Energy storage capacity optimization model considering transferable load characteristics of rural power distribution network area

Abstract

The invention relates to the technical field of energy storage capacity optimization, in particular to an energy storage capacity optimization model considering the transferable load characteristic of a rural power distribution network district, and the model comprises the following steps: 1, obtaining the most representative solar photovoltaic output curve in the whole year based on a k-means clustering algorithm; step 2, demand response collaborative optimization; specifically, demand side characteristic analysis and rural area transferable load characteristics are included; 3, a light storage capacity configuration model considering transferable load characteristics; the parameters specifically comprise an objective function and constraint conditions; and step 4, solving by applying a particle swarm algorithm model. According to the energy storage capacity optimization method considering the transferable load characteristic of the rural power distribution network area, the transferable load characteristic of the rural power distribution network area is considered, a novel photovoltaic energy storage capacity configuration model is established, and photovoltaic absorption and reliable operation of a rural area distribution network are achieved.

Description

Technical Field

[0001] The present invention relates to the technical field of energy storage capacity optimization, and in particular to an energy storage capacity optimization model that takes into account the transferable load characteristics of a rural distribution network station area. Background Art

[0002] As the scale of power grid expands year by year, users have higher and higher requirements for power supply quality. On the one hand, high-proportion distributed generation (DG) represented by photovoltaic power generation is greatly affected by external factors and has a high degree of randomness; on the other hand, with the continuous development of rural areas, rural load categories and load demand are increasing year by year, and the electricity consumption situation is becoming more complicated and diversified. At this time, the access of distributed generation will bring about voltage over-limit and serious abandonment of light. In order to deal with the contradiction between the reliability of distribution network and the demand for energy conservation and emission reduction in the development of new energy power generation, in the past mode, it is usually started from the balance of supply and demand, and a battery energy storage system (BESS) is established. By using its characteristics that can break the real-time balance limit of power supply and demand to a certain extent, it can achieve peak shaving and valley filling, smooth intermittent power fluctuations, improve voltage quality, and promote large-scale consumption of new energy and achieve energy conservation and emission reduction. However, most rural distribution networks do not consider the resource scheduling on the demand side. Therefore, in the new rural distribution network, the load-storage is flexibly adjusted through demand-side response to reduce the peak-to-valley difference of load, so as to meet the needs of high-proportion new energy generation access and efficient consumption to the greatest extent.

[0003] As an important means of regulating active distribution networks, energy storage systems have been extensively studied by many scholars at home and abroad. The energy storage configuration method has changed from "storage follows load" to "source, load, storage" multi-party coordinated dynamic power balance. Starting from the energy storage mode, the literature studies multiple energy storage modes such as wind / solar / diesel / storage, and analyzes the impact of energy storage types on photovoltaic consumption. Starting from the BESS operation strategy, the literature considers factors such as the charge and discharge balance constraints and the full life cycle of the energy storage system, divides the peak and valley of the daily load into charging and discharging periods, and adjusts the charging and discharging power as the decision variable. On the basis of considering the source-load balance, the literature establishes a value evaluation model, solves the model with the goal of maximizing the value, and proves the economic feasibility of the energy storage system. By adjusting the BESS capacity and charge and discharge power, the literature improves voltage stability and reduces grid losses, thereby improving the economic benefits of the distribution network while keeping the distributed photovoltaic (PV) capacity and access points unchanged. The literature takes network losses into consideration and adopts a two-layer planning charging and discharging strategy. The upper layer optimizes the charging and discharging period, and the lower layer optimizes the discharge power to achieve accurate division of the charging and discharging period. The above literature shows that the reasonable configuration and regulation of the energy storage system in the distribution network can effectively improve voltage fluctuations, reduce the amount of abandoned light in the distribution network, and have a significant impact on the penetration rate of DG in the future power grid. However, most of them only consider the conservation of BESS cycle charging and discharging, but less consideration is given to the balance of continuous discharge. Since the charging and discharging amount of BESS remains unchanged with its adjacent electricity each time, they are relatively independent. Some energy storage designs do not consider the globality of long-term charging and discharging, and their strategies have not fully achieved the expected effect in maintaining power balance and peak-shaving and valley-filling. In addition, in terms of source-load balance, only the participation of industrial and commercial users with large electricity consumption in the traditional distribution network architecture is considered, and the distribution network loads in some rural areas and remote mountainous areas are not taken into consideration. Nowadays, with the rapid development and popularization of smart power consumption equipment, the dispatching of transferable loads in rural areas can achieve peak-shaving and valley-filling for rural load characteristics and optimize the source-load balance. Summary of the invention

[0004] In view of the defects and shortcomings of the prior art, the present invention provides an energy storage capacity optimization model that takes into account the transferable load characteristics of rural distribution network stations.

