A peak-shaving layered optimization control method for energy storage cluster-assisted thermal power units

Abstract

The present invention belongs to the field of energy storage cluster assisted thermal power peak regulation. Specifically for wind power high penetration areas, a layered optimization control method for peak regulation of energy storage cluster assisted thermal power units is proposed; the present invention proposes a two-layer optimization control strategy for peak regulation of energy storage cluster assisted thermal power units in high-proportion wind power systems. The upper layer is a thermal storage peak regulation optimization control strategy based on the peak regulation surplus coefficient of each energy storage, which obtains the optimal allocation of peak regulation resources and system flexibility between energy storage clusters and thermal power units in each time period. The lower layer is an optimization control strategy for each energy storage power station under multi-objective constraints considering the life of energy storage, and the optimal output of each energy storage power station is obtained under the triple objectives. Simulation analysis and comparison of dynamic regulation effects based on the above methods demonstrate the orderliness and scientificity of this method.

Description

Technical Field

[0001] The present invention belongs to the field of energy storage cluster-assisted thermal power peak regulation, and specifically proposes a layered optimization control method for peak regulation of thermal power units assisted by energy storage clusters for areas with high wind power penetration. Background Art

[0002] Accelerate the construction of a new power system that adapts to the proportion of new energy to gradually increase. However, due to the uncertainty of wind power output, its large-scale grid connection will inevitably lead to a substantial increase in the net load peak-to-valley difference of the system. In order to maintain the system source-load power balance and maximize the level of new energy consumption, the demand for peak-shaving resources in the power system is increasing day by day. With the decline in battery costs, energy storage, as a fast-responding flexibility regulation resource, can effectively assist thermal power units in participating in system peak regulation, provide flexibility for the system, and alleviate the system peak regulation pressure. With the large-scale commissioning of wind farms in the power system and the introduction of various favorable energy storage policies, energy storage power stations in the power system have exploded in the past two years, and energy storage clusters have been formed in many regions, but there are problems of low utilization and poor returns. The regulation strategy of energy storage clusters has become the key.

[0003] Therefore, a new technical solution is urgently needed in the prior art to solve this problem. Summary of the invention

[0004] The purpose of the present invention is to propose a hierarchical optimization control strategy for peak-shaving of energy storage clusters to assist thermal power units in order to cope with the severe challenges brought by large-scale wind power grid connection to the peak-shaving of the power system. In the thermal-storage regulation layer, a thermal-storage peak-shaving optimization control strategy based on the peak-shaving surplus coefficient of the energy storage cluster is proposed. By introducing the peak-shaving surplus coefficient of the energy storage, a reasonable reserve of energy storage peak-shaving output space is made to improve the overall peak-shaving effect; in the internal regulation layer of the energy storage cluster, in order to slow down the aging of energy storage power stations, an optimization control strategy for each energy storage power station under multi-objective constraints is proposed, taking into account technical and economic constraints such as network flow constraints, regulation consistency constraints, and output fairness constraints, and taking into account the slow decay of energy storage life, low node load rate balance, and low operating cost. The multi-objective model uses a genetic algorithm to calculate the initial output value of each energy storage power station, and iterates with the required output of energy storage in the upper-level peak-shaving task allocation to finally determine the optimal output of each energy storage power station.

[0005] In order to achieve the above object, the present invention provides the following technical solutions:

[0006] 1. A hierarchical optimization control strategy for peak load regulation of thermal power units assisted by energy storage clusters, comprising the following steps, which are performed in sequence:

[0007] Step 1: Analyze and model the peak load status of the system including the energy storage cluster.

[0008] 1) Single thermal power plant participates in peak load regulation, so as to obtain the peak load regulation demand in each peak load regulation period when single thermal power plant participates in peak load regulation;

[0009] 2) Thermal energy and energy storage jointly participate in peak load regulation to obtain the corresponding actions of energy storage in each peak load regulation period;

[0010] 3) analyzing the operation mode of the energy storage cluster to obtain multiple operation modes of the energy storage cluster;

[0011] 4) Modeling the operation status of the combined thermal-storage peak load regulation, and obtaining the peak load surplus coefficient of energy storage at each moment based on the established scenario probability model Step 2: Energy storage cluster assists thermal power units in peak load regulation with double-layer optimization.

[0012] Step 3: Formulate upper-level strategies: Optimize the peak-shaving control of thermal storage based on the peak-shaving surplus coefficient of each energy storage.

[0013] 1) Formulate a combined thermal-storage peak-shaving control strategy based on the energy storage peak-shaving surplus coefficient,

[0014] 2) Constructing a thermal-storage peak load power economic allocation model,

[0015] Step 4: Formulate lower-level strategies: Optimal control of each energy storage power station under multi-objective constraints

[0016] 1) Construct a multi-objective peak load optimization model between energy storage power stations

[0017] 2) Solving based on multi-objective genetic algorithm model

[0018] 3) Establishment of control effect evaluation indicators

[0019] Preferably, the single thermal power plant in step 1) of 1) participates in peak load regulation:

[0020] Figure 2 Figure 3 is a schematic diagram of thermal power units participating in peak regulation. When a single thermal power unit participates in peak regulation, the peak regulation demands in each peak regulation time period are as follows: in [t1-t2], the thermal power unit meets the peak regulation demand. In [t2-t3], due to the rapid change of wind power, the maximum ramp rate of the thermal power unit cannot track the load, resulting in an upward flexibility demand. In [t3-t4], the capacity of the thermal power unit is insufficient, resulting in an upward peak regulation demand. In the interval [t4-t5], the thermal power unit can track the equivalent load. In [t5-t6], the wind power generation decreases sharply, and the adjustment speed of the thermal power unit is slow, resulting in a downward flexibility demand. In [t6-t7], the equivalent load curve is lower than the minimum output of the thermal power unit, resulting in a downward peak regulation demand. In the interval [t7-t8], the downward ramping ability of the thermal power unit does not meet the demand. In the interval [t8-t9], the downward peak regulation ability of the thermal power unit is insufficient and cannot track the equivalent load.

[0021] Preferably, the fire-storage combination in step 1 (2) participates in peak load regulation:

[0022] Based on the power constraints and capacity constraints of energy storage power stations in actual operation, the energy storage cluster regulation capability model is constructed as shown in the following formula:

[0023]

[0024] In the formula, are the discharge power and charging power of energy storage station j at time t; SOC j (t) are the actual value, lower limit and upper limit of SOC of energy storage station j at time t; E N,j is the rated capacity of energy storage power station j; P N,j is the rated charging and discharging power of energy storage station j; are the discharge efficiency and charging efficiency of the energy storage station.

[0025] The regulation capability model of extending the output of the above single energy storage power station to a multi-type energy storage cluster is as follows:

[0026]

[0027] In the formula, is the maximum charging and discharging power of the energy storage cluster in the region at time t

[0028] The corresponding actions of energy storage in each peak load period and other time periods correspond to the following six situations:

[0029] ①[t1-t2] Energy storage does not work

[0030]

[0031] in, is the charging and discharging power of the energy storage system; in the interval [t1-t2], the thermal power unit can meet the equivalent load demand and the energy storage does not operate.

[0032] ②[t2-t3] Energy storage provides upward flexibility

[0033]

[0034] in, To meet the upward climbing demand; at time t2, due to the start and stop of the thermal power units, the maximum climbing rate of thermal power is insufficient, resulting in insufficient upward flexibility of the system in the [t2-t3] interval. At this time, the discharge of the energy storage cluster meets the upward flexibility demand.

[0035] ③[t3-t4] Energy storage provides peak load regulation resources

[0036]

[0037] In the [t3-t4] interval, thermal power units are unable to meet the peak load demand, and energy storage provides peak load resources to meet equivalent load operation.

[0038] ④[t4-t5] Thermal power charges the energy storage system

[0039]

[0040] In the interval [t4-t5], the thermal power units can track the equivalent load, but it is predicted that there will be a large demand for peak-shaving resources at the next moment. Therefore, excess thermal power resources are used to charge the energy storage cluster during this period to meet the power demand in the next period t5-t7.

[0041] ⑤[t7-t8] Energy storage provides downward adjustment flexibility

[0042]

[0043] in, To meet the climbing demand; in the interval [t7-t8], the climbing capacity of the thermal power unit cannot meet the demand. At this time, energy storage improves the flexibility of the system by absorbing excess wind power.

[0044] ⑥[t8-t9] Energy storage provides peak load reduction resources

[0045]

[0046] In the interval [t8-t9], the thermal power units have insufficient peak-shaving capacity and are unable to track the equivalent load, so the energy storage power station supplements the peak-shaving resources.

[0047] in, The up / down flexibility that energy storage plants should supplement; It is the up / down climbing ability of the thermal power unit; It is the maximum value of up / down ramp of thermal power unit; It is the upper / lower limit of thermal power unit output; It is the up / down peak-shaving resource that should be supplemented by the energy storage power station; is the total up / down peak regulation demand; P G (t) is the peak output of thermal power units; The system's thermal power units can provide charging power for the energy storage power station.

