Power distribution network energy storage system multi-scene scheduling method and device based on data driving

Abstract

The invention relates to a power distribution network energy storage system multi-scene scheduling method and device based on data driving, and the method comprises the following steps: obtaining the time sequence load data and distributed photovoltaic time sequence power generation data of each node of a power distribution network, and employing a data driving method to respectively establish an electrical load uncertainty set and a photovoltaic power generation uncertainty set; establishing multi-scene data required by non-planed emergency island operation; based on the uncertainty set and the multi-scene data, constructing a distributed energy storage scheduling model based on opportunity operation cost; and solving the distributed energy storage scheduling model, and obtaining the optimal reserve capacity and the auxiliary service response mode of the energy storage system. Compared with the prior art, the method has the advantages of being high in emergency response capability, good in model universality and the like, and the defects that in the prior art, the response capability is insufficient, and long-distance new energy consumption of the power transmission network level is not fully considered are overcome.

Description

Technical Field

[0001] The present invention belongs to the technical field of power system auxiliary services, and relates to a distribution network energy storage scheduling method, and in particular to a data-driven distribution network energy storage system multi-scenario scheduling method and device. Background Art

[0002] Power system ancillary services are the cornerstone for ensuring the safe and economical operation of the power system. In traditional power systems, these key services are mainly provided by core equipment such as power plants and capacitors, and play a vital role in maintaining the stability of the power grid. Ancillary services can be divided into several categories: first, grid congestion management ancillary services, which aim to ensure the smooth operation of the power grid and effectively prevent overloads; second, voltage control ancillary services, which ensure the quality and stability of power supply by accurately adjusting the voltage level; and finally, frequency regulation ancillary services, whose core goal is to maintain the stability of the grid frequency and ensure the smooth operation of the power system. At the distribution network level, the focus of ancillary services is on key tasks such as alleviating congestion, accurately controlling voltage, and reducing network losses.

[0003] However, with the large-scale grid connection of renewable energy sources such as wind power and photovoltaics, and the gradual elimination of traditional fossil fuel generators, the power system is facing unprecedented challenges. This change urgently requires us to find a solution that can flexibly regulate power and is cost-effective to effectively deal with problems such as voltage fluctuations and network congestion in the distribution network. In the current situation where the demand-side response mechanism and distributed power market are not yet fully mature, distributed battery energy storage systems are gradually emerging as an innovative alternative. The system can provide key reactive power support, voltage regulation, congestion relief, and unplanned island emergency operation capabilities for the distribution network, significantly improving the stability and regulation capabilities of the power grid.

[0004] At present, the methods of using energy storage systems to respond to grid ancillary services are mainly concentrated in two aspects: one is to respond to unplanned ancillary service needs, and the other is to respond to planned ancillary service needs. Among them, the dispatching method for unplanned ancillary service needs mainly solves frequency regulation, voltage regulation and black start problems; while the dispatching method for planned ancillary service needs mainly solves grid congestion and peak load reduction problems.

[0005] For the voltage regulation problem of unplanned auxiliary services, the paper "Real-time coordinated voltage control of pv inverters and energy storage for weak networks with high pv penetration" (Wang, L., Bai, F., Yan, R., and Saha, TK (2018). Power Syst. 33, 3383–3395.) realizes the voltage regulation function of distributed renewable energy active output lossless by adjusting the active and reactive output of the energy storage system, which can effectively avoid the occurrence of node overvoltage and low voltage scenarios in the distribution network. For the frequency regulation problem, the paper "Primary frequency control with bess considering adaptive soc recovery" (Tan, Z., Li, X., He, L., Li, Y., and Huang, J. (2020). Int. J. Electr. Power & Energy Syst. 117, 105588.) combines adaptive droop control and SoC recovery control methods to design a variable frequency controller to respond to the grid frequency regulation needs. For the frequency black start problem, the literature Performance evaluation of a BESS unit for black start and seamless islanding operation (Izadkhast, S., Cossent, R., Frías, P., García-González, P., and Rodríguez-Calvo, A. (2022). Energies 15, 1736.) designed a control method to evaluate the voltage and frequency support performance of the energy storage system during black start. For the network congestion problem of planned ancillary services, the literature "Stochastic planning and scheduling of energy storage systems for congestion management in electric power systems including renewable energy resources" (Hemmati, R., Saboori, H., and Jirdehi, MA (2017). Energy 133, 380–387.) proposed a decentralized control algorithm to manage distributed BESS, using the concept of multi-agent to realize the congestion signal communication between different areas in the power network.To address the problem of peak load reduction, the literature “Optimal peak shaving control using dynamic demand and feed-in limits for grid-connected pv sources with batteries” (Manojkumar, R., Kumar, C., Ganguly, S., and. JP (2021). IEEE Syst. J. 15, 5560–5570.) proposed a rule-based peak load control strategy, which uses dynamic load and grid-connected electricity price to predict the day-ahead load and photovoltaic profile, and then uses it to optimize the charging and discharging of BESS to reduce peak load.

