharge more), the sigmoid function is converted into the SOC function of charge and discharge through scaling and translation.

[0166] The specific SOC function expression for charging and discharging is as follows:

[0167]

[0168] Embodiment 2 of the present invention provides a distributed energy storage optimization scheduling system with minimum peak-to-valley difference in a distribution network, comprising:

[0169] Data acquisition module, used to collect power grid data, including load data and distributed energy storage data;

[0170] A mathematical model building module is used to establish a distributed energy storage mathematical model based on the collected data and analyze the distributed energy storage state parameters;

[0171] A clustering grouping module is used to cluster distributed energy storage devices according to state parameters of distributed energy storage to obtain K distributed energy storage clusters, and establish an equivalent centralized energy storage model for each cluster;

[0172] The optimization scheduling module is used to optimize the charging and discharging power of distributed energy storage by taking the equivalent centralized energy storage model as the control unit and minimizing the peak-to-valley difference of the distribution network load as the objective function of the optimization target.

[0173] A terminal comprises a processor and a storage medium; the storage medium is used to store instructions; the processor is used to operate according to the instructions to execute the steps of the method.

[0174] A computer-readable storage medium stores a computer program, which implements the steps of the method when executed by a processor.

[0175] The beneficial effects of the present invention are as follows:

[0176] 1. The present invention proposes a method for large-scale distributed energy storage to participate in the operation of the distribution network, which solves the problem of aggregate optimization of a large number of distributed energy storage resources connected to the power grid in the large-scale distributed energy storage scenario. Most existing studies have established a multi-cycle optimization scheduling model for distributed energy storage. However, when the number of optimization cycles and the amount of energy storage are large, many variables will appear in the optimization model, which greatly affects the optimization effect and time.

[0177] 2. The present invention reduces the variable dimension of distributed energy storage, reduces the difficulty of solving, is conducive to the reasonable scheduling of distributed energy storage, reduces the peak-to-valley difference in distribution network operation, ensures the safe and stable operation of the power grid, and improves the utilization rate of distributed energy storage during the regulation and operation process.

[0178] 3. The clustering algorithm of the present invention optimizes the selection of initial values. First, a seed data is randomly selected, and then the distance from other data points to this data point is calculated. Data points with a larger distance are selected to replace the randomly selected data points as new data points. The above steps are iterated until K data points are selected. This algorithm process solves the problem of slow convergence caused by random selection of initial data points, and better realizes efficient regulation of distributed energy storage.

[0179] The present disclosure may be a system, a method and / or a computer program product. The computer program product may include a computer-readable storage medium carrying computer-readable program instructions for causing a processor to implement various aspects of the present disclosure.

[0180] A computer-readable storage medium may be a tangible device that can hold and store instructions used by an instruction execution device. A computer-readable storage medium may be, for example, but not limited to, an electrical storage device, a magnetic storage device, an optical storage device, an electromagnetic storage device, a semiconductor storage device, or any suitable combination of the foregoing. More specific examples of computer-readable storage media (a non-exhaustive list) include: a portable computer disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), a static random access memory (SRAM), a portable compact disk read-only memory (CD-ROM), a digital versatile disk (DVD), a memory stick, a floppy disk, a mechanical encoding device, such as a punch card or a raised structure in a groove on which instructions are stored, and any suitable combination of the foregoing. As used herein, a computer-readable storage medium is not to be interpreted as a transient signal per se, such as a radio wave or other freely propagating electromagnetic wave, an electromagnetic wave propagating through a waveguide or other transmission medium (e.g., a light pulse through a fiber optic cable), or an electrical signal transmitted through a wire.

[0181] The computer-readable program instructions described herein can be downloaded from a computer-readable storage medium to each computing / processing device, or downloaded to an external computer or external storage device via a network, such as the Internet, a local area network, a wide area network, and / or a wireless network. The network can include copper transmission cables, optical fiber transmissions, wireless transmissions, routers, firewalls, switches, gateway computers, and / or edge servers. The network adapter card or network interface in each computing / processing device receives the computer-readable program instructions from the network and forwards the computer-readable program instructions for storage in the computer-readable storage medium in each computing / processing device.