[0048] Preferably, the energy storage cluster operation mode analysis in step 1 (3) is as follows:

[0049] The energy storage cluster has a first operating mode and a second operating mode. When the energy storage cluster is in the first operating mode, the thermal power units operate at minimum output, and the energy storage makes up for the remaining peak-shaving demand. When the energy storage cluster is in the second operating mode, the output of the thermal power units is increased, and the output duration of the energy storage power station is extended from the original [t1-t2] to [t1-t3], thereby utilizing the remaining peak-shaving capacity of the thermal power units to increase the response capability of the energy storage cluster.

[0050] Preferably, the fire-storage combined peak-shaving operation state modeling of step 1 (4) is as follows:

[0051] For the system, there are m (m=7) possible operating states at the next moment. The scenario set is Q, where 0 means the requirements are met and 1 means the requirements are not met. From left to right, they are upward peak demand, downward peak demand, insufficient upward climbing, and insufficient downward climbing. For the system, the seven situations are converted into a matrix as shown in formula (9).

[0052]

[0053] Construct the fuzzy set χ of the probability distribution of the scene:

[0054]

[0055] In the formula, Q i is the i-th uncertain set; E R is the mathematical expectation; r i is the probability of being in the i-th uncertain scenario at the next moment; R is the probability distribution of the scenario variable ξ; χ 0 (Q m ) is ξ based on Q m The set of all probability distributions of .

[0056] According to the above scenario probability model, the probability of occurrence of each operating state can be obtained, and the obtained results can be used to calculate the energy storage peak regulation surplus coefficient at each time

[0057]

[0058] Surplus coefficient of energy storage power station The amount of electricity that the energy storage power station reserves for the next moment of insufficient peak-shaving demand. The system has up-slope and up-peak-shaving demand scenarios obtained by historical data analysis, which are Q1, Q2, and Q6, and the probability of the two scenarios is r1, r2, and r6; similarly, the scenarios and probabilities of down-slope and down-peak-shaving demand are Q3, Q4, Q5 and r3, r4, and r5; Q7 and r7 are the scenarios and probabilities when the thermal power units meet the peak-shaving demand and system flexibility. The up / down / no-adjustment demand sets are Ru={r1,r2,r6}, Rd={r3,r4,r5}, and Rb={r7}, respectively, and the internal elements are represented by r u 、r d 、r b Indicates that the total number of elements is u n d n , b n .

[0059] Preferably, the energy storage cluster assists the thermal power unit in peak load regulation double-layer optimization in step 2, specifically:

[0060] (1) Thermal-storage peak-shaving optimization control layer based on the peak-shaving surplus coefficient of the energy storage cluster: The system operating status is analyzed, and the system peak-shaving operation scenario is modeled based on historical data. The energy storage peak-shaving surplus coefficient of each time period is introduced to regulate the power of the energy storage cluster to assist the system peak-shaving and improve the system flexibility. A thermal-storage peak-shaving power economic allocation model considering the uncertainty of wind power is constructed, and the optimal allocation of peak-shaving resources between the energy storage cluster and the thermal power units is completed under the premise of meeting the economic objective function.

[0061] (2) Optimization control layer of each energy storage power station under multi-objective constraints: A multi-objective peak-shaving power optimization model between energy storage power stations within the energy storage cluster considering battery life is proposed, and the peak-shaving power demand of the energy storage cluster obtained by the upper layer is allocated among the multiple types of energy storage power stations in the lower layer. Under the consideration of network flow constraints and the balance of supply and demand of system peak-shaving resources, the energy storage cycle life coefficient, the balance of the load rate of the energy storage power station node and the operating cost of the energy storage power station are used as multi-objective functions to achieve the optimal peak-shaving output distribution among multiple types of energy storage power stations.

[0062] Preferably, in step 3, 1) a fire-storage combined peak load control strategy is formulated, specifically:

[0063] To simplify the formula, P xq (t) is the peak load demand at time t; ΔP xq (t) is the remaining peak load demand at time t; is the peak load power supplied by the energy storage power station at time t; is the climbing constraint of the thermal power unit; ΔP lhx (t) is the system flexibility requirement at time t; E S(t) is the power of the energy storage station at time t; SOC tj Prepare the required value for the energy storage SOC; The median value of the energy storage SOC preparation requirement; E N is the rated capacity of the energy storage power station; P S (t) is the output of the energy storage power station; Δt is the time interval between sampling points. The system operation status is shown in the figure below: Figure 6 As shown:

[0064] (1)Q1: When the thermal power unit's ramp-up capability is met and the peak-shaving capacity is less than the equivalent load demand, the remaining peak-shaving demand is supplied by the energy storage system:

[0065]

[0066] (2)Q2: When the thermal power unit has insufficient ramp-up capability but sufficient peak-adjustment capacity, the energy storage power station can supplement the flexibility of the peak-adjustment at that moment:

[0067]

[0068] (3)Q3: When the thermal power unit has insufficient ramp-down capability and the peak-shaving capacity is less than the equivalent load demand, the remaining peak-shaving demand is supplemented by the energy storage system:

[0069]

[0070] (4)Q4: When the ramp-down capability of the thermal power unit is met and the peak-down capacity is less than the equivalent load demand, the remaining peak-down resources are replenished by the energy storage system.

[0071]

[0072] (5)Q5: When the thermal power unit has insufficient ramp-down capability and the peak capacity is adjusted down to meet the equivalent load demand, the energy storage power station can supplement the downward adjustment flexibility at this moment:

[0073]

[0074] (6)Q6: When the thermal power unit has insufficient ramp-up capability and the upward peak capacity is less than the equivalent load demand, the energy storage power station can supplement the downward flexibility at that moment:

[0075]

[0076] (7)Q7: When the thermal power unit has sufficient ramp capability and the peak load regulation capacity meets the equivalent load demand:

[0077] a) When the energy storage peak regulation surplus coefficient At this time, the thermal power unit is used to charge the energy storage system so that the energy storage system meets SOC tj , there is a certain flexibility in upward adjustment. If the energy storage power station SOC>SOC tj Then SOC(t+1)=SOC(t), otherwise:

[0078]

[0079] b) when When there is a certain probability of insufficient upward adjustment, in order to avoid wasting resources, the energy storage power station does not need to be equal to the upward adjustment coefficient. If the energy storage power station Then SOC(t+1)=SOC(t), otherwise:

[0080]

[0081] c) when At this time, the next moment, the downward adjustment capacity is insufficient, and the energy storage power station does not act. S (t) = 0, mainly based on the output of thermal power units:

[0082]

[0083] Preferably, the step 3 (2) constructs a fire-storage peak load power economic allocation model, specifically:

[0084] (1) Objective function

[0085] The objective function aims to minimize the overall operating cost of the system, which includes the operating cost of thermal power generators and the penalty cost caused by wind curtailment and load loss, as follows:

[0086] min(F 1 )=F1+F2 (21)

[0087] Where F1 is the operating cost of thermal power units; F2 is the penalty cost caused by wind abandonment and load loss.

[0088]

[0089] Among them, i is the startup cost of generator set i; c i is the fixed operating cost of generator set i; a i , b iare the coal consumption coefficients of thermal power unit i; P G,i (t) is the output of thermal power unit i at time t; z i,t is the decision variable for starting and stopping the generator set; its value is 1 for starting and 0 for not starting; μ i,t It is the decision variable of the generator set state. When the value is 1, it means running, and when it is 0, it means shutdown.

[0090]

[0091] In the formula, c wast is the wind abandonment penalty coefficient, is the wind abandonment volume at time t, c loss is the load loss penalty coefficient, is the load loss at time t, where:

[0092]

[0093] Where P W (t) is the wind power generated by the wind farm at time t; P is the wind power connected to the grid at time t; G,i (t) is the output of thermal power plant i at time t; P S,j (t) is the output of energy storage power station j at time t; the total number of thermal power units and total number of energy storage power stations in the region are I, J, P L (t) is the grid load that needs to be met at time t.

[0094] (2) Constraints

[0095] 1) Power balance constraints

[0096]

[0097] 2) Constraints on thermal power units

[0098] Output upper and lower limits of thermal power units:

[0099]

[0100] Thermal power unit climbing constraints:

[0101]

[0102] Thermal power unit start and stop constraints:

[0103]

[0104] In the formula, are the minimum and maximum output of thermal power unit i respectively; is the climbing limit of thermal power unit i; u i,tis the start / stop status of thermal power unit i, 1 and 0 are start and stop respectively; t on ,t off The minimum startup and shutdown time.

[0105] 3) Energy storage constraints

[0106] Energy storage power station SOC:

[0107]

[0108] Energy storage power station charging and discharging upper and lower limit constraints:

[0109]

[0110] Constraints on the consistency of the power at the beginning and end of the energy storage power station:

[0111] SOC T =SOC0 (34)

[0112] In the formula, E j,t is the energy of energy storage station j at time t; E N is the rated capacity of the energy storage cluster; SOC0 is the SOC value of the energy storage cluster at the initial moment; SOC T It is the discharge terminal; it ensures the stable operation of the energy storage system.

[0113] 4) Wind curtailment constraints

[0114]

[0115] In the formula, To abandon the upper realm.