[0006] Nevertheless, in previous research and practical applications, we found that energy storage systems still have significant shortcomings in providing multi-scenario auxiliary services. A major problem is that the energy storage system's response capability to planned and unplanned auxiliary services is not ideal. At present, energy storage systems mainly rely on the difference between peak and valley electricity prices for arbitrage operations, and are committed to minimizing network losses and peak loads on the distribution network side. Although this mode can meet the needs of planned auxiliary services, once a large number of unplanned reactive power compensation or node voltage regulation requirements appear, it may lead to conflicts in working modes. At the same time, considering that unplanned auxiliary service demands such as reactive power loss, low voltage, and emergency island operation are highly random and time-uncertain, energy storage systems may not be able to respond in time due to insufficient power reserves. If the influence of factors such as the aging of the energy storage system is added, these problems of insufficient response capabilities will be more serious, posing a potential threat to the stable operation of the power grid.

[0007] In addition, the current research on energy storage operation based on the distribution system has not fully considered the problem of accommodating remote renewable energy at the transmission network level. Specifically, when the load of the distribution network is stable or low, the unit combination model based on the minimum power generation cost usually recommends shutting down coal-fired or gas-fired units with higher marginal power generation costs and choosing lower-cost renewable energy generators for power production. However, if the renewable energy power source on the transmission network side is far away from the distribution center, it may lead to insufficient reactive power compensation on the distribution side. In order to cope with this situation, the dispatching side may choose to keep the coal-fired or gas-fired units on the load side running, and implement wind and solar abandonment strategies for remote renewable energy power sources, thereby causing the problem of unsalable remote renewable energy power generation. This type of distribution network-side auxiliary service scenario for the purpose of accommodating remote renewable energy power sources has not been fully considered in previous studies and patents. Summary of the invention

[0008] The purpose of the present invention is to provide a data-driven multi-scenario scheduling method and device for a distribution network energy storage system in order to overcome the defects of the above-mentioned prior art, such as insufficient response capability and failure to fully consider the problem of accommodating long-distance renewable energy at the transmission network level.

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

[0010] A data-driven multi-scenario dispatching method for a distribution network energy storage system comprises the following steps:

[0011] Obtain the time-series load data of each node in the distribution network and the time-series power generation data of distributed photovoltaics, and use the data-driven method to establish the uncertainty set of power load and the uncertainty set of photovoltaic power generation respectively;

[0012] Establish multi-scenario data for unplanned emergency island operation needs;

[0013] Based on the uncertainty set and multi-scenario data, a distributed energy storage scheduling model based on opportunity operating cost is constructed, wherein the objective function of the distributed energy storage scheduling model considers the marginal opportunity cost of planned ancillary services and unplanned ancillary services, and the constraint conditions include reactive power constraints of the energy storage system supporting remote new energy consumption services;

[0014] Solve the distributed energy storage scheduling model to obtain the optimal backup capacity and auxiliary service response mode of the energy storage system.

[0015] Furthermore, the time-series load data of each node in the distribution network includes load active power data and load reactive power data.

[0016] Furthermore, based on the acquired data, the process of establishing the uncertainty set using a data-driven approach includes:

[0017] Extract the data of each node at time t and store it in a new sequence;

[0018] Using a data-driven method to fit the new series to obtain a probability density function;

[0019] Based on the preset confidence interval rolling mining, the uncertainty set of all nodes in the distribution network at each sampling time is obtained, and the probability that the uncertainty set satisfies the probability density function is greater than or equal to the confidence level.

[0020] Furthermore, the multi-scenario data for establishing unplanned emergency island operation requirements specifically includes:

[0021] A generative model is established to simulate the unplanned emergency island operation scenario, and a 0, 1 random number assignment method is used to generate an unplanned emergency island operation scenario matrix as the multi-scenario data.

[0022] Furthermore, in the unplanned emergency island operation scenario matrix, the emergency island operation demand is generated only once in each unplanned emergency island operation scenario.

[0023] Furthermore, the objective function expression of the distributed energy storage scheduling model is:

[0024]

[0025] Among them, VA b The charging and discharging efficiency of the energy storage system, VA LR To benefit the energy storage system from participating in remote renewable energy power supply, VA Isl Vl is the efficiency of energy storage system in responding to unplanned emergency island operation. o Opportunity loss when energy storage systems respond to unplanned ancillary services.