[0005] In order to achieve the above purpose, the technical solutions provided are as follows:

[0006] The energy storage capacity optimization model considering the transferable load characteristics of the rural distribution network area is characterized by comprising the following steps:

[0007] Step 1: Obtain the most representative daily photovoltaic output curve throughout the year based on the k-means clustering algorithm; specifically, the following steps are included:

[0008] Step (1), based on the historical data of the substation, pre-process the typical operation data of the substation's photovoltaic load, divide the data of one year into four data clusters as the photovoltaic output data of the four seasons, and perform maximum value normalization processing on them;

[0009] Step (2), calculate the sum of the squared Euclidean distances J(C) of all data to the cluster center of the corresponding category, and classify the category with the closest distance into one category;

[0010]

[0011] Where: u mn is a binary variable, u mn =1 means the nth sample belongs to the mth class, and 0 means the opposite; d(c m ,x n ) is the distance from the sample to the center of the cluster to which it belongs; c m is the sample cluster center; x n For other sample data in the same category;

[0012] Step (3): According to the division result of step (2), k cluster centers c m To update the generation, Lagrange's basic principle is used to introduce J(C) as a constraint, and the objective function is selected as d(c m ,x n ), optimize it by the least square method, and then iteratively update the 4 cluster centers according to the following formula;

[0013]

[0014] The judgment is made according to the basic convergence conditions. If it does not meet the convergence conditions, the iteration conditions are continued in equation (2). If it meets the convergence conditions, the photovoltaic output data of four typical days are output.

[0015] Step 2: Demand response collaborative optimization; specifically including demand side feature analysis and rural area transferable load characteristics;

[0016] In the demand side characteristic analysis, flexible loads can be divided into three categories from the perspective of load regulation characteristics, specifically: ① Transferable loads: The total power consumption of the periodic type remains unchanged, but the power consumption time is flexible, such as electric vehicle charging piles, ice storage appliances, and some residential loads; ② Shiftable loads: Due to the constraints of the production process, the power load can only be shifted in a continuous period, such as large industrial users; ③ Reducible loads: The power consumption can be reduced at a specific time, such as air conditioners and other household electrical equipment;

[0017] In the characteristics of the transferable load in the rural area, the transferable load can be more flexibly scheduled in terms of energy consumption time and load size changes compared with other types of flexible loads, achieving better coordinated optimization of the source and load sides, improving the comprehensive benefits of the new energy system, and increasing the absorption of renewable energy such as photovoltaics; based on the energy consumption characteristics of the transferable load equipment, the transfer range of the transferable load and the constraints on the increase in electricity prices are determined, and the load affiliation index is established. The transferable load is scheduled with the minimum load peak-to-valley difference as the goal, reducing the source-load imbalance and improving the system safety;

[0018] Constrain the electricity price and realize the optimal dispatch of transferable load, denoted by P t tran is the load that can be transferred at a certain moment. The transfer period of the transferable load is from 6 pm to 6 am. 0-1 is used to indicate whether the transferable load is transferred during a certain period, and β is used to mark it;

[0019] (1) Transferable load power range

[0020]

[0021] In the formula, is the minimum value of the transferable load equipment power, The maximum value of the transferable load equipment power;

[0022] (2) Calculate the membership degree for 24 hours a day

[0023] The difference between the load power of each hour and the average power in a day is selected as the membership degree μ i ;

[0024]

[0025] Where P L (t) is the load power of the substation;

[0026] (3) Transferable load dispatch constraints

[0027] The operation and scheduling of transferable loads will have an impact on the user's life. Therefore, a penalty function based on transferable load scheduling is proposed to constrain it.

[0028]

[0029] Where f is the coefficient of the penalty function;

[0030] (4) Constraints on electricity price fluctuations

[0031] Through electricity price regulation, transferable load scheduling is achieved, with electricity prices increased during peak load periods and reduced during valley load periods;

[0032]

[0033] In the formula, p f is the average electricity price, i.e. the electricity price during normal periods, Δp i The adjustment range of peak and valley electricity prices;

[0034] (5) Objective function

[0035] The product of the membership degree of the corresponding time period and the power of the corresponding time period.

[0036]

[0037] By establishing a membership index that correlates load changes and electricity price fluctuations, the goal of optimizing the load curve with the objective function is achieved.

[0038] Step 3: A photovoltaic storage capacity configuration model considering transferable load characteristics, including an objective function and constraints;

[0039] Preferably, in the objective function, the charging and discharging period and the charging and discharging power are used as decision variables, and the constraints such as the area voltage and the charging and discharging power are considered to solve the maximum photovoltaic absorption rate as the goal, so as to realize the reasonable division of the charging and discharging strategies within 24 hours and maximize the photovoltaic absorption rate; the objective function is the photovoltaic absorption rate within the area within 24 hours:

[0040]

[0041] P SC (t) = min{P PV (t),P out (t)} (9)

[0042] Where: R sel is the local consumption rate of new energy; P PV (t) is the photovoltaic power output of the new energy in the area, P L-a (t) is the new energy output consumed by the load in real time in the distribution network of the substation; P bat-charge (t) is the charging power of the energy storage during time period t, and P bat-charge (t)>0.