[0116] 5) Network operation constraints

[0117] Since the present invention studies the peak load dispatching strategy of the power system, the active power constraint of the transmission line needs to be considered.

[0118]

[0119] In the formula, is the upper limit of active power transmitted by line ab; P ab (t) is the active power flowing from node a to node b on line ab during period t, and is calculated as follows:

[0120]

[0121] Where, X an , X bn , X al , X bl is the node impedance matrix element of the network topology; x n 、x lis the reactance value of line ab at node n and node l; P n (t) is the active power of power node n at time t; D l (t) Active power aggregate of load node l.

[0122] Preferably, the multi-objective energy storage power station peak load optimization model in step 4 1) is specifically:

[0123] (1) Objective function

[0124] 1) Energy storage cycle life

[0125] The battery life decay rate can be expressed as:

[0126]

[0127] Where U j,max (DOD j (u)) represents the battery cycle life (number of times) corresponding to the u-th discharge depth at power station j.

[0128] 2) Load rate

[0129] The calculation formula for node load balance is as follows:

[0130]

[0131] Among them, μ j,t is the load rate of the j-th energy storage power station in period t; The load borne by this substation; μ ave,t is the average load rate of multiple energy storage power stations in period t.

[0132] 3) Energy storage operating costs

[0133] In order to significantly improve the economic operation efficiency of energy storage power stations, an energy storage power station operation cost model was constructed to help energy storage power stations maximize their economic benefits:

[0134]

[0135] In the formula, c price The operating price for energy storage.

[0136] (2) Constraints

[0137] 1) State of charge constraints:

[0138]

[0139] Where P S,j (t) is the output of energy storage power station j at time t; P S '(t) is the total power demand of the energy storage cluster.

[0140] 2) The energy storage operation constraints are shown in equations (32) and (33).

[0141] 3) Network operation constraints are shown in equations (36) and (37):

[0142] 4) Regulating consistency constraints

[0143] In order to ensure that the energy storage stations are charged and discharged at the same time, that is, the regulation consistency, the energy storage station power cannot have the reverse increase or decrease:

[0144]

[0145] Where ΔE S,j (t) is the capacity change of energy storage station j in period t; E S,j (t) is the capacity of energy storage station j at time t; E S,j (t-1) is the capacity of energy storage station j at t-1; ΔE S '(t) is the capacity change of the energy storage cluster in period t; E S '(t) is the capacity of the energy storage cluster at time t; E S '(t-1) is the capacity of the energy storage cluster at time t-1.

[0146] 5) Output fairness constraints

[0147] The real-time SOC of each energy storage power station is normalized to obtain the remaining capacity b of the energy storage power station at each moment. j (t), propose b bias,j (t), represents the remaining capacity of energy storage power station j and the average value b of each power station at time t-1 mean Deviation of (t-1):

[0148]

[0149] b bias,j (t)=(b j (t-1)+Δb j (t)-b mean (t-1)) 2 (44)

[0150] The calculation methods of each part are as follows:

[0151]

[0152] In the formula, Δt is the time interval between two moments; Δb j (t) is the change in the remaining capacity of the energy storage station j at time t; ΔSOC j (t) is the SOC change of energy storage station j at time t.

[0153] Therefore, the output fairness constraint is:

[0154]

[0155] In the formula, γ max It is the maximum value of total deviation allowed by the system.

[0156] Preferably, step 4 (2) is solved based on a multi-objective genetic algorithm model:

[0157] 1) For every two solutions belonging to the same Pareto front, equation (49) holds:

[0158]

[0159] For every solution x that belongs to the Pareto frontier S i , there is at least one solution x j , which is better than x in at least one objective function (here called n) i In other words, there is no optimal solution on the Pareto front considering the set of all objective functions.

[0160]

[0161] 2) For each solution belonging to the upper Pareto frontier and the lower Pareto frontier, the following holds:

[0162]

[0163] Only one of the three objective functions in the present invention is retained, and the other two objective functions are transformed under the constraint of the constraint threshold ε1.

[0164] min(F (1) ,F (2) ) is converted into a constraint problem:

[0165]

[0166] Preferably, the control effect evaluation index of step 4 3) is specifically:

[0167] (1) SOC balance between energy storage power stations

[0168]

[0169] In the formula, SOC mean (t) is the average SOC value among energy storage clusters, SOC ph To balance the SOC among the power stations in the energy storage cluster, use SOC ph The SOC balance degree, operation fairness, and SOCph The smaller it is, the better the fairness.

[0170] (2) Peak load tracking

[0171]

[0172] Where P(t) is the total power output, H gz H is the degree of tracking of the system power supply to the load, gz The smaller the value, the better the strategy tracking.

[0173] (3) Unit energy consumption cost

[0174]

[0175] In the formula, C B (t) is the total energy consumption cost; P B (t) is the total energy consumption, including thermal power, energy storage and wind power, U cb It is the unit energy consumption cost during the total scheduling period. The smaller it is, the higher the energy utilization rate is.

[0176] Through the above design scheme, the present invention can bring the following beneficial effects:

[0177] 1. In order to cope with the severe challenges brought by large-scale wind power grid connection to the peak load regulation of the power system, the present invention proposes a hierarchical optimization control strategy for energy storage clusters to assist thermal power units in peak load regulation. At the thermal-storage regulation layer, a thermal-storage peak load regulation optimization control strategy based on the peak load surplus coefficient of the energy storage cluster is proposed. By introducing the peak load surplus coefficient of the energy storage, reasonable space for energy storage peak load regulation output is reserved to improve the overall peak load regulation effect; at the internal regulation layer of the energy storage cluster, in order to slow down the aging of energy storage power stations, an optimization control strategy for each energy storage power station under multi-objective constraints is proposed, taking into account technical and economic constraints such as network flow constraints, regulation consistency constraints, and output fairness constraints, and taking into account the slow decay of energy storage life, low node load rate balance, and low operating cost. The multi-objective model uses a genetic algorithm to calculate the initial output value of each energy storage power station, and iterates with the required output of energy storage in the upper-level peak load regulation task allocation to finally determine the optimal output of each energy storage power station. Through example simulation, it is found that the strategy proposed in this paper can reduce the peak demand by 87.40MW during peak hours. Under the premise of fully absorbing new energy, the total operating cost is reduced by 3.62%, and the cycle life of each energy storage power station has increased to varying degrees, among which the lithium battery energy storage power station has increased by 4.38%. The effectiveness of the strategy proposed in this paper has been verified from both economic and peak-shaving technical aspects.

[0178] 2. The present invention proposes a two-layer optimization control strategy for peak load regulation of energy storage clusters assisting thermal power units in high-proportion wind power systems. The upper layer is a thermal storage peak load optimization control strategy based on the peak load surplus coefficient of each energy storage, which obtains the optimal allocation of peak load resources and system flexibility between energy storage clusters and thermal power units in each time period. The lower layer is an optimization control strategy for each energy storage power station under multi-objective constraints considering the energy storage life, and obtains the optimal output of each energy storage power station under the triple objectives. Simulation analysis and comparison of dynamic control effects based on the above methods demonstrate the orderliness and scientificity of this method. BRIEF DESCRIPTION OF THE DRAWINGS

[0179] Figure 1 This is a schematic diagram of the peak-shaving system of energy storage clusters assisting thermal power units;

[0180] Figure 2 This is a schematic diagram of peak load regulation of thermal power units;

[0181] Figure 3 This is a schematic diagram of energy storage assisting thermal power units in peak regulation;

[0182] Figure 4 Schematic diagram of different matching schemes;

[0183] Figure 5 This is the overall framework diagram of the double-layer control strategy;

[0184] Figure 6 It is a schematic diagram of the system operation status;

[0185] Figure 7 It is the relationship between the number of battery cycles and the depth of discharge;

[0186] Figure 8 It is a flow chart of the double-layer optimization control strategy;

[0187] Fig. 9 To improve the IEEE-39 node grid structure diagram;

[0188] Fig.10 It is the hourly output diagram of thermal power units;

[0189] Fig.11 This is the start and stop state diagram of the thermal power unit;

[0190] Fig.12 This is the thermal power peak regulation state analysis diagram;

[0191] Fig.13 is the surplus demand coefficient of the energy storage power station;

[0192] Fig.14 For the remaining peak-shaving demand;

[0193] Fig.15 The SOC change of the energy storage system;

[0194] Fig.16 This is the output diagram of each energy storage power station in Strategy 1;

[0195] Fig.17 The SOC diagram of each energy storage power station;

[0196] Fig.18 It is the energy storage cycle life diagram under different strategies;

[0197] Fig.19 To adjust the peak demand according to SOC tj and Change graph;

[0198] Fig. 20 As the operating cost increases with SOC tj and Change graph;

[0199] Fig.21 The energy storage cycle life varies with SOC tj and Change graph. DETAILED DESCRIPTION

[0200] The present invention will be further described in detail below in conjunction with specific implementation modes. The following examples are used to illustrate the present invention, but it should be understood that the protection scope of the present invention is not limited by the specific implementation modes.