[0026] Furthermore, the energy storage system charging and discharging efficiency VA b The expression is:

[0027]

[0028] Among them, N bus is the total number of nodes in the system, T e,t is the real-time electricity price, T is the total number of moments, P bch,i,t represents the battery charging active power of node i at time t, P bds,i,t represents the battery discharge active power of node i at time t;

[0029] The energy storage system participates in the remote renewable energy power supply benefit VA LR The expression is:

[0030]

[0031] Among them, Q n,i,t represents the reactive power emitted / absorbed by the battery of node i at time t;

[0032] The energy storage system responds to unplanned emergency island operation benefits VA Isl The expression is:

[0033]

[0034] Among them, A el,t Indicates the unplanned auxiliary service start signal. When the system has no unplanned auxiliary service demand, A el,t =1, on the contrary, when unplanned ancillary service demand occurs, A el,t =R c ,R c is the proportion of important loads in the distribution network, T dis The cost of power outage for important loads, is the average load active power of node i;

[0035] The opportunity loss Vl when the energy storage system responds to unplanned auxiliary services o The expression is:

[0036] V o =Vl reserve +Vl inc-Chr

[0037] Among them, Vl reserve To reserve capacity for arbitrage losses, Vl inc-Chr The charging cost increases due to participation in voltage regulation and reactive power support.

[0038] Furthermore, the constraints of the distributed energy storage scheduling model include grid flow constraints, distributed photovoltaic constraints and energy storage constraints, wherein the energy storage constraints include reactive power constraints of the energy storage system supporting remote new energy consumption services.

[0039] Furthermore, the reactive power constraint of the energy storage system supporting remote new energy consumption service is expressed as:

[0040] A pq,t Q b,i,t ≥Q s,t

[0041] Among them, A pq,t Indicates the planned remote new energy consumption service response signal. When the system has no unplanned auxiliary service demand, A pq,t =1, on the contrary, when the demand for remote new energy consumption service appears, A pq,t =0,Q b,i,t Represents the reactive power emitted / absorbed by the battery of node i at time t.

[0042] Furthermore, the power grid flow constraints include node active power constraints, node reactive power constraints, node voltage constraints, line current constraints, and node voltage upper and lower limit constraints;

[0043] The energy storage constraints also include energy storage active and reactive constraints, energy storage system charging active constraints, energy storage system energy constraints and energy storage system SOC constraints.

[0044] The present invention also provides a data-driven multi-scenario dispatching device for a distribution network energy storage system, comprising:

[0045] The uncertainty set construction module is used to obtain the time-series load data of each node in the distribution network and the time-series power generation data of distributed photovoltaics, and use the data-driven method to establish the uncertainty set of power load and the uncertainty set of photovoltaic power generation respectively;

[0046] Multi-scenario data building module, used to build multi-scenario data for unplanned emergency island operation requirements;

[0047] A scheduling model building module is used to build a distributed energy storage scheduling model based on opportunity operating costs based on the uncertainty set and multi-scenario data. The objective function of the distributed energy storage scheduling model considers the marginal opportunity costs of planned auxiliary services and unplanned auxiliary services. The constraints include reactive power constraints of energy storage systems that support remote new energy consumption services.

[0048] The solution module is used to solve the distributed energy storage scheduling model to obtain the optimal spare capacity and auxiliary service response mode of the energy storage system.

[0049] Compared with the prior art, the present invention has the following beneficial effects:

[0050] (1) The present invention considers the marginal opportunity costs of planned ancillary services and unplanned ancillary services in the objective function of the scheduling model, improves the scheduling objective function and constraints of the energy storage system, solves the problem of insufficient response capability to unplanned ancillary services in traditional control methods, and improves response capability.

[0051] (2) The present invention adopts for the first time an energy storage scheduling target based on the opportunity benefit principle in the scheduling of distribution network energy storage systems, solving the problem of insufficient emergency response capabilities of traditional methods that only consider direct benefits.

[0052] (3) In the present invention, the design of the retention capacity is consistent with the aging target of the energy storage system. The nonlinear aging functions of different energy storage systems can be ignored in the solution process, which improves the versatility of the model. BRIEF DESCRIPTION OF THE DRAWINGS

[0053] Figure 1 It is a schematic diagram of the process of the present invention;

[0054] Figure 2 The load collection and photovoltaic daily average data used in this embodiment;

[0055] Figure 3 This is the energy storage mobilization strategy based on heating and working day load in this embodiment. DETAILED DESCRIPTION

[0056] The present invention is described in detail below in conjunction with the accompanying drawings and specific embodiments. This embodiment is implemented based on the technical solution of the present invention, and provides a detailed implementation method and specific operation process, but the protection scope of the present invention is not limited to the following embodiments.