[0043] Preferably, the constraints include:

[0044] (1) Substation voltage constraints

[0045] The substation must maintain a single-phase power supply voltage deviation of +7% and -10% for 0.22kv, and a power supply voltage deviation of ±7% of the rated voltage for 10kv and below, to ensure the power quality of BESS operation.

[0046] U i,min ≤U i ≤U i,maxi=1,2,...,n (10)

[0047] U i,min , U i,max are the upper and lower limits of the voltage at node i;

[0048] (2) Energy storage system operation constraints

[0049] In order to ensure the safe operation of energy storage, the energy storage charging and discharging power is guaranteed not to exceed the charging and discharging power of the battery;

[0050] -P bess ≤P bat (t)≤P bess (11)

[0051] Among them, P bat (t) is the charging and discharging power of the energy storage system at each time period, P bess The maximum allowable charging and discharging power of energy storage. Positive values ​​represent discharging, and negative values ​​represent charging.

[0052] Charging process:

[0053] Discharge process:

[0054] Among them, SOC(t+Δt) is the remaining power level at time t+Δt and t, that is, the state of charge; ε is the loss rate of the remaining power of BESS per hour, referred to as the self-discharge rate, in % / h; α and β are the charging and discharging efficiencies of BESS, respectively; E e is the capacity of BESS; t is the simulation time; Δt is the sampling interval, i.e. the time period during simulation, which is 1h here;

[0055] (3) Energy storage system charge state constraints

[0056] To avoid the impact of excessive charging and discharging of the battery on its service life, the charge state is limited to a certain range to ensure that the battery is neither fully charged nor fully discharged.

[0057]

[0058] Where S min , S max are the minimum and maximum states of charge of the energy storage, respectively; S(0) and S(24) are the states of charge of the energy storage system at time periods 0 and 24, respectively;

[0059] (4) Dischargeable Capacity Constraints

[0060] In order to make the solution applicable in actual situations and avoid the situation where the continuous discharge period is greater than the remaining power in the energy storage, or the continuous charge amount is greater than the energy storage content, certain constraints are imposed on the charge and discharge amount.

[0061] E dis,i ≤E soc,i (15)

[0062] Among them, E dis,i is the dischargeable capacity in the i-th continuous discharge time period; E soc,i is the remaining power of BESS before discharge in the i-th continuous discharge time period;

[0063] (5) Energy conservation constraints for charging and discharging within a daily cycle

[0064] The substation must maintain a balance between the real-time power it interacts with users and the real-time power it interacts with the distribution network and new energy sources.

[0065] P PV (t)+P grid (t)+P bat (t) = P L (t) (16)

[0066] P PV (t) is the photovoltaic power output of the new energy in the area, P grid (t) is the power flowing from one node to the next node in the distribution network, P bat (t) is the charging and discharging power of the energy storage during period t, P L (t) is the load power of the substation.

[0067] Step 4: Apply the particle swarm algorithm model to solve; the transferable load scheduling is a typical integer quadratic programming problem. The transferable load scheduling is solved by the CPLEX solver in the Matlab toolbox to obtain the load curve after the transferable load scheduling, so as to reduce the negative impact of the access to the distribution network system in the substation area;

[0068] The source-load-storage collaborative planning model of BESS is solved by PSO intelligent optimization algorithm. PSO algorithm has good global search performance and fast convergence speed. It can quickly solve the energy storage capacity and power configuration when the photovoltaic absorption rate in the platform area is the largest. PV (t) and load forecast power demand P0(t) are known quantities. The power of energy storage in the model is P bess and capacity E bess is the decision variable to be optimized;

[0069] The specific operation process of target optimization scheduling is as follows:

[0070] (1) Data initialization: input the system structure of the distribution network in the substation, the photovoltaic output curve, the load curve after flexible scheduling, etc., and initialize the particle population at the same time;

[0071] (2) Input each system variable and use it as a particle individual to calculate the particle fitness under each constraint condition;

[0072] (3) Taking individual fitness as the input of the optimization model, the offspring particle group is obtained through particle movement;

[0073] (4) Calculate the individual optimal value based on the objective function and compare it with the group optimal value. If it is better than the group optimal value, set the individual optimal value as the group optimal value. If not, continue to iterate towards the individual optimal value.

[0074] (5) Return to step 3 until the fitness of all particles is the same as the optimal value of the group or the maximum number of iterations is reached, and the final optimization scheduling result is output.

[0075] The beneficial effects of the present invention are:

[0076] (1) The present invention proposes a method for optimizing energy storage capacity by taking into account the transferable load characteristics of rural distribution network substations. Taking into account the transferable load characteristics of rural distribution network substations, a new photovoltaic energy storage capacity configuration model is established to achieve photovoltaic absorption and reliable operation of rural distribution network substations.

[0077] (2) After considering the transferable load scheduling, the introduction of transferable load reduces the load peak-valley difference and improves the consumption of new energy. In this case, the energy storage charging and discharging period is optimized and the energy storage capacity is reduced. Compared with the case where the transferable load scheduling is not considered, the energy storage capacity is reduced by 25% and the service life of the energy storage equipment is extended by about 18%.