[0201] The present invention is a multi-region peak-shaving resource optimization scheduling strategy considering network trends, including: analyzing and modeling the peak-shaving state of the system containing the energy storage cluster, proposing a double-layer optimization control strategy for peak-shaving of the energy storage cluster-assisted thermal power unit, building a multi-region peak-shaving simulation model using the MATLAB simulation platform, and performing simulation. The specific steps are:

[0202] Step 1: Analyze and model the peak load status of the system including the energy storage cluster.

[0203] The first step is to analyze the peak load regulation of single thermal power plants.

[0204] In a high-proportion wind power system, thermal power units, as a single peak-shaving resource, cannot meet the system's peak-shaving needs. The output of thermal power units in the system is hourly, and the hourly regulation scale cannot meet the actual equivalent load fluctuations. Figure 2This is a schematic diagram of the peak load regulation of thermal power units. First, the start and stop plan of the thermal power units is obtained through the power plant's day-ahead plan, so as to know the output range of the thermal power units in each period (gray shaded area), and then the uncertainty in the system is modeled through historical data. The more accurate equivalent load data of the day before 15 minutes is obtained by the iteration of the model (red curve), and the 15-minute thermal power unit output plan is re-formulated (as shown by the black curve in the figure). Due to the limitations of the upper / lower climbing rate and the upper / lower limit of the output, a single thermal power unit cannot fully track the equivalent load for output, so there is a peak load shortage (shown in the colored part).

[0205] Figure 2 This is a schematic diagram of thermal power units participating in peak load regulation. In each peak load regulation period: at [t1-t2], the thermal power units meet the peak load regulation demand.

[0206] [t2-t3] Due to the rapid change of wind power, the maximum ramp rate of thermal power units cannot track the load, resulting in an upward flexibility demand; at [t3-t4] the capacity of thermal power units is insufficient, resulting in an upward peak demand; [t5-t6] wind power generation drops sharply, and the adjustment speed of thermal power units is slow, resulting in a downward flexibility demand; [t6-t7] the equivalent load curve is lower than the minimum output of thermal power units, resulting in a downward peak demand.

[0207] The second step is to analyze the combined participation of fire and storage in peak load regulation.

[0208] Based on the power constraints and capacity constraints of energy storage power stations in actual operation, the energy storage cluster regulation capability model is constructed as shown in the following formula:

[0209]

[0210] In the formula, are the discharge power and charging power of energy storage station j at time t; SOC j (t) are the actual value, lower limit and upper limit of SOC of energy storage station j at time t; E N,j is the rated capacity of energy storage power station j; P N,j is the rated charging and discharging power of energy storage station j; are the discharge efficiency and charging efficiency of the energy storage station.

[0211] The regulation capability model of extending the output of the above single energy storage power station to a multi-type energy storage cluster is as follows:

[0212]

[0213] In the formula, is the maximum charging and discharging power of the energy storage cluster in the region at time t

[0214] The energy storage system can play the dual role of peak load resource provider and system flexibility provider in the peak load regulation process. The unit output plan is based on the hourly time scale, and the system flexibility is 15 minutes. The energy storage-assisted peak load regulation system of thermal power units is analyzed as follows: Figure 2 , there are 6 situations:

[0215] ① Energy storage does not work

[0216]

[0217] in, is the charging and discharging power of the energy storage system; in the interval [t1-t2], the thermal power unit can meet the equivalent load demand and the energy storage does not operate.

[0218] ② Energy storage provides upward flexibility

[0219]

[0220] in, To meet the upward climbing demand; at time t2, due to the start and stop of the thermal power units, the maximum climbing rate of thermal power is insufficient, resulting in insufficient upward flexibility of the system in the [t2-t3] interval. At this time, the discharge of the energy storage cluster meets the upward flexibility demand.

[0221] ③ Energy storage provides peak-shaving resources

[0222]

[0223] In the [t3-t4] interval, thermal power units are unable to meet the peak load demand, and energy storage provides peak load resources to meet equivalent load operation.

[0224] ④ Thermal power charges the energy storage system

[0225]

[0226] In the interval [t4-t5], the thermal power units can track the equivalent load, but it is predicted that there will be a large demand for peak-shaving resources at the next moment. Therefore, excess thermal power resources are used to charge the energy storage cluster during this period to meet the power demand in the next period t5-t7.

[0227] ⑤ Energy storage provides downward flexibility

[0228]

[0229] in, To meet the climbing demand; in the interval [t7-t8], the climbing capacity of the thermal power unit cannot meet the demand. At this time, energy storage improves the flexibility of the system by absorbing excess wind power.

[0230] ⑥ Energy storage provides peak-shaving resources

[0231]

[0232] In the interval [t8-t9], the thermal power units have insufficient peak-shaving capacity and are unable to track the equivalent load, so the energy storage power station supplements the peak-shaving resources.

[0233] in, The up / down flexibility that energy storage plants should supplement; It is the up / down climbing ability of the thermal power unit; It is the maximum value of up / down ramp of thermal power unit; It is the upper / lower limit of thermal power unit output; It is the up / down peak-shaving resource that should be supplemented by the energy storage power station; is the total up / down peak regulation demand; P G (t) is the peak output of thermal power units; The system's thermal power units can provide charging power for the energy storage power station.

[0234] The third step is to analyze the operation mode of the energy storage cluster.

[0235] The impact of the internal output distribution mode of the energy storage cluster on the peak load regulation effect and operation economy is analyzed. Figure 4 , sub-figures (a) and (d) show the operating time of the energy storage cluster under different modes. Mode 1 is that the thermal power unit operates at the minimum output, and the energy storage makes up for the remaining peak-shaving demand. Mode 2 makes full use of the surplus peak-shaving capacity of the thermal power unit to increase the output of the thermal power unit. The output time of the energy storage power station is extended from the original [t1-t2] to [t1-t3]. The remaining peak-shaving capacity of the thermal power unit is used to increase the response capability of the energy storage cluster, thereby improving the peak-shaving performance of the system.

[0236] Figure 4 Sub-figures (b) and (e) show the distribution of output among energy storage power stations within the energy storage cluster. Figure (c) shows that the operating cost of mode 2 is lower than that of mode 1, and figure (f) shows that the total attenuation degree of each energy storage power station in the energy storage cluster under operating mode 2 is less than that of mode 1.

[0237] The fourth step is to model the operation status of the combined peak load regulation of fire and storage.

[0238] For the system, there are m (m=7) possible operating states at the next moment, and the scenario set is Q, 0 means meeting the requirements, and 1 means not meeting the requirements. From left to right, they are upward peak demand, downward peak demand, insufficient upward climbing, and insufficient downward climbing. The above 7 situations are converted into a matrix as shown in formula (9).

[0239]

[0240] Based on a large amount of historical data, the occurrence probability information of each scene in each period is obtained, and then the fuzzy set χ of the scene occurrence probability distribution is constructed. Compared with the commonly used variables in probability distribution, the fuzzy set of occurrence probability distribution has lower calculation complexity and lower sensitivity to noise when considering factors such as wind power load, which is suitable for the actual operation of the system in this study:

[0241]

[0242] In the formula, Q i is the i-th uncertain set; E R is the mathematical expectation; r i is the probability of being in the i-th uncertain scenario at the next moment; R is the probability distribution of the scenario variable ξ; χ 0 (Q m ) is ξ based on Q m The set of all probability distributions of .

[0243] According to the above scenario probability model, the probability of occurrence of each operating state can be obtained, and the obtained results can be used to calculate the energy storage peak regulation surplus coefficient at each time

[0244]

[0245] Surplus coefficient of energy storage power station The amount of electricity that the energy storage power station reserves for the next moment of insufficient peak-shaving demand. The system has up-slope and up-peak-shaving demand scenarios obtained by historical data analysis, which are Q1, Q2, and Q6, and the probability of the two scenarios is r1, r2, and r6; similarly, the scenarios and probabilities of down-slope and down-peak-shaving demand are Q3, Q4, Q5 and r3, r4, and r5; Q7 and r7 are the scenarios and probabilities when the thermal power units meet the peak-shaving demand and system flexibility. The up / down / no-adjustment demand sets are Ru={r1,r2,r6}, Rd={r3,r4,r5}, and Rb={r7}, respectively, and the internal elements are represented by r u 、r d 、r b Indicates that the total number of elements is u n d n , b n .

[0246] Step 2: Analyze the double-layer optimization of peak load regulation of thermal power units assisted by energy storage clusters.

[0247] (1) Thermal-storage peak-shaving optimization control layer based on the peak-shaving surplus coefficient of the energy storage cluster: The system operating status is analyzed, and the system peak-shaving operation scenario is modeled based on historical data. The energy storage peak-shaving surplus coefficient of each time period is introduced to regulate the power of the energy storage cluster to assist the system peak-shaving and improve the system flexibility. A thermal-storage peak-shaving power economic allocation model considering the uncertainty of wind power is constructed, and the optimal allocation of peak-shaving resources between the energy storage cluster and the thermal power units is completed under the premise of meeting the economic objective function.