[0182] The computer program instructions for performing the operation of the present disclosure may be assembly instructions, instruction set architecture (ISA) instructions, machine instructions, machine-related instructions, microcode, firmware instructions, state setting data, or source code or object code written in any combination of one or more programming languages, including object-oriented programming languages, such as Smalltalk, C++, etc., and conventional procedural programming languages, such as "C" language or similar programming languages. Computer-readable program instructions may be executed completely on a user's computer, partially on a user's computer, as an independent software package, partially on a user's computer, partially on a remote computer, or completely on a remote computer or server. In the case of a remote computer, the remote computer may be connected to the user's computer via any type of network, including a local area network (LAN) or a wide area network (WAN), or may be connected to an external computer (e.g., using an Internet service provider to connect via the Internet). In some embodiments, an electronic circuit, such as a programmable logic circuit, a field programmable gate array (FPGA), or a programmable logic array (PLA), may be customized by utilizing the state information of the computer-readable program instructions, and the electronic circuit may execute the computer-readable program instructions, thereby realizing various aspects of the present disclosure.

[0183] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention rather than to limit it. Although the present invention has been described in detail with reference to the above embodiments, ordinary technicians in the relevant field should understand that the specific implementation methods of the present invention can still be modified or replaced by equivalents, and any modifications or equivalent replacements that do not depart from the spirit and scope of the present invention should be covered within the scope of protection of the claims of the present invention.

Claims

1. A distributed energy storage optimization scheduling method with minimum peak-to-valley difference in distribution network, characterized in that: The method comprises: Step 1: Collect grid data, including load data and distributed energy storage data; Step 2: Establish a distributed energy storage mathematical model based on the data collected in step 1, and analyze the distributed energy storage state parameters; Step 3: Cluster the distributed energy storage devices according to the state parameters of the distributed energy storage obtained by the analysis in step 2 to obtain K distributed energy storage clusters, and establish an equivalent centralized energy storage model for each cluster; Step 4: Take the equivalent centralized energy storage model established in step 3 as the control unit, and take the minimization of the peak-to-valley difference of the distribution network load as the objective function for optimization, and optimize the charging and discharging power of the distributed energy storage.

2. A distributed energy storage optimization scheduling method with minimum peak-to-valley difference in distribution network according to claim 1, characterized in that: In step 1, the active power of the load in the power grid, the state of charge SOC, charging and discharging power, charging and discharging efficiency and rated capacity of the distributed energy storage are collected.

3. The distributed energy storage optimization scheduling method with the minimum peak-to-valley difference of the distribution network according to claim 1 is characterized by: In step 2, the distributed energy storage mathematical model established based on the data collected in step 1 is: SOC min ≤SOC i ≤SOC max Among them, SOC0, SOC t , SOC t-1 is the SOC value at the initial energy storage, t, and t-1; P ch and P dis They are respectively the charging and discharging power of energy storage from time t to time t-1; η ch and η dis are the energy storage charging and discharging efficiency respectively; Δt is a continuous charging and discharging period; S rat is the rated capacity of the energy storage; SOC min , SOC max are the minimum and maximum allowed values ​​of energy storage SOC, respectively; P dismax and P chmax are the maximum charging power and discharging power of the energy storage respectively.

4. The distributed energy storage optimization scheduling method with the minimum peak-to-valley difference of the distribution network according to claim 1 is characterized by: In step 2, the distributed energy storage state parameters in the distributed energy storage mathematical model include: power, state of charge, and available capacity.

5. The distributed energy storage optimization scheduling method with the minimum peak-to-valley difference of the distribution network according to claim 1 is characterized by: In step 3, the distributed energy storage devices are clustered according to the state parameters of the distributed energy storage to obtain K distributed energy storage clusters: Step 3.1: Use the state parameters of each energy storage device to form a feature vector, and use the feature vectors of n energy storage devices to form a sample in is the i-th point in the sample, representing the characteristic vector composed of the state parameters of energy storage device i, and Where V i1 、V i2 、V i3 for The three aggregated characteristic quantities represent the values ​​of the three state parameters of the energy storage device i; From the sample Randomly select a point As the initial aggregation center, where k is 1, 2, .., or n; Step 3.2: Calculate the distance D from the nearest initial clustering center for each point in the sample; Step 3.3: Add the farthest point as the next clustering center; Step 3.4: Repeat steps 3.2 and 3.3 until K clustering centers are selected; Step 3.5: Calculate the distance from each point in the sample to any aggregation center distance; Assign each point to the cluster corresponding to the aggregation center with the smallest distance, and update the aggregation center of each cluster: in for The updated aggregation center of the cluster, W m 、N m are the number of data points with existing aggregation centers and the number of total cluster centers respectively; Step 3.6: Repeat step 3.5 until the average function E converges, end clustering, and output K clusters, i.e., K distributed energy storage clusters, where the average function E is: Where p is all points in space, is the average value of cluster cm.