[0057] Example 1

[0058] This embodiment provides a data-driven multi-scenario scheduling method for a distribution network energy storage system, comprising the following steps:

[0059] Step S1, obtaining the time series load data of each node in the distribution network and the time series power generation data of distributed photovoltaics, and using a data-driven method to establish a power load uncertainty set and a photovoltaic power generation uncertainty set respectively;

[0060] Step S2, establishing multi-scenario data for unplanned emergency island operation requirements;

[0061] Step S3: Based on the uncertainty set and multi-scenario data, a distributed energy storage scheduling model based on opportunity operating cost is constructed, wherein the objective function of the distributed energy storage scheduling model considers the marginal opportunity cost of planned auxiliary services and unplanned auxiliary services, and the constraint conditions include reactive power constraints of the energy storage system supporting remote new energy consumption services;

[0062] Step S4: Solve the distributed energy storage scheduling model to obtain the optimal backup capacity and auxiliary service response mode of the energy storage system.

[0063] In step S1, the acquired time series load data of each node in the distribution network includes load active power data and load reactive power data:

[0064] set up

[0065]

[0066] is the load active power data collected at node i in the distribution network, where P l,d,i,t It is the load active power data collected at the ith time on the dth day at the ith node of the distribution network;

[0067] set up

[0068]

[0069] is the load reactive power data collected at node i in the distribution network, where Q l,d,i,t It is the load reactive power data collected at the i-th moment on the d-th day at the i-th node of the distribution network.

[0070] The obtained distributed photovoltaic time-series power generation data is expressed as:

[0071]

[0072] Among them, P pv,d,i,t It is the photovoltaic power data collected at the i-th time on the d-th day at the i-th node of the distribution network.

[0073] Based on the acquired data, the process of establishing an uncertainty set using a data-driven method includes: extracting the data of each node at the tth moment and storing it in a new sequence, and fitting the new sequence with a data-driven method to obtain a probability density function; based on a preset confidence interval, rolling mining is performed to obtain an uncertainty set for all nodes in the distribution network at each sampling moment, and the probability that the uncertainty set satisfies the probability density function is greater than or equal to the confidence level.

[0074] Specifically, in this embodiment, the process of establishing the power load uncertainty set and the photovoltaic power generation uncertainty set includes:

[0075] 101) For the collected active load data P l At each sampling moment, an uncertainty set is established and the data at the tth moment (∈(1,t)) is extracted and stored in a new sequence P l,i,t In, P l,i,t = {P l,1,i,t ,P l,2,i,t …P l,d,i,t}, using data-driven methods to target P l,i,t Fitting probability density function to a sequence And use the confidence interval [p l,i,t,low ,p l,i,t,high Mining uncertainty sets is the average load active power of node i, is the average load active power of node i on day d, so that the probability density function is satisfied Probability α is the confidence level, and so on, rolling from time 1 to time t, the uncertainty set of active load data at each sampling time of all nodes in the distribution network is obtained;

[0076] 102) For the collected reactive load data Q l Establish an uncertainty set at each sampling moment, and extract the data at the tth moment (∈(1,t)) and store it in a new sequence Q l,i,t In, Q l,i,t = {Q l,1,i,t ,Q l,2,i,t …Q l,d,i,t}, using a data-driven approach to target Q l,i,t Fitting probability density function to a sequence And use the confidence interval [q i,i,t,low ,q l,i,t,high Mining uncertainty sets is the average load reactive power of node i, is the average load reactive power of node i on day d, so that the probability density function is satisfied Probability α is the confidence level; and so on, rolling from time 1 to time t, the uncertainty set of reactive load data at each sampling time of all nodes in the distribution network is obtained;

[0077] 103) For the collected photovoltaic power generation data P pv At each sampling moment, an uncertainty set is established and the data at the tth moment (∈(1,t)) is extracted and stored in a new sequence P pv,i,t In, P pv,i,t = {P pv,1,i,t ,P pv,2,i,t …P pv,d,i,t}, using data-driven methods to target P pv,i,t Fitting probability density function to a sequence And use the confidence interval [p pv,i,t,low ,pv pv,i,t,high Mining uncertainty sets is the average photovoltaic power generation power of node i, is the photovoltaic power generation of node i on the dth day, so that the probability density function is satisfied Probability α is the confidence level; and so on, rolling from time 1 to time t, the uncertainty set of photovoltaic power generation data at each sampling time of all nodes in the distribution network is obtained.