[0248] (2) Optimization control layer of each energy storage power station under multi-objective constraints: A multi-objective peak-shaving power optimization model between energy storage power stations within the energy storage cluster considering battery life is proposed, and the peak-shaving power demand of the energy storage cluster obtained by the upper layer is allocated among the multiple types of energy storage power stations in the lower layer. Under the consideration of network flow constraints and the balance of supply and demand of system peak-shaving resources, the energy storage cycle life coefficient, the balance of the load rate of the energy storage power station node and the operating cost of the energy storage power station are used as multi-objective functions to achieve the optimal peak-shaving output distribution among multiple types of energy storage power stations.

[0249] Step 3: Formulate upper-level strategies: Optimize the peak-shaving control of thermal storage based on the peak-shaving surplus coefficient of each energy storage.

[0250] The first step is to design the control strategy for combined thermal-storage peak load regulation.

[0251] When the peak load surplus coefficient of energy storage is higher than the upper limit When the thermal power unit adjusts its output, the SOC of the energy storage power station remains at the same level as the energy storage surplus coefficient SOC tj ;

[0252] When the peak-shaving surplus coefficient of energy storage is between the upper and lower bounds When the energy storage power station SOC is adjusted to It can ensure that the energy storage power station can generate maximum power;

[0253] When the energy storage peak regulation surplus coefficient is less than the lower limit When the thermal power unit tracks the equivalent load, the energy storage power station does not participate.

[0254] To simplify the formula, the two-way constraints such as upper and lower demands are expressed in only one formula, P xq (t) is the peak load demand at time t; ΔP xq (t) is the remaining peak load demand at time t; is the peak load power supplied by the energy storage power station at time t; is the climbing constraint of the thermal power unit; ΔP lhx (t) is the system flexibility requirement at time t; E S (t) is the power of the energy storage station at time t; SOC tj Prepare the required value for the energy storage SOC; The median value of the energy storage SOC preparation requirement; EN is the rated capacity of the energy storage power station; P S (t) is the output of the energy storage power station; Δt is the time interval between sampling points. The system operation status is shown in the figure below: Figure 6 As shown:

[0255] The detailed analysis of each operation is as follows:

[0256] (1)Q1: When the thermal power unit's ramp-up capability is met and the peak-shaving capacity is less than the equivalent load demand, the remaining peak-shaving demand is supplied by the energy storage system:

[0257]

[0258] (2)Q2: When the thermal power unit has insufficient ramp-up capability but sufficient peak-adjustment capacity, the energy storage power station can supplement the flexibility of the peak-adjustment at that moment:

[0259]

[0260] (3)Q3: When the thermal power unit has insufficient ramping capability and the peak load regulation capacity is less than the equivalent load demand, the remaining peak load regulation demand is supplied by the energy storage system:

[0261]

[0262] (4)Q4: When the thermal power unit's ramp-down capability is met and the peak-down capacity is less than the equivalent load demand, the remaining peak-down resources are replenished by the energy storage system.

[0263]

[0264] (5)Q5: When the thermal power unit has insufficient ramp-down capability and the peak capacity is adjusted down to meet the equivalent load demand, the energy storage power station can supplement the downward adjustment flexibility at this moment:

[0265]

[0266] (6)Q6: When the thermal power unit has insufficient ramp-up capability and the upward peak capacity is less than the equivalent load demand, the energy storage power station can supplement the downward flexibility at this moment:

[0267]

[0268] (7)Q7: When the thermal power unit's ramping capability is met and the peak load regulation capacity meets the equivalent load demand:

[0269] a) When the energy storage peak regulation surplus coefficient At this time, the thermal power unit is used to charge the energy storage system so that the energy storage system meets SOC tj , there is a certain flexibility in upward adjustment. If the energy storage power station SOC>SOC tj Then SOC(t+1)=SOC(t), otherwise:

[0270]

[0271] b) when When there is a certain probability of insufficient upward adjustment, in order to avoid wasting resources, the energy storage power station does not need to be equal to the upward adjustment coefficient. If the energy storage power station Then SOC(t+1)=SOC(t), otherwise:

[0272]

[0273] c) when At this time, the next moment, the downward adjustment capacity is insufficient, and the energy storage power station does not act. S (t) = 0, mainly based on the output of thermal power units:

[0274]

[0275] The second step is to construct the economic allocation model of thermal-storage peak load regulation power.

[0276] (1) Objective function

[0277] The objective function aims to minimize the overall operating cost of the system, which includes the operating cost of thermal power generators and the penalty cost caused by wind curtailment and load loss, as follows:

[0278] min(F 1 )=F1+F2 (21)

[0279] Where F1 is the operating cost of thermal power units; F2 is the penalty cost caused by wind abandonment and load loss.

[0280]

[0281] Among them, i is the startup cost of generator set i; c i is the fixed operating cost of generator set i; a i , b i are the coal consumption coefficients of thermal power unit i; P G,i (t) is the output of thermal power unit i at time t; z i,tis the decision variable for starting and stopping the generator set; its value is 1 for starting and 0 for not starting; μ i,t It is the decision variable of the generator set state. When the value is 1, it means running, and when it is 0, it means shutdown.

[0282]

[0283] In the formula, c wast is the wind abandonment penalty coefficient, is the wind abandonment volume at time t, c loss is the load loss penalty coefficient, is the load loss at time t, where:

[0284]

[0285] Where P W (t) is the wind power generated by the wind farm at time t; P is the wind power connected to the grid at time t; G,i (t) is the output of thermal power plant i at time t; P S,j (t) is the output of energy storage power station j at time t; the total number of thermal power units and total number of energy storage power stations in the region are I, J, P L (t) is the grid load that needs to be met at time t.

[0286] (2) Constraints

[0287] 1) Power balance constraints

[0288]

[0289] 2) Constraints on thermal power units

[0290] Output upper and lower limits of thermal power units:

[0291]

[0292] Thermal power unit climbing constraints:

[0293]

[0294] Thermal power unit start-up and shutdown constraints:

[0295]

[0296] In the formula, are the minimum and maximum output of thermal power unit i respectively; is the climbing limit of thermal power unit i; u i,t is the start / stop status of thermal power unit i, 1 and 0 are start and stop respectively; t on ,t off The minimum startup and shutdown time.

[0297] 3) Energy storage constraints

[0298] Energy storage power station SOC:

[0299]

[0300]

[0301] Energy storage power station charging and discharging upper and lower limit constraints:

[0302]

[0303] Constraints on the consistency of the power at the beginning and end of the energy storage power station:

[0304] SOC T =SOC0 (34)

[0305] In the formula, E j,t is the energy of energy storage station j at time t; E N is the rated capacity of the energy storage cluster; SOC0 is the SOC value of the energy storage cluster at the initial moment; SOC T It is the discharge terminal; it ensures the stable operation of the energy storage system.

[0306] 4) Wind curtailment constraints

[0307]

[0308] In the formula, To abandon the upper realm.

[0309] 5) Network operation constraints

[0310] Since the present invention studies the peak load dispatching strategy of the power system, the active power constraint of the transmission line needs to be considered.

[0311]

[0312] In the formula, is the upper limit of active power transmitted by line ab; P ab (t) is the active power flowing from node a to node b on line ab during period t, and is calculated as follows:

[0313]

[0314] Where, X an , X bn , X al , X bl is the node impedance matrix element of the network topology; x n 、x l is the reactance value of line ab at node n and node l; Pn (t) is the active power of power node n at time t; D l (t) Active power aggregate of load node l.

[0315] Step 4: Formulate lower-level strategies: Optimize control of each energy storage power station under multi-objective constraints.

[0316] The first step is to build a multi-objective peak load optimization model between energy storage power stations.

[0317] (1) Objective function

[0318] 1) Energy storage cycle life

[0319] It is complicated to analyze the battery life by considering all factors. This paper ignores the influence of temperature and charge and discharge rate on battery life and only considers the influence of charge and discharge depth and cycle number on battery life. [25-26] The relationship is fitted by a 5th-order function as follows: Figure 7 As shown. Using the rain flow counting method, the u-cycle DOD of the battery working DOD can be obtained according to the battery SOC curve. j (u), the battery life decay rate can be expressed as:

[0320]

[0321] Where U j,max (DOD j (u)) represents the battery cycle life (number of times) corresponding to the u-th discharge depth at power station j.

[0322] 2) Load rate

[0323] In order to maximize the effective utilization of energy storage power station resources, the output of the generator set at each time point and the output power of each energy storage power station are included in the multi-energy storage power station coordinated peak-shaving optimization operation model as decision variables. In order to accurately evaluate the output level of each energy storage power station, the "node load rate" is introduced. The load rate is an indicator that describes the ratio between the actual load carried by the transformer and its maximum capacity. The smaller the value of the load rate balance, the more balanced the load distribution. The calculation formula for the node load rate balance is as follows:

[0324]

[0325] Among them, μ j,t is the load rate of the j-th energy storage power station in period t; The load borne by this substation; μ ave,t is the average load rate of multiple energy storage power stations in period t.

[0326] 3) Energy storage operating costs

[0327] In order to significantly improve the economic operation efficiency of energy storage power stations, an energy storage power station operation cost model was constructed to help energy storage power stations maximize their economic benefits:

[0328]

[0329] In the formula, c price The operating price for energy storage.