6. The distributed energy storage optimization scheduling method with the minimum peak-to-valley difference of the distribution network according to claim 1 is characterized by: In step 3, the equivalent centralized energy storage model established for each cluster is: SOC min ≤SOC all ≤SOC max SOCIETY first =SOC last Among them, P allrate is the rated power of the equivalent centralized energy storage, P irate is the rated power of energy storage device i, N is the total number of energy storage devices; S allrate is the rated capacity of the equivalent centralized energy storage; S irate is the rated capacity of energy storage device i; η allch and η alldis are the charging efficiency and discharging efficiency of equivalent centralized energy storage respectively; η ich and η idis are the charging efficiency and discharging efficiency of energy storage device i, respectively; SOC all is the SOC value of the equivalent centralized energy storage, SOC min and SOC max are the minimum and maximum values ​​allowed by the equivalent centralized energy storage SOC, SOC first and SOC last They are the initial SOC and final SOC of the equivalent centralized energy storage, respectively.

7. The distributed energy storage optimization scheduling method with the minimum peak-to-valley difference of the distribution network according to claim 1 is characterized by: In step 4, the equivalent centralized energy storage model is used as the control unit, and the objective function with the minimum peak-to-valley difference of the distribution network load as the optimization target is: Where f Δp is the peak-to-valley difference of the load in the distribution network at each period; K is the number of load power intervals; M j is the number of loads in load power interval j; P javrload is the average power within the load power interval j; P avrload is the average power of the distribution network in all periods; P jess It is the total active output of the equivalent centralized energy storage in the load power interval j. A positive value indicates discharge, and a negative value indicates charging.

8. The distributed energy storage optimization scheduling method with the minimum peak-to-valley difference of the distribution network according to claim 1 is characterized by: The constraint conditions of the objective function include power flow constraint and equivalent centralized energy storage model; The power flow constraint is: Among them, P ji,s is the active power injected by node i in load power interval j; U j,I, is the voltage from node j to node i, U j,l is the voltage from node j to node l, G i,l and B il are the conductance and susceptance from node i to node l, respectively; θ i,l is the phase angle difference between node i and node l; P ji,avrload is the active power of the load at node i in load power interval j; P ji,ess is the active power of the energy storage installed at node i in the load power interval j.

9. The distributed energy storage optimization scheduling method with the minimum peak-to-valley difference of the distribution network according to claim 1 is characterized by: In step 4, the objective function and its constraints are solved to obtain the total active output P of the equivalent centralized energy storage model. jess , for P jess The power of each energy storage device in the equivalent centralized energy storage model is allocated to achieve the optimal scheduling of the charging and discharging power of the distributed energy storage. The specific allocation method is: P jess When it is a negative value, the charging power is distributed as follows: P jess When it is positive, the discharge power distribution is as follows: P jess When it is 0, no power allocation is performed; Where P irate is the rated power of energy storage device i; P Jrate is the rated power of the energy storage device J; P ich and P idis are the charging and discharging power of energy storage device i respectively; P Jch and P Jdis are the charging and discharging power of energy storage device J respectively; P allch and P alldis is the total charging and discharging power required; N is the total number of energy storage devices; f ch (SOC i ) and f dis (SOC i ) are the SOC functions of charging and discharging of energy storage device i; f ch (SOC J ) and f dis (SOC J ) are the SOC functions of the energy storage device J for charging and discharging, respectively.

10. A distributed energy storage optimization dispatching system with minimum peak-to-valley difference in distribution network, using the method described in any one of claims 1 to 9, characterized in that: The distributed energy storage optimization dispatching system comprises: Data acquisition module, used to collect power grid data, including load data and distributed energy storage data; A mathematical model building module is used to establish a distributed energy storage mathematical model based on the collected data and analyze the distributed energy storage state parameters; A clustering grouping module is used to cluster distributed energy storage devices according to state parameters of distributed energy storage to obtain K distributed energy storage clusters, and establish an equivalent centralized energy storage model for each cluster; The optimization scheduling module is used to optimize the charging and discharging power of distributed energy storage by taking the equivalent centralized energy storage model as the control unit and minimizing the peak-to-valley difference of the distribution network load as the objective function of the optimization target.

11. A terminal comprising a processor and a storage medium; characterized in that: The storage medium is used to store instructions; The processor is configured to operate according to the instructions to execute the steps of the method according to any one of claims 1-9.

12. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the program is executed by a processor, the steps of the method according to any one of claims 1 to 9 are implemented.

Images