[0078] In step S2, the multi-scenario data for establishing unplanned emergency island operation requirements specifically includes:

[0079] A generative model is established to simulate the unplanned emergency island operation scenario, and a 0, 1 random number assignment method is used to generate an unplanned emergency island operation scenario matrix as the multi-scenario data.

[0080] In this embodiment, assuming that the daily sampling time T = {1, 2, 3, 4 ... t}, t is the maximum number of samples per day, the constructed unplanned emergency island operation scenario matrix is ​​expressed as:

[0081]

[0082] in, is a variable of 0 or 1, N s Indicates the total number of randomly generated scenes.

[0083] It should be noted that considering the high overall reliability of the power system, it is very rare for two faults to occur in one day. That is, only one emergency island operation demand is generated in each unplanned emergency island operation scenario. For example, in the first set of operation scenarios generated, S ils , 1 ={0,1,1,0…0}. In addition, considering that most faults in the power system can be eliminated in a short period of time, the emergency island operation time is within the interval [0.5,1].

[0084] In step S3, the objective function of constructing the distributed energy storage scheduling model takes into account the charging and discharging benefits of the energy storage system, the benefits of the energy storage system participating in the remote new energy power supply, the benefits of the energy storage system responding to unplanned emergency island operation, and the opportunity loss when the energy storage system responds to unplanned auxiliary services, thereby maximizing the opportunity benefits of the energy storage system. Among them:

[0085] Energy storage system charging and discharging benefit VA b The expression is:

[0086]

[0087] Among them, N bus is the total number of nodes in the system, T e,t is the real-time electricity price, T is the total number of moments, P bch,i,t represents the battery charging active power of node i at time t, P bds,i,t represents the battery discharge active power of node i at time t;

[0088] Energy storage system participates in remote renewable energy power supply benefits VA LR The expression is:

[0089]

[0090] Among them, Q b,i,t represents the reactive power emitted / absorbed by the battery of node i at time t;

[0091] Energy storage system response to unplanned emergency island operation benefits VA Isl The expression is:

[0092]

[0093] Among them, A el,t Indicates the unplanned auxiliary service start signal. When the system has no unplanned auxiliary service demand, A el,t =1, on the contrary, when unplanned ancillary service demand occurs, A el,t =R c ,R c is the proportion of important loads in the distribution network, T dis The cost of power outage for important loads, is the average load active power of node i;

[0094] Opportunity loss Vl when the energy storage system responds to unplanned ancillary services o The expression is:

[0095] V o =Vl reserve +Vl inc-Chr

[0096] Among them, Vl reserve To reserve capacity for arbitrage losses, Vl inc-Chr The charging cost increases due to participation in voltage regulation and reactive power support.

[0097] Then the objective function expression of the distributed energy storage scheduling model is:

[0098]

[0099] That is, maximize the opportunity cost of the scheduling model.

[0100] In step S3, the constraints of the distributed energy storage scheduling model are established based on the optimal power flow theory and second-order cone relaxation, including power grid power flow constraints, distributed photovoltaic constraints and energy storage constraints, wherein the energy storage constraints include the reactive power constraints of the energy storage system supporting the remote new energy consumption service, specifically:

[0101] 301) The power grid flow constraints include node active power constraints, node reactive power constraints, node voltage constraints, line current constraints, and node voltage upper and lower limit constraints.

[0102] Active power constraint of node i:

[0103]

[0104] P i,k,t represents the active power flowing from node i to node k at time t, P j,i,t represents the active power flowing from node j to node i at time t, P bch,i,t represents the battery charging active power of node i at time t, P bds,i,t represents the battery discharge active power of node i at time t, R j,i represents the line resistance between node j and node i, L j,i,t A represents the square of the current amplitude flowing from node j to node i at time t. el,t Indicates the unplanned auxiliary service start signal. When the system has no unplanned auxiliary service demand, A el,t =1, on the contrary, when unplanned ancillary service demand occurs, A el,t =R c , R c is the proportion of important loads in the distribution network;

[0105] Reactive power constraint of node i:

[0106]

[0107] Q i,k,t represents the reactive power flowing from node i to node k at time t, Q j,i,trepresents the reactive power flowing from node j to node i at time t, Q b,i,t represents the reactive power emitted / absorbed by the battery of node i at time t, X j,i represents the line reactance between node j and node i, L j,i,t represents the square of the current amplitude flowing from node j to node i at time t, represents the photovoltaic reactive power of node i;

[0108] Voltage constraint at node i:

[0109] V j,t =V i,t -2(R i,j P i,j,t +X i,j Q i,j,t )+(R i,j 2 +X i,j 2 )L i,j,t

[0110] V i,t , V j,t represents the square of the voltage amplitude between node i and the connected node j;

[0111] Line ij current constraint:

[0112]

[0113] L i,j,t ≤I max,i,j 2

[0114] I max,i,j represents the maximum current carrying capacity of the line between node i and node j, ‖‖ 2 The symbol indicates the two-norm calculation;

[0115] Node voltage upper and lower limit constraints:

[0116]

[0117] V min ,V max They represent the lower and upper voltage limits in the distribution network respectively.