[0330] (2) Constraints

[0331] 1) State of charge constraints:

[0332]

[0333] Where P S,j (t) is the output of energy storage power station j at time t; P S '(t) is the total power demand of the energy storage cluster.

[0334] 2) The energy storage operation constraints are shown in equations (32) and (33).

[0335] 3) Network operation constraints are shown in equations (36) and (37):

[0336] 4) Regulating consistency constraints

[0337] In order to ensure that the energy storage stations are charged and discharged at the same time, that is, the regulation consistency, the energy storage station power cannot have the reverse increase or decrease:

[0338]

[0339] Where ΔE S,j (t) is the capacity change of energy storage station j in period t; E S,j (t) is the capacity of energy storage station j at time t; E S,j (t-1) is the capacity of energy storage station j at t-1; ΔE S '(t) is the capacity change of the energy storage cluster in period t; E S '(t) is the capacity of the energy storage cluster at time t; E S '(t-1) is the capacity of the energy storage cluster at time t-1.

[0340] 5) Output fairness constraints

[0341] In order to avoid the situation where the more economical energy storage power stations are always in operation, while the less economical energy storage power stations are always underutilized, resulting in energy waste, the remaining capacity consistency constraint of the energy storage power stations is introduced. First, the real-time SOC of each energy storage power station is normalized to obtain the remaining capacity b of the energy storage power station at each moment. j(t), propose b bias,j (t), represents the remaining capacity of energy storage power station j and the average value b of each power station at time t-1 mean Deviation of (t-1):

[0342]

[0343] b bias,j (t)=(b j (t-1)+Δb j (t)-b mean (t-1)) 2 (44)

[0344] The calculation methods of each part are as follows:

[0345]

[0346] In the formula, Δt is the time interval between two moments; Δb j (t) is the change in the remaining capacity of the energy storage station j at time t; ΔSOC j (t) is the SOC change of energy storage station j at time t.

[0347] Therefore, the output fairness constraint is:

[0348]

[0349] In the formula, γ max It is the maximum value of total deviation allowed by the system.

[0350] The second step is to solve the problem based on the multi-objective genetic algorithm model.

[0351] (1) Pareto Optimality

[0352] In a multi-objective optimization problem, all possible Pareto optimal solutions form the so-called Pareto front. The concept of Pareto optimality aims to provide a set of acceptable compromises for the decision maker to choose from. This set contains all the optimal solutions that cannot be dominated by other solutions, and it is these that make up the Pareto front. Imagine facing a minimization problem involving two conflicting objectives; in this case, consider the following three-objective optimization problem:

[0353] 1) For every two solutions belonging to the same Pareto front, equation (49) holds:

[0354]

[0355] For every solution x that belongs to the Pareto frontier S i , there is at least one solution x j, which is better than x in at least one objective function (here called n) i In other words, there is no optimal solution on the Pareto front considering the set of all objective functions.

[0356]

[0357] 2) For each solution belonging to the upper Pareto frontier and the lower Pareto frontier, the following holds:

[0358]

[0359] For each solution belonging to the upper Pareto front, there exists at least one solution in the lower Pareto front that is not inferior in any objective function and is better in at least one objective function.

[0360] (2) Multi-objective optimization solution algorithm

[0361] The present invention adopts the ε-constraint method to solve the multi-objective optimization problem in 4.1. Compared with other multi-objective solution methods, it can simplify the problem, be more intuitive, reduce the computational burden, and meet the actual operation requirements of the research system. When solving the minimum optimization problem, it minimizes one of the original objective functions and converts the other objective functions into constraints. Only one of the three objective functions in the present invention is retained, and the other two objective functions are transformed under the constraint of the constraint threshold ε1, min(F (1) ,F (2) ) is converted into a constraint problem:

[0362]

[0363] The third step is to establish the control effect evaluation index.

[0364] (1) SOC balance between energy storage power stations

[0365]

[0366] In the formula, SOC mean (t) is the average SOC value among energy storage clusters, SOC ph To balance the SOC among the power stations in the energy storage cluster, use SOC ph The SOC balance degree, operation fairness, and SOC ph The smaller it is, the better the fairness.

[0367] (2) Peak load tracking

[0368]

[0369] Where P(t) is the total power output, Hgz H is the degree of tracking of the system power supply to the load, gz The smaller the value, the better the strategy tracking.

[0370] (3) Unit energy consumption cost

[0371]

[0372] In the formula, C B (t) is the total energy consumption cost; P B (t) is the total energy consumption, including thermal power, energy storage and wind power, U cb It is the unit energy consumption cost during the total scheduling period. The smaller it is, the higher the energy utilization rate is.

[0373] Step 5: Use the analysis method of the present invention to perform corresponding simulation analysis on a specific example.

[0374] The present invention uses an improved IEEE-39 node network as the experimental basis to verify the actual effect of the proposed method. A 24-hour day is selected as the control cycle, and the optimization step is set to 15 minutes to ensure the accuracy and real-time performance of the control strategy. The topology of the system is as follows: Fig. 9 As shown in the figure, it covers 6 thermal power units, 5 wind power units and 5 energy storage power stations. The installed capacity of thermal power is 495MW, the installed capacity of wind power is 600MW, and the installed capacity of energy storage is 50MW / 100MWh according to the principle of wind power storage. It is worth mentioning that each wind farm is equipped with a corresponding energy storage power station to improve energy utilization efficiency.

[0375] Firstly, the historical data are analyzed to obtain the probability of thermal power units in various operating states in different time periods. The energy storage peak regulation surplus coefficient is obtained through formula (11). Then, the effectiveness of the strategy proposed in this paper is proved by comparing the traditional strategy with this strategy in terms of peak regulation effect and operating economy. Fig.10 is the day-ahead power generation plan of the thermal power unit. It can be seen that there are regulation problems in some periods. Fig.11 It is concluded that when a single thermal power plant participates in peak load regulation, there are periods when the peak load demand is not met due to the start and stop time restrictions of the units and insufficient downward wind regulation of the thermal power units. In order to strengthen the role of energy storage resources in system regulation, the present invention analyzes historical data. Fig.12 The yellow blocks in the figure indicate the time intervals when there is regulation demand, and the green blocks indicate that the thermal power units meet the regulation demand of the system. According to the proposed research strategy, the surplus demand coefficient of the energy storage power station in each period is obtained by analyzing the above historical data. Fig.13It can be seen that the coefficients with large values ​​are mainly concentrated in 8:00-21:00, which is consistent with the analysis in the previous article, that is, during this period, the unit output must be adjusted for the previous normal operation period with demand. Under Strategy 1, during the continuous high equivalent load stage, due to insufficient energy storage, it is impossible to output all the time, resulting in load loss, while for Strategy 2, there is no load loss during this period. Strategy 2 extends the time for energy storage to participate in peak regulation by adjusting the energy storage peak regulation surplus coefficient. Fig.14 The remaining peak-shaving demand under different strategies is shown in the figure. As can be seen from the figure, there is a large amount of insufficient upward wind during 18:00-22:00 under strategy 1, that is, load loss, while there is no upward peak-shaving demand during the entire peak-shaving period under strategy 2, which effectively ensures the safe operation of the power grid and achieves the goal of peak shaving and valley filling. As can be seen from the figure, under strategy 1, it is basically fully charged and fully discharged, while for strategy 2, the energy storage power station is not always in a fully charged and fully discharged state.

[0376] Secondly, in order to verify the effectiveness of this strategy in reducing the life of energy storage power stations, the following example is given: Fig.16 The output of each energy storage power station under different strategies. Energy storage 1-4 is electrochemical energy storage, and energy storage unit 5 is a pumped storage power station. The pumped storage power station has a large capacity and undertakes the main peak-shaving task most of the time. Fig.17 Under the goal of balancing the node load rate, the consistency index between energy storage power stations in Strategy 2 is 0.0082, which is 26.79% higher than that of Strategy 1, reflecting the fairness of the strategy to each power station. Fig.18 It is observed that various types of electrochemical energy storage power stations have a certain reduction in the number of cycle life by reducing the discharge depth.

[0377] Finally, the rigorous experimental results verify the SOC tj and The impact on the strategy, the energy storage SOC reserve requirement value (SOC tj ) and energy storage peak regulation surplus coefficient The peak regulation effect and energy storage cycle life of the proposed strategy have a certain impact. Fig.19 The peak load demand decreases with the increase of the two types of coefficients. For the energy storage SOC reserve demand value, SOC tj The increase in will increase the capacity of energy storage power stations, improve the ability to cope with peak loads, reduce peak load demand, and the peak load surplus coefficient of energy storage The increase will reduce the charging opportunities of energy storage power stations, thereby increasing the peak demand. Fig. 20 The system operating cost varies with SOC tj and The relationship between the upper and lower limits is that the basic trend is to decrease first and then increase. tjFor example, in the higher stage, the energy storage may be charged but not fully utilized, which will increase the operating cost. In the later stage, the energy storage peak-shaving resources are insufficient, which will cause penalty costs such as load loss and wind abandonment. tj The larger the value, the more beneficial it is to the system peak load regulation operation. However, based on the constraints of system operation cost and energy storage power station cycle life attenuation, the value of this invention is 0.88; for the peak load regulation of energy storage peak load surplus coefficient In terms of, the smaller the value, the better it is for system peak regulation and energy storage cycle life attenuation. However, based on the regional operating cost constraints, the present invention takes a value of 0.3, that is, the rationality of the values ​​of the two coefficients in this paper is obtained through the above analysis.