[0078] (3) Scenario analysis shows that the optimization of distribution network energy storage based on the dispatch of transferable load characteristics has improved the utilization rate of new energy to 96.20%, which is two percentage points higher than the scenario without considering the dispatch of transferable loads, and has improved the utilization rate of photovoltaic energy in rural areas. BRIEF DESCRIPTION OF THE DRAWINGS

[0079] Figure 1 It is the overall research framework of the present invention;

[0080] Figure 2 is the photovoltaic output curve in the present invention;

[0081] Figure 3 It is the operation process of the particle swarm algorithm in the present invention;

[0082] Figure 4 A distribution diagram of user-side transferable load before optimization in an embodiment of the present invention;

[0083] Figure 5 This is a distribution diagram of optimized transferable loads on the user side in an embodiment of the present invention;

[0084] Figure 6 The energy storage charge state in the following scenario in the embodiment of the present invention;

[0085] Figure 7 The power diagram for each time period in the following scenario in an embodiment of the present invention;

[0086] Figure 8 The energy storage charge state in scenario 2 in the embodiment of the present invention;

[0087] Fig. 9 This is a power diagram for each time period under scenario 2 in an embodiment of the present invention;

[0088] Fig.10 The energy storage charge state in scenario three in the embodiment of the present invention;

[0089] Fig.11 This is a power diagram for each time period under scenario three in an embodiment of the present invention. DETAILED DESCRIPTION

[0090] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0091] This paper takes the photovoltaic access to the grid by users in the substation as the background, and makes reasonable configuration of photovoltaic energy storage capacity for the rural substation distribution network, and considers the scheduling of transferable loads to optimize the load curve. Figure 1 shown.

[0092] Firstly, the most representative daily photovoltaic output curve of the whole year is obtained based on the k-means clustering algorithm; secondly, taking the photovoltaic absorption rate of the distribution network as the objective function and considering the load curve after the transferable load is optimized, a photovoltaic energy storage model is established to analyze the impact of DG access to the grid on the operation of BESS; finally, through a case analysis of a certain area power grid, the results show that the proposed model can reduce the voltage fluctuation of the distribution network, improve the photovoltaic absorption rate of the system, and reduce the regulation pressure of the power grid, verifying the feasibility of the model.

[0093] Example 1: Photovoltaic clustering scenario

[0094] The output of photovoltaic new energy is closely related to the light intensity and installed capacity in the scene, which seriously affects the real-time balance of the distribution network power after photovoltaic access to the distribution network. In view of the fact that the photovoltaic output in the rural distribution network needs to screen a large number of scenes to reduce the unstable impact of the photovoltaic access belt and avoid the problem of reduced accuracy caused by the processing of a large amount of photovoltaic data, the present invention selects the k-means clustering method to obtain the typical operation curve of the light load of the substation power grid, reduce the photovoltaic output scene, and improve the system certainty. The k-means clustering algorithm has the characteristics of simple algorithm, fast convergence and local optimality, and can still be reused after the data is updated. The specific process detailed steps are summarized as follows:

[0095] (1) Based on the historical data of the substation, the typical operation data of the substation's photovoltaic load is preprocessed, and the data for one year is divided into four data clusters as the photovoltaic output data for the four seasons, and the maximum value normalization processing is performed on them.

[0096] (2) Calculate the sum of the squared Euclidean distances J(C) of all data to the cluster center of the category to which they belong. Classify the category with the closest distance into one category.

[0097]

[0098] Where: u mn is a binary variable, u mn =1 means the nth sample belongs to the mth class, and 0 means the opposite; d(c m ,x n ) is the distance from the sample to the center of the cluster to which it belongs; c m is the sample cluster center; x n is other sample data in the same category.

[0099] (1) According to the division results in (2), k cluster centers c m To update the generation, Lagrange's basic principle is used to introduce J(C) as a constraint, and the objective function is selected as d(c m ,x n ), optimize it with the least squares method, and then iteratively update the four cluster centers according to the following formula.

[0100]

[0101] The judgment is made based on the basic convergence conditions. If the conditions are not met, the iteration condition is continued in step (2). If the conditions are met, the PV output data of four typical days are output.

[0102] The clustering algorithm is used to analyze the 8760-hour light data of a certain place. The results are as follows: Figure 2 shown.

[0103] Example 2 Demand Response Collaborative Optimization

[0104] Due to the dual carbon goals, the proportion of new energy in the distribution network has increased year by year, and the optimization, dispatching and control of the distribution network have also been continuously deepened. Demand-side management has also derived demand-side response.

[17] Unlike traditional load control, it is mainly based on the power users' active adjustment of their own load demand or power consumption mode according to the power market situation, so as to improve the reliability and stability of the power system.

[18] As a demand-side management method, demand-side response has expanded its regulation targets from focusing on the participation of industrial and commercial users at the beginning to changing the electricity consumption plans of urban residents through electricity price signals. Now, it can optimize the distribution network load curve through rural flexible loads, reduce the peak-to-valley gap of the system, and increase the local consumption rate of new energy.