[0118] 302) Distributed photovoltaic constraints:

[0119]

[0120] Q pv,i,t represents the photovoltaic reactive power at the i-th node at time t, S pvmax,i Represents the rated apparent power of the PV at the i-th node.

[0121] 303) Energy storage constraints include active and reactive power constraints of energy storage, active power constraints when the energy storage system is charging, energy storage constraints of the energy storage system, SOC constraints of the energy storage system and reactive power constraints of the energy storage system for supporting remote new energy consumption services.

[0122] Active and reactive energy storage constraints:

[0123]

[0124] S bcmax,i represents the rated power of the energy storage system converter at node i;

[0125] Active power constraints when charging the energy storage system:

[0126] 0≤P bch,i,t -P bds,i,t ≤P bmax,i

[0127] P bmax,i represents the maximum active output power of the energy storage system at node i;

[0128] Energy storage system energy storage constraints:

[0129] B s,i,t =B s,i,t-1 +Δt(P bch,i,t *α--P bds,i,t / α)

[0130] B s,i,t represents the current power of the energy storage system at node i at time t (in MWh), Δt is the daily sampling interval of the system, for example, 144 data points are sampled per day, (hours), α is the charging and discharging efficiency of the energy storage system;

[0131] Energy storage system SOC constraints:

[0132] Soc i,t =B s,i,t / B max,i,t

[0133] Soc min +A eb,t Soc reserve.i,t ≤Soc i,t ≤Soc max

[0134] Soc i,t is the charge state of the energy storage system at node i at time t, Soc reserve.i,t It is the emergency capacity of the energy storage system. The emergency capacity does not participate in the response plan type auxiliary service. min is the minimum charge state of the energy storage system, Soc maxis the maximum charge state of the energy storage system, A eb,t Indicates the unplanned ancillary service response signal. When the system has no unplanned ancillary service demand, A eb,t =1, on the contrary, when unplanned ancillary service demand occurs, A eb,t =0;

[0135] Reactive power constraints for energy storage systems supporting remote renewable energy consumption services:

[0136] A pq,t Q b,i,t ≥Q s,t

[0137] A pq,t Indicates the planned remote new energy consumption service response signal. When the system has no unplanned auxiliary service demand, A pq,t =1, on the contrary, when the demand for remote new energy consumption service appears, A pq,t =0.

[0138] In step S4, an optimization framework considering uncertainty constraints is used to solve the distributed energy storage scheduling model, to explore the optimal backup capacity and auxiliary service response mode of the energy storage system, to schedule the energy storage system, and to maximize the grid-side benefits brought by the energy storage system.

[0139] In this embodiment, the IEEE 33-node distribution network is used for simulation verification, in which the load and distributed photovoltaic data use the photovoltaic power generation data of a certain area in Australia for one year, with a sampling frequency of 15 minutes. In this embodiment, 4 groups of distributed photovoltaics are connected to nodes 4, 15, 21 and 32, and the capacity of a single distributed photovoltaic group is 1MW, and the rated voltage of the access node is 11kV; 2 groups of energy storage systems are connected to nodes 2 and 33, and the capacity of a single energy storage system is 1MWh, and the inverter capacity is 1.2MVA. In order to protect the energy storage system, the upper and lower boundaries of SOC are set to be within the range of [0.1, 0.9].

[0140] Depend on Figure 3 It can be obtained that, in this embodiment, the energy storage system only responds to the absolute peak process for 6 hours in the process of responding to peak shaving, and ensures that the energy storage system retains about 47% of the backup capacity to respond to the occurrence of unplanned emergency islanding events. When the unplanned emergency islanding event occurs, the energy storage discharge power increases to 0.5MW to provide power supply for important loads in the island. In addition, in this embodiment, the remote new energy consumption signal appears above 47% SOC. On this basis, the energy storage system sacrifices the active charging and discharging process, fully compensates for the reactive power demand in the system, and improves the remote new energy consumption capacity.