[0378] The embodiments of the present invention are not exhaustive and do not constitute a limitation on the protection scope of the claims. Based on the inspiration gained from the embodiments of the present invention, those skilled in the art can think of other substantially equivalent alternatives without creative work, all of which are within the protection scope of the present invention.

Claims

1. A hierarchical optimization control method for peak load regulation of thermal power units assisted by energy storage clusters, characterized in that: It includes the following steps: Step 1: Analyze and model the peak load status of the system including the energy storage cluster: 1) Analyze the peak load regulation of a single thermal power plant, and obtain the peak load regulation demand of each peak load regulation time period within a planned period when a single thermal power plant participates in peak load regulation according to the corresponding relationship between the thermal power unit and the equivalent load; 2) Analyze the joint participation of thermal energy and energy storage in peak load regulation to obtain the corresponding actions of energy storage in each peak load regulation time period; 3) Analyzing the operation mode of the energy storage cluster to obtain multiple operation modes of the energy storage cluster; 4) Modeling the operation status of the combined thermal-storage peak load regulation, and obtaining the peak load surplus coefficient of energy storage at each moment based on the established scenario probability model Step 2: Energy storage cluster assists thermal power units in peak load regulation with double-layer optimization; Step 3: Formulate the upper-level method: Optimal control of thermal storage peak load regulation based on the peak load surplus coefficient of each energy storage. 1) Formulate a combined thermal-storage peak-shaving control method based on the energy storage peak-shaving surplus coefficient, 2) Constructing a power economic allocation model for thermal-storage peak load regulation; Step 4: Formulate the lower-level method: Optimal control of each energy storage power station under multi-objective constraints 1) Construct a multi-objective peak load optimization model between energy storage power stations 2) Solving based on multi-objective genetic algorithm model; The step 1) of step 1 includes: According to the corresponding relationship between thermal power units and equivalent loads, the peak load demand of each peak load time period within a planned period is obtained when a single thermal power plant participates in peak load regulation; When a single thermal power plant participates in peak regulation, the peak regulation demands of each peak regulation time period in a planned period are as follows: in [t1-t2], the thermal power unit meets the peak regulation demand; in [t2-t3], the maximum ramp rate of the thermal power unit cannot track the load, resulting in an upward flexibility demand; in [t3-t4], the capacity of the thermal power unit is insufficient, resulting in an upward peak regulation demand; in the [t4-t5] interval, the thermal power unit can track the equivalent load; in [t5-t6], the thermal power unit has a slow adjustment speed, resulting in a downward flexibility demand; in [t6-t7], the equivalent load curve is lower than the minimum output of the thermal power unit, resulting in a downward peak regulation demand; in the [t7-t8] interval, the thermal power unit's downward ramp rate does not meet the demand; in the [t8-t9] interval, the thermal power unit's downward peak regulation capacity is insufficient and cannot track the equivalent load; The step 1) of step 2) comprises: Based on the power constraints and capacity constraints of energy storage power stations in actual operation, the energy storage cluster regulation capability model is constructed as shown in the following formula: In the formula, are the discharge power and charging power of energy storage station j at time t; SOC j (t) are the actual value, lower limit and upper limit of SOC of energy storage station j at time t; E N,j is the rated capacity of energy storage power station j; P N,j is the rated charging and discharging power of energy storage station j; is the discharge efficiency and charging efficiency of the energy storage station; The regulation capability model of extending the output of the above single energy storage power station to a multi-type energy storage cluster is as follows: In the formula, is the maximum charging and discharging power of the energy storage cluster in the region at time t The corresponding actions of energy storage in each peak load period and other time periods correspond to the following six situations: ①[t1-t2] Energy storage does not work in, is the charging and discharging power of the energy storage system; in the interval [t1-t2], the thermal power unit can meet the equivalent load demand, and the energy storage system does not operate; ②[t2-t3] Energy storage provides upward flexibility in, The demand for upward ramping is met. At time t2, the maximum ramping of thermal power is insufficient due to the start and stop of the thermal power units, resulting in insufficient flexibility for upward adjustment of the system in the interval [t2-t3]. At this time, the discharge of the energy storage cluster meets the flexibility demand for upward adjustment. ③[t3-t4] Energy storage provides peak load regulation resources In the interval [t3-t4], thermal power units are unable to meet the peak load demand, and energy storage provides peak load resources to meet equivalent load operation. ④[t4-t5] Thermal power charges the energy storage system In the interval [t4-t5], the thermal power units can track the equivalent load. Based on the prediction that there will be a large demand for peak-shaving resources at the next moment, the excess thermal power resources are used to charge the energy storage cluster to meet the power demand in the next period [t5-t7]. ⑤[t7-t8] Energy storage provides downward adjustment flexibility in, The climbing demand; in the interval [t7-t8], the climbing capacity of the thermal power unit does not meet the demand. At this time, the energy storage improves the flexibility of the system by absorbing excess wind power; ⑥[t8-t9] Energy storage provides peak load reduction resources In the interval [t8-t9], the thermal power units have insufficient peak-shaving capacity and cannot track the equivalent load, so the energy storage power station supplements the peak-shaving resources. in, The up / down flexibility that energy storage plants should supplement; It is the up / down climbing ability of the thermal power unit; It is the maximum value of up / down ramp of thermal power unit; It is the upper / lower limit of thermal power unit output; It is the up / down peak-shaving resource that should be supplemented by the energy storage power station; is the total up / down peak regulation demand; P G (t) is the peak output of thermal power units; The charging power of the system thermal power unit for the energy storage power station; The energy storage cluster operation mode analysis of step 1 (3) includes: the energy storage cluster has a first operation mode and a second operation mode. When the energy storage cluster is in the first operation mode, the thermal power units operate at minimum output, and the energy storage makes up for the remaining peak-shaving demand; when the energy storage cluster is in the second operation mode, the output of the thermal power units is increased, and the output duration of the energy storage power station is extended from the original [t1-t2] to [t1-t3], and the remaining peak-shaving capacity of the thermal power units is used to increase the response capacity of the energy storage cluster; The modeling of the fire-storage combined peak-shaving operation state of step 1 (4) includes: For the system, there are m, m=7 possible operating states that will appear at the next moment. The m possibilities are converted into a matrix as shown in formula (9); The scenario set is Q, 0 means the requirement is met, 1 means the requirement is not met; from left to right, they are upward peak demand, downward peak demand, insufficient upward ramp, and insufficient downward ramp; Construct the fuzzy set χ of the probability distribution of the scene: In the formula, Q i is the i-th uncertain set; E R is the mathematical expectation; r i is the probability of being in the i-th uncertain scenario at the next moment; R is the probability distribution of the scenario variable ξ; χ 0 (Q m ) is ξ based on Q m The probability distribution set of all scenarios; According to the above scenario probability model, the probability of each operating state can be obtained, and the obtained results can be used to calculate the energy storage peak regulation surplus coefficient at each time : Surplus coefficient of energy storage power station is the amount of electricity that the energy storage power station reserves for the next moment of insufficient peak-shaving demand. The system has up-slope and up-peak-shaving demand scenarios obtained by historical data analysis, which are Q1, Q2, and Q6, and the probability of the two scenarios is r1, r2, and r6; similarly, the scenarios and probabilities of down-slope and down-peak-shaving demand are Q3, Q4, Q5 and r3, r4, and r5, respectively; Q7 and r7 are the scenarios and probabilities when the thermal power units meet the peak-shaving demand and system flexibility; the up / down / no-regulation demand sets are Ru={r1, r2, r6}, Rd={r3, r4, r5}, and Rb={r7}, respectively, and the internal elements are represented by r u 、r d 、r b Indicates that the total number of elements is u n d n , b n .

2. A hierarchical optimization control method for peak load regulation of thermal power units assisted by energy storage clusters as claimed in claim 1, characterized in that: The energy storage cluster assisting thermal power generation unit peak load regulation double-layer optimization in step 2 includes: (1) Thermal-storage peak-shaving optimization control layer based on the peak-shaving surplus coefficient of the energy storage cluster: Analyze the system operation status, model the system peak-shaving operation scenario based on historical data, and adjust the power of the energy storage cluster to assist the system peak-shaving and improve the system flexibility by introducing the peak-shaving surplus coefficient of the energy storage cluster in each period. A thermal-storage peak-shaving power economic allocation model considering the uncertainty of wind power is constructed, and the optimal allocation of peak-shaving resources between the energy storage cluster and the thermal power units is completed under the premise of meeting the economic objective function; (2) Optimization control layer of each energy storage power station under multi-objective constraints: A multi-objective peak-shaving power optimization model between energy storage power stations within the energy storage cluster considering battery life is proposed, and the peak-shaving demand power of the energy storage cluster obtained by the upper layer is allocated among the multiple types of energy storage power stations in the lower layer; the energy storage cycle life coefficient, the balance of the load rate of the energy storage power station node and the operating cost of the energy storage power station are used as multi-objective functions to achieve the optimal peak-shaving output distribution among multiple types of energy storage power stations under the consideration of network flow constraints and the balance of supply and demand of system peak-shaving resources.