[0105] 2.1 Analysis of demand-side characteristics

[0106] Nowadays, with the advancement of intelligent distribution network equipment construction and the rapid development of new energy applications, the flexible loads that can participate in the dispatch of power systems are increasing, and the demand-side response is gradually transitioning towards flexible load regulation. From the perspective of load regulation characteristics, flexible loads can be divided into three categories:

[19] ①Transferable load: The total power consumption of the periodic type remains unchanged, but the power consumption time is flexible, such as electric vehicle charging piles, ice storage appliances, and some residential loads. ②Transferable load: Due to the constraints of the production process, the power load can only be transferred in a continuous period, such as large industrial users. ③Reducible load: The power consumption can be reduced at a specific time, such as air conditioners and other household electrical equipment.

[0107] 2.2 Characteristics of transferable load in rural areas

[0108] With the development of rural economy, the number of electrical appliances in rural areas has gradually increased. Such household electrical appliances are affected by electricity prices and are flexible in scheduling. In terms of load characteristics, they belong to the transferable load category.

[20] As a dispatching resource between sources and loads, the advantages of transferable loads participating in the dispatching of rural distribution network systems are: transferable loads can be more flexibly dispatched in terms of energy consumption time and load size changes compared to other types of flexible loads, achieving better coordinated optimization of both the source and the load, and improving the comprehensive benefits of the new energy system.

[21] , and increase the consumption of renewable energy such as photovoltaics.

[0109] Based on the energy consumption characteristics of the transferable load equipment, the transfer range of the transferable load and the constraints on the increase in electricity prices are determined, the load affiliation index is established, and the transferable load is dispatched with the goal of minimizing the load peak-to-valley difference, thereby reducing the source-load imbalance and improving system safety.

[0110] Constrain electricity prices and achieve optimal scheduling of transferable loads.

[0111] Note: t tran It is the transferable load at a certain moment. The transfer period of the transferable load is from 6 pm to 6 am. 0-1 is used to indicate whether the transferable load is transferred in a certain period, and β is used to mark it.

[0112] 1) Transferable load power range

[0113]

[0114] In the formula, is the minimum value of the transferable load equipment power, It is the maximum value of the transferable load equipment power.

[0115] 2) Calculate the membership degree for 24 hours a day

[0116] The difference between the load power of each hour and the average power in a day is selected as the membership degree μ i .

[0117]

[0118] Where P L (t) is the load power of the substation.

[0119] 3) Transferable load dispatch constraints

[0120] The operation scheduling of transferable loads will have an impact on the user's life. Therefore, a penalty function based on transferable load scheduling is proposed to constrain it.

[0121]

[0122] Where f is the coefficient of the penalty function.

[0123] 4) Constraints on electricity price fluctuations

[0124] Through electricity price regulation, transferable load scheduling can be achieved, increasing electricity prices during load peaks and reducing electricity prices during load valleys.

[0125]

[0126] In the formula, p f is the average electricity price, i.e. the electricity price during normal periods, Δp i Adjustment range of peak and valley electricity prices.

[0127] 5) Objective function

[0128] The product of the membership degree of the corresponding time period and the power of the corresponding time period.

[0129]

[0130] The present invention achieves the goal of optimizing the load curve by the objective function by establishing a membership index that is mutually related to load variation and electricity price fluctuation.

[0131] Example 3: Photovoltaic storage capacity configuration model considering transferable load characteristics

[0132] Based on k-means clustering, scenario analysis is performed to obtain a typical daily photovoltaic output curve, which is compared with the load curve to determine the charging and discharging period. When the load curve is at the peak, discharge, and when the load curve is at the trough, discharge. Generally, the energy storage system uses the charging and discharging power as the control variable, and determines the BESS charging and discharging strategy through certain constraints and objective functions. Most of the existing literature uses this method, which can ensure the balance of cycle charging and discharging, but does not consider the continuous change of energy storage capacity within the cycle, that is, it cannot guarantee that the discharge or charge amount in a continuous discharge period or charging period is less than or equal to the previous BESS remaining power or capacity, which may cause E dis,1 >E e,1 situation, resulting in an erroneous solution that does not match the actual situation.

[0133] The discharge time period is determined by the load and photovoltaic output curve, and the discharge capacity constraint within the power period is added to the charge and discharge power optimization solution method to ensure the feasibility of the solution.

[0134] 3.1 Objective function of the photovoltaic energy storage capacity configuration model

[0135] Taking the charging and discharging period and charging and discharging power as decision variables, considering the constraints such as the station voltage and charging and discharging power, the maximum photovoltaic absorption rate is solved. A reasonable division of charging and discharging strategies within 24 hours is achieved, and the photovoltaic absorption rate is maximized.

[0136] The objective function is the photovoltaic consumption rate within 24 hours in the substation area:

[0137]

[0138] P SC (t) = min{P PV (t),P out (t)} (9)

[0139] Where: R sel is the local consumption rate of new energy; P PV (t) is the photovoltaic power output of the new energy in the area, P L-a (t) is the new energy output consumed by the load in real time in the distribution network of the substation; P bat-charge (t) is the charging power of the energy storage during time period t, and P bat-charge (t)>0.