[0141] In summary, the energy storage scheduling strategy specified by the present invention has multi-scenario and multi-dimensional response capabilities for power grid auxiliary services. The design of the backup capacity avoids the problem of insufficient response capability when responding to planned and unplanned auxiliary service demands. In addition, if the daily probability of unplanned emergency island demand is 1%, compared with the single-day full deep charge and discharge model, the design of the reserved capacity increases the number of battery cycles by 2.3 times. The reserved capacity design target is consistent with the battery life protection target, and nonlinear energy storage aging constraints are not embedded in the scheduling model, which enhances the versatility of the present invention.

[0142] If the above method is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, or the part that contributes to the prior art or the part of the technical solution, can be embodied in the form of a software product. The computer software product is stored in a storage medium, including several instructions for a computer device (which can be a personal computer, a server, or a network device, etc.) to perform all or part of the steps of the method described in each embodiment of the present invention. The aforementioned storage medium includes: U disk, mobile hard disk, read-only memory (ROM, Read-Only Memory), random access memory (RAM, Random Access Memory), disk or optical disk and other media that can store program codes.

[0143] Example 2

[0144] This embodiment provides a data-driven multi-scenario scheduling device for a distribution network energy storage system, including: an uncertainty set construction module, used to obtain time-series load data of each node in the distribution network and distributed photovoltaic time-series power generation data, and use a data-driven method to establish a power load uncertainty set and a photovoltaic power generation uncertainty set respectively; a multi-scenario data construction module, used to establish multi-scenario data of unplanned emergency island operation requirements; a scheduling model construction module, used to construct a distributed energy storage scheduling model based on opportunity operation cost based on the uncertainty set and multi-scenario data, the objective function of the distributed energy storage scheduling model considers the marginal opportunity cost of planned auxiliary services and unplanned auxiliary services, and the constraint conditions include reactive power constraints of the energy storage system that supports remote new energy consumption services; a solution module, used to solve the distributed energy storage scheduling model to obtain the optimal backup capacity of the energy storage system and the auxiliary service response method.

[0145] The rest is the same as in Example 1.

[0146] It will be appreciated by those skilled in the art that embodiments of the present invention may be provided as methods, systems, or computer program products. Therefore, the present invention may take the form of a complete hardware embodiment, a complete software embodiment, or an embodiment combining software and hardware. Moreover, the present invention may take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program codes. The schemes in the embodiments of the present invention may be implemented in various computer languages, for example, object-oriented programming language Java and literal scripting language JavaScript, etc.

[0147] These computer program instructions may also be stored in a computer-readable memory capable of directing a computer or other programmable data processing device to operate in a specific manner, so that the instructions stored in the computer-readable memory produce an article of manufacture comprising an instruction device, which implements the process Figure 1 A process or multiple processes and / or boxes Figure 1 A function specified in one or more boxes.

[0148] These computer program instructions can also be loaded onto a computer or other programmable data processing device so that a series of operating steps are executed on the computer or other programmable device to produce a computer-implemented process, thereby providing instructions for implementing the process. Figure 1 A process or multiple processes and / or boxes Figure 1 The steps for the functions specified in one or more boxes.

[0149] Although the preferred embodiments of the present invention have been described, those skilled in the art may make other changes and modifications to these embodiments once they have learned the basic creative concept. Therefore, the appended claims are intended to be interpreted as including the preferred embodiments and all changes and modifications that fall within the scope of the present invention.

[0150] Obviously, those skilled in the art can make various changes and modifications to the present invention without departing from the spirit and scope of the present invention. Thus, if these modifications and variations of the present invention fall within the scope of the claims of the present invention and their equivalents, the present invention is also intended to include these modifications and variations.

Claims

1. A data-driven multi-scenario scheduling method for distribution network energy storage system, characterized in that: The following steps are involved: Obtain the time-series load data of each node in the distribution network and the time-series power generation data of distributed photovoltaics, and use the data-driven method to establish the uncertainty set of power load and the uncertainty set of photovoltaic power generation respectively; Establish multi-scenario data for unplanned emergency island operation needs; Based on the uncertainty set and multi-scenario data, a distributed energy storage scheduling model based on opportunity operating cost is constructed, wherein the objective function of the distributed energy storage scheduling model considers the marginal opportunity cost of planned ancillary services and unplanned ancillary services, and the constraint conditions include reactive power constraints of the energy storage system supporting remote new energy consumption services; Solve the distributed energy storage scheduling model to obtain the optimal backup capacity and auxiliary service response mode of the energy storage system.

2. The data-driven multi-scenario scheduling method for distribution network energy storage system according to claim 1 is characterized in that: The time series load data of each node in the distribution network includes load active power data and load reactive power data.