3. A hierarchical optimization control method for peak load regulation of thermal power units assisted by energy storage clusters as claimed in claim 2, characterized in that: The step 3 of 1) formulating a combined fire-storage peak-shaving control method based on the energy storage peak-shaving surplus coefficient comprises: is the peak load demand at time t; ΔP xq (t) is the remaining peak load demand at time t; is the peak load power supplied by the energy storage power station at time t; is the climbing constraint of the thermal power unit; ΔP lhx (t) is the system flexibility requirement at time t; E S (t) is the power of the energy storage power station at time t; SOC tj Prepare the required value for the energy storage SOC; The median value of the energy storage SOC preparation requirement; E N is the rated capacity of the energy storage power station; P S (t) is the output of the energy storage power station; Δt is the time interval between sampling points; the operating conditions of the system are: (1)Q1： When the thermal power unit's ramp-up capability is met and the peak-shaving capacity is less than the equivalent load demand, the remaining peak-shaving demand is supplied by the energy storage system: (2)Q2： When the thermal power unit has insufficient ramp-up capability but sufficient peak-adjustment capacity, the energy storage power station can supplement the flexibility of the peak-adjustment at that moment: (3)Q3： When the thermal power unit has insufficient ramp-down capability and the peak-shaving capacity is less than the equivalent load demand, the remaining peak-shaving demand is supplemented by the energy storage system: (4)Q4： When the ramp-down capability of the thermal power unit is met and the peak-down capacity is less than the equivalent load demand, the remaining peak-down resources are replenished by the energy storage system. (5)Q5： When the thermal power unit has insufficient ramp-down capability and the peak capacity is adjusted down to meet the equivalent load demand, the energy storage power station can supplement the downward adjustment flexibility at this moment: (6)Q6： When the thermal power unit has insufficient ramp-up capability and the upward peak capacity is less than the equivalent load demand, the energy storage power station can supplement the downward flexibility at that moment: (7)Q7： When the thermal power unit has sufficient ramp capability and the peak load regulation capacity meets the equivalent load demand: a) :When the energy storage peak regulation surplus coefficient At this time, the thermal power unit is used to charge the energy storage system so that the energy storage system meets SOC tj , there is a certain flexibility in upward adjustment. If the energy storage power station SOC>SOC tj Then SOC(t+1)=SOC(t), otherwise: b) :when When there is a certain probability of insufficient upward adjustment, in order to avoid wasting resources, the energy storage power station does not need to be equal to the upward adjustment coefficient. If the energy storage power station Then SOC(t+1)=SOC(t), otherwise: c) when At this time, the next moment, the downward adjustment capacity is insufficient, and the energy storage power station does not act. S (t) = 0, mainly based on the output of thermal power units:

4. A hierarchical optimization control method for peak load regulation of thermal power units assisted by energy storage clusters as claimed in claim 3, characterized in that: The step 3 (2) of constructing a thermal-storage peak load power economic allocation model includes: (1) Determine the objective function: The objective function aims to minimize the overall operating cost of the system, which includes the operating cost of thermal power generators and the penalty cost caused by wind curtailment and load loss, as follows: <h2 style=";text-align:left;direction:ltr">min(F<h2 style=";text-align:left;direction:ltr"> 1 <h2 style=";text-align:left;direction:ltr"> )=F1+F2 (21) Where, F1 is the operating cost of thermal power units; F2 is the penalty cost caused by wind abandonment and load loss; Among them, i is the startup cost of generator set i; c i is the fixed operating cost of generator set i; a i , b i are the coal consumption coefficients of thermal power unit i; P G,i (t) is the output of thermal power unit i at time t; zi,t is the decision variable for starting and stopping the generator set; its value is 1 for starting and 0 for not starting; μ i,t is the decision variable of the generator set state, when the value is 1, it means running, and when it is 0, it means shutdown; In the formula, c wast is the wind abandonment penalty coefficient, is the wind abandonment volume at time t, c loss is the load loss penalty coefficient, is the load loss at time t, where: Where P W (t) is the wind power generated by the wind farm at time t; P is the wind power connected to the grid at time t; G,i (t) is the output of thermal power plant i at time t; P S,j (t) is the output of energy storage power station j at time t; the total number of thermal power units and total number of energy storage power stations in the region are I, J, P L (t) is the grid load that needs to be met at time t; (2) Determine the constraints: 1) Power balance constraints 2) Constraints on thermal power units Output upper and lower limits of thermal power units: Thermal power unit climbing constraints: Thermal power unit start and stop constraints: In the formula, are the minimum and maximum output of thermal power unit i respectively; is the climbing limit of thermal power unit i; ui,t is the start / stop status of thermal power unit i, 1 and 0 are start and stop respectively; t on ,t off The minimum startup and shutdown time; 3) Energy storage constraints Energy storage power station SOC: Energy storage power station charging and discharging upper and lower limit constraints: Constraints on the consistency of the first and last power of energy storage power stations: SOC T =SOC0 (34) In the formula, E j,t is the energy of energy storage station j at time t; E N is the rated capacity of the energy storage cluster; SOC0 is the SOC value of the energy storage cluster at the initial moment; SOC T It is the discharge terminal; it ensures the stable operation of the energy storage system; 4) Wind curtailment constraints In the formula, For abandoning the wind to the upper realm; 5) Network operation constraints Considering the active power constraint of the transmission line: In the formula, is the upper limit of active power transmitted by line ab; P ab (t) is the active power flowing from node a to node b on line ab during period t, and is calculated as follows: In the formula, X an , X bn , X al , X bl is the node impedance matrix element of the network topology; x n 、x l For line ab at node n and the reactance value at node l; P n (t) is the active power of power node n at time t; D l (t) Active power aggregate of load node l.

5. A hierarchical optimization control method for peak load regulation of thermal power units assisted by energy storage clusters as claimed in claim 4, characterized in that: The step 4 of 1) constructing a multi-objective energy storage power station peak load optimization model includes: (1) Determine the objective function: 1) Energy storage cycle life The battery life decay rate can be expressed as: Where U j,max (DOD j (u)) represents the battery cycle life (number of times) corresponding to the u-th discharge depth at power station j; 2) Load rate The calculation formula for node load balance is as follows: Among them, μ j,t is the load rate of the j-th energy storage power station in period t; The load borne by this substation; μ ave,t is the average load rate of multiple energy storage power stations in period t; 3) Energy storage operating costs In order to significantly improve the economic operation efficiency of energy storage power stations, an energy storage power station operation cost model was constructed to help energy storage power stations maximize their economic benefits: In the formula, c price The price for energy storage operation; (2) Determine the constraints 1) State of charge constraints: Where P S,j (t) is the output of energy storage power station j at time t; P S '(t) is the total power demand of the energy storage cluster; 2) The energy storage operation constraints are shown in equations (32) and (33); 3) Network operation constraints are shown in equations (36) and (37): 4) Regulating consistency constraints In order to ensure that the energy storage stations are charged and discharged at the same time, that is, the regulation consistency, the energy storage station power cannot have the reverse increase or decrease: In the formula, ΔE S,j (t) is the capacity change of energy storage station j in period t; E S,j (t) is the capacity of energy storage station j at time t; E S,j (t-1) is the capacity of energy storage station j at t-1; ΔE S '(t) is the capacity change of the energy storage cluster in period t; E S '(t) is the capacity of the energy storage cluster at time t; E S '(t-1) is the capacity of the energy storage cluster at time t-1; 5) Output fairness constraints The real-time SOC of each energy storage power station is normalized to obtain the remaining capacity b of the energy storage power station at each moment. j (t), propose b bias,j (t), represents the remaining capacity of energy storage power station j and the average value b of each power station at time t-1 mean Deviation of (t-1): b bias,j (t)=(b j (t-1)+Δb j (t)-b mean (t-1)) 2 (44) The calculation methods of each part are as follows: In the formula, Δt is the time interval between two moments; Δb j (t) is the change in the remaining capacity of the energy storage station j at time t; ΔSOC j (t) is the SOC change of energy storage station j at time t; Therefore, the output fairness constraint is: In the formula, γ max It is the maximum value of total deviation allowed by the system.

6. A hierarchical optimization control method for peak load regulation of thermal power units assisted by energy storage clusters as claimed in claim 5, characterized in that: The solution process of step 4 2) based on the multi-objective genetic algorithm model is: 1) For every two solutions belonging to the same Pareto front, equation (49) holds: For every solution x that belongs to the Pareto frontier S i , there is at least one solution x j , which is better than x in at least one objective function (here called n) i ; 2) For each solution belonging to the upper Pareto frontier and the lower Pareto frontier, the following holds: Only one of the three objective functions is retained, and the other two objective functions are transformed under the constraint of the constraint threshold ε1, min(F (1) ,F (2) ) is converted into a constraint problem:

Images