[0140] 3.2 Constraints

[0141] (1) Substation voltage constraints

[0142] The substation must maintain a single-phase power supply voltage deviation of +7% and -10% for 0.22kv, and a power supply voltage deviation of ±7% of the rated voltage for 10kv and below, to ensure the power quality of BESS operation.

[0143] U i,min ≤U i ≤U i,max i=1,2,…,n (10)

[0144] U i,min , U i,max are the upper and lower limits of the voltage at node i;

[0145] (2) Energy storage system operation constraints

[0146] In order to ensure the safe operation of energy storage, the energy storage charging and discharging power is guaranteed not to exceed the charging and discharging power of the battery;

[0147] -P bess ≤P bat (t)≤P bess (11)

[0148] Among them, P bat (t) is the charging and discharging power of the energy storage system at each time period, P bess It is the maximum allowable charging and discharging power of energy storage. Positive value represents discharging and negative value represents charging.

[0149] Charging process:

[0150] Discharge process:

[0151] Among them, SOC(t+Δt) is the remaining power level at time t+Δt and t, that is, the state of charge; ε is the loss rate of the remaining power of BESS per hour, referred to as the self-discharge rate, in % / h; α and β are the charging and discharging efficiencies of BESS, respectively; E e is the capacity of BESS; t is the simulation time; Δt is the sampling interval, i.e. the time period during simulation, which is 1h here;

[0152] (3) Energy storage system charge state constraints

[0153] To avoid the impact of excessive charging and discharging of the battery on its service life, the state of charge is limited to a certain range to ensure that the battery is neither fully charged nor fully discharged;

[0154]

[0155] Where Smin , S max are the minimum and maximum states of charge of the energy storage, respectively; S(0) and S(24) are the states of charge of the energy storage system at time periods 0 and 24, respectively;

[0156] (4) Dischargeable Capacity Constraints

[0157] In order to make the solution applicable in actual situations and avoid the situation where the continuous discharge period is greater than the remaining power in the energy storage, or the continuous charge amount is greater than the energy storage content, certain constraints are imposed on the charge and discharge amount.

[0158] E dis,i ≤E soc,i (15)

[0159] Among them, E dis,i is the dischargeable capacity in the i-th continuous discharge time period; E soc,i is the remaining capacity of BESS before discharge in the i-th continuous discharge time period.

[0160] (5) Energy conservation constraints for charging and discharging within a daily cycle

[0161] The substation must maintain a balance between the real-time power it interacts with users and the real-time power it interacts with the distribution network and new energy sources.

[0162] P PV (t)+P grid (t)+P bat (t) = P L (t) (16)

[0163] P PV (t) is the photovoltaic power output of the new energy in the area, P grid (t) is the power flowing from one node to the next node in the distribution network, P bat (t) is the charging and discharging power of the energy storage during period t, P L (t) is the load power of the substation.

[0164] Example 4: Application of Particle Swarm Optimization Model to Solve

[0165] Transferable load scheduling is a typical integer quadratic programming problem. The present invention solves the transferable load scheduling through the CPLEX solver in the Matlab toolbox, obtains the load curve after the transferable load scheduling, and reduces the negative impact of the access to the distribution network system in the substation area.

[0166] The source-load-storage collaborative planning model of BESS uses the PSO intelligent optimization algorithm to solve it. The PSO algorithm has good global search performance and fast convergence speed, and can quickly solve the energy storage capacity and power configuration when the photovoltaic absorption rate in the platform area is the maximum. The solution process is as follows Figure 3As shown. Among them, the predicted output of new energy photovoltaic power generation P PV (t) and load forecast power demand P0(t) are known quantities. The power of energy storage in the model is P bess and capacity E bess is the decision variable to be optimized.

[0167] The specific operation process of target optimization scheduling is as follows:

[0168] (1) Data initialization. Input the system structure of the distribution network in the substation, the photovoltaic output curve, the load curve after flexible compliance scheduling, etc., and initialize the particle population at the same time.

[0169] (2) Input various system variables and use them as individual particles to calculate the particle fitness under various constraints.

[0170] (3) The individual fitness is used as the input of the optimization model, and the offspring particle swarm is obtained through particle movement.

[0171] (4) According to the objective function calculation, the individual optimal value is obtained and compared with the group optimal value. If it is better than the group optimal value, the individual optimal value is set as the group optimal value. If not, it continues to iterate towards the individual optimal value.

[0172] (5) Return to step 3 until the fitness of all particles is the same as the optimal value of the group or the maximum number of iterations is reached, and the final optimization scheduling result is output.

[0173] Example 5 Analysis

[0174] 5.1 Scene Setup

[0175] The example selected a 0.4KV rural distribution station, where the load demand change trend is not obvious. The transferable load data obtained by the clustering algorithm is solved by cplex. Through the constraint of the membership function, the peak-to-valley difference of the load curve is reduced, the load curve is improved for photovoltaic absorption, the volatility of the power grid output is reduced, and the distribution network safety of the station area is improved. The transferable load scheduling is shown in Table 1.