3. The data-driven multi-scenario scheduling method for distribution network energy storage system according to claim 1 is characterized in that: Based on the acquired data, the process of establishing the uncertainty set using a data-driven approach includes: Extract the data of each node at time t and store it in a new sequence; Using a data-driven method to fit the new series to obtain a probability density function; Based on the preset confidence interval rolling mining, the uncertainty set of all nodes in the distribution network at each sampling time is obtained, and the probability that the uncertainty set satisfies the probability density function is greater than or equal to the confidence level.

4. The data-driven multi-scenario scheduling method for distribution network energy storage system according to claim 1 is characterized in that: The multi-scenario data for establishing unplanned emergency island operation requirements specifically includes: A generative model is established to simulate the unplanned emergency island operation scenario, and a 0, 1 random number assignment method is used to generate an unplanned emergency island operation scenario matrix as the multi-scenario data.

5. The data-driven multi-scenario scheduling method for distribution network energy storage system according to claim 4 is characterized in that: In the unplanned emergency island operation scenario matrix, the emergency island operation demand is generated only once in each unplanned emergency island operation scenario.

6. The data-driven multi-scenario scheduling method for distribution network energy storage system according to claim 1 is characterized in that: The objective function expression of the distributed energy storage scheduling model is: Among them, VA b The charging and discharging efficiency of the energy storage system, VA LR To benefit the energy storage system from participating in remote renewable energy power supply, VA Isl Vl is the efficiency of energy storage system in responding to unplanned emergency island operation. o Opportunity loss when energy storage systems respond to unplanned ancillary services.

7. The data-driven multi-scenario scheduling method for distribution network energy storage system according to claim 6 is characterized in that: The energy storage system charging and discharging benefit VA b The expression is: Among them, N bus is the total number of nodes in the system, T e,t is the real-time electricity price, T is the total number of moments, P bch,i,t represents the battery charging active power of node i at time t, P bds,i,t represents the battery discharge active power of node i at time t; The energy storage system participates in the remote renewable energy power supply benefit VA LR The expression is: Among them, Q b,i,t represents the reactive power emitted / absorbed by the battery of node i at time t; The energy storage system responds to unplanned emergency island operation benefits VA Isl The expression is: Among them, A el,t Indicates the unplanned auxiliary service start signal. When the system has no unplanned auxiliary service demand, A el,t =1, on the contrary, when unplanned ancillary service demand occurs, A el,t =R c ,R c is the proportion of important loads in the distribution network, T dis The cost of power outage for important loads, is the average load active power of node i; The opportunity loss Vl when the energy storage system responds to unplanned auxiliary services o The expression is: Fl o =Vl reserve +Fl inc-Chr Among them, Vl reserve To reserve capacity for arbitrage losses, Vl inc-Chr The charging cost increases due to participation in voltage regulation and reactive power support.

8. The data-driven multi-scenario scheduling method for distribution network energy storage system according to claim 1 is characterized in that: The constraints of the distributed energy storage scheduling model include grid flow constraints, distributed photovoltaic constraints and energy storage constraints, wherein the energy storage constraints include reactive power constraints of the energy storage system supporting remote new energy consumption services.

9. The data-driven multi-scenario scheduling method for distribution network energy storage system according to claim 1 or 8, characterized in that: The reactive power constraint of the energy storage system supporting remote new energy consumption service is expressed as: A pq，t Q b，i,t ≥Q s,t Among them, A pq,t Indicates the planned remote new energy consumption service response signal. When the system has no unplanned auxiliary service demand, A pq,t =1, on the contrary, when the demand for remote new energy consumption service appears, A pq,t =0,Q b,i,t Represents the reactive power emitted / absorbed by the battery of node i at time t.

10. A data-driven multi-scenario dispatching device for distribution network energy storage system, characterized in that: include: The uncertainty set construction module is used to obtain the time-series load data of each node in the distribution network and the time-series power generation data of distributed photovoltaics, and use the data-driven method to establish the uncertainty set of power load and the uncertainty set of photovoltaic power generation respectively; Multi-scenario data building module, used to build multi-scenario data for unplanned emergency island operation requirements; A scheduling model building module is used to build a distributed energy storage scheduling model based on opportunity operating costs based on the uncertainty set and multi-scenario data. The objective function of the distributed energy storage scheduling model considers the marginal opportunity costs of planned auxiliary services and unplanned auxiliary services. The constraints include reactive power constraints of energy storage systems that support remote new energy consumption services. The solution module is used to solve the distributed energy storage scheduling model to obtain the optimal spare capacity and auxiliary service response mode of the energy storage system.

Images