[0176] Table 1 Transferable load parameters

[0177]

[0178] The distribution diagram of user-side transferable load before and after optimization is as follows Figure 4 and Figure 5 As shown,

[0179] contrast Figure 4 and Figure 5, the transferable load is mainly distributed in the time period of 10:00-16:00, which is just at the peak of the load curve, while the optimized time period is more dispersed, reducing the overall peak-to-valley difference of the load curve. Since the transferable load can flexibly change the load duration and transfer power, it has flexibility in optimizing scheduling, and has the effect of shaving peaks and filling valleys on the load curve, which can reduce the configured energy storage capacity and reduce the required cost.

[0180] Since photovoltaic output varies significantly with the seasons, this substation is divided into four operating conditions according to photovoltaic changes, and the final selected time scenario is analyzed through the three parameters of new energy consumption rate, energy storage capacity and energy storage power. The new energy consumption rate reflects the impact of photovoltaics on the substation's distribution network during this period. The two indicators of energy storage capacity and energy storage power reflect the demand for energy storage configuration in the current period. The required configuration capacity of the substation is finally determined by comparing the three indicators in four different time periods.

[0181] Table 2 Photovoltaic energy storage configuration in four seasons

[0182]

[0183] As shown in Table 2, the energy storage capacity required on a typical day in summer is the largest. This is because the photovoltaic intensity is the highest and the photovoltaic output is the largest in summer. The substation area needs a larger energy storage capacity than other seasons to ensure the local consumption of photovoltaic energy. The energy storage capacity required on a typical day in winter is the smallest, and the new energy consumption rate is the largest, reaching 98.39%, but the new energy output is the smallest at this time. Therefore, in order to ensure the photovoltaic consumption rate and distribution network operation safety of the distribution substation area, the summer typical day is selected as the optimization energy storage model, and the following three scenarios are set to analyze the optimization method of substation energy storage capacity configuration.

[0184] Scenario 1: Without considering the constraints of energy storage capacity conditions and without limiting the charging and discharging power of energy storage, this is the current distribution network operation scenario in the substation.

[0185] Scenario 2: Without considering flexible load scheduling, the energy storage model is optimized and scheduled based on the original load curve as the current distribution network operation scenario.

[0186] Scenario 3: Taking into account the energy storage discharge capacity constraints and flexible load scheduling proposed in the present invention, the energy storage model is optimized and scheduled.

[0187] 5.2 Results Analysis

[0188] The energy storage optimization dispatch results of the above three operation scenarios are as follows: Figure 4-Figure 11 As shown, Figure 8-Figure 11 This is a comparison picture of scene two and scene three.

[0189] Figure 6 and Figure 7Based on the configuration of scenario 1, the energy storage charging and discharging situation after the flexible load scheduling is optimized without considering the energy storage restriction. Without the restriction of energy storage capacity and charging and discharging power, the energy storage shows a high degree of participation in ensuring the balance of supply and demand of the distribution network in the substation. The number of discharges increases significantly, and the division of charging and discharging periods is not clear, showing a relatively continuous alternating state of charging and discharging. Although this high-frequency charging and discharging behavior improves the photovoltaic absorption rate, the capacity of the energy storage equipment increases significantly. Fig.10 The energy storage capacity of the 480kW·h battery is 650kW·h, and multiple charging and discharging will also shorten its service life. Although the photovoltaic absorption rate has increased to 97.68%, the improvement in photovoltaic absorption rate is not significant compared with scenario three. This result shows that simply increasing the energy storage capacity without considering the coordinated optimization of energy storage constraints and transferable loads cannot effectively promote the effective improvement of photovoltaic absorption rate, and will also lead to the problem of shortening the service life of energy storage equipment.

[0190] Figure 8 and Fig. 9 In scenario 2, the flexible load is not dispatched, and the energy storage charging and discharging after the optimization of the distribution network operation in the substation is considered with energy storage limitation. At this time, the photovoltaic consumption rate is 95.31% and the energy storage capacity is 540kW·h. Figure 6 It can be seen that the energy storage charging and discharging period at this time is more reasonable than that in scenario 1, but the grid output fluctuates greatly, and it is difficult to guarantee the safe operation of the substation grid. In order to ensure the charge state and discharge capacity constraints of the energy storage, it does not have a good effect on photovoltaic absorption.

[0191] Compared with Scenario 2, Scenario 3 has a more moderate power demand from the power grid in the distribution network of the substation, and has less impact on the power quality of the system. Compared with Scenario 1, its charging period arrangement is more reasonable, with fewer charging and discharging periods in a day, and the service life of the energy storage equipment is longer. The energy storage capacity of Scenario 3 is 480kW·h, which is better than Scenario 2. The photovoltaic absorption rate is 96.20%, which is also higher than Scenario 2, and can ensure that the system absorption rate is above a certain level.

[0192] Table 3 Energy storage configuration and photovoltaic absorption rate under different scenarios

[0193]

[0194] In summary, compared with other scenarios, the optimization scheduling model and solution method proposed in the present invention can significantly optimize the photovoltaic absorption rate of the distribution network in the substation area. In the above-mentioned transferable load optimization scheduling operation analysis, the transferable load plays an important role in reducing the abandonment rate of light in the rural distribution network. When the transferable load meets the demand-side response, the capacity demand of the energy storage system is reduced to 480kW·h, which reduces the number of charging and discharging times by half, and reduces the proportion of transferable loads in a single period, thereby increasing the photovoltaic absorption rate of the system and ensuring the safe operation of the system.

[0195] Conclusion: Based on the photovoltaic characteristics of the distribution network in the substation, this paper proposes an energy storage capacity optimization model that takes into account the transferable load characteristics of the rural distribution network substation.

[0196] (1) The present invention proposes a method for optimizing energy storage capacity by taking into account the transferable load characteristics of rural distribution network substations. Taking into account the transferable load characteristics of rural distribution network substations, a new photovoltaic energy storage capacity configuration model is established to achieve photovoltaic absorption and reliable operation of rural distribution network substations.

[0197] (2) After considering the transferable load scheduling, the introduction of transferable load reduces the load peak-valley difference and improves the consumption of new energy. In this case, the energy storage charging and discharging period is optimized and the energy storage capacity is reduced. Compared with the case where the transferable load scheduling is not considered, the energy storage capacity is reduced by 25% and the service life of the energy storage equipment is extended by about 18%.

[0198] (3) Scenario analysis shows that the optimization of distribution network energy storage based on the dispatch of transferable load characteristics has improved the utilization rate of new energy to 96.20%, which is two percentage points higher than the scenario without considering the dispatch of transferable loads, and has improved the utilization rate of photovoltaic energy in rural areas.

[0199] The energy storage capacity optimization method only optimizes the scheduling of transferable loads. With the development of electrical equipment, the types of load resources on the demand side of the rural distribution network will gradually increase. Subsequent research can consider the optimal scheduling of transferable loads and reducible loads, and continuously optimize and improve the energy storage configuration of the rural distribution network.

[0200] It will be apparent to those skilled in the art that the invention is not limited to the details of the exemplary embodiments described above and that the invention can be implemented in other specific forms without departing from the spirit or essential features of the invention. Therefore, the embodiments should be considered exemplary and non-limiting in all respects, and the scope of the invention is defined by the appended claims rather than the foregoing description, and it is intended that all variations falling within the meaning and scope of the equivalent elements of the claims be included in the invention. Any reference numeral in a claim should not be considered as limiting the claim to which it relates.

Claims

1. An energy storage capacity optimization model considering the transferable load characteristics of rural distribution network stations is characterized by: The steps include: Step 1: Obtain the most representative daily photovoltaic output curve throughout the year based on the k-means clustering algorithm; Step 2: Demand response collaborative optimization; specifically including demand side feature analysis and rural area transferable load characteristics; Step 3: A photovoltaic storage capacity configuration model considering transferable load characteristics, including objective functions and constraints; Step 4: Apply the particle swarm algorithm model to solve the problem.

2. The energy storage capacity optimization model considering the transferable load characteristics of the rural distribution network area according to claim 1 is characterized by: In step 1, the specific steps of obtaining the most representative daily photovoltaic output curve throughout the year based on the k-means clustering algorithm are as follows: Step (1), based on the historical data of the substation, pre-process the typical operation data of the substation's photovoltaic load, divide the data of one year into four data clusters as the photovoltaic output data of the four seasons, and perform maximum value normalization processing on them; Step (2), calculate the sum of the squared Euclidean distances J(C) of all data to the cluster center of the corresponding category, and classify the category with the closest distance into one category; Where: u mn is a binary variable, u mn =1 means the nth sample belongs to the mth class, and 0 means the opposite; d(c m ,x n ) is the distance from the sample to the center of the cluster to which it belongs; c m is the sample cluster center; x n For other sample data in the same category; Step (3): According to the division result of step (2), k cluster centers c m To update the generation, Lagrange's basic principle is used to introduce J(C) as a constraint, and the objective function is selected as d(c m ,x n ), optimize it by the least square method, and then iteratively update the 4 cluster centers according to the following formula; The judgment is made according to the basic convergence conditions. If it does not meet the convergence conditions, the iteration conditions are continued in equation (2). If it meets the convergence conditions, the photovoltaic output data of four typical days are output.

3. The energy storage capacity optimization model considering the transferable load characteristics of the rural distribution network area according to claim 1 is characterized by: In step 2, in the demand side characteristic analysis, flexible loads can be divided into three categories from the perspective of load regulation characteristics, specifically: ① Transferable loads: the total power consumption of the periodic type remains unchanged, but the power consumption time is flexible, such as electric vehicle charging piles, ice storage appliances, some residential loads, etc.; ② Shiftable loads: loads that are constrained by the production process and can only be shifted in continuous time periods, such as large industrial users, etc.; ③ Load reduction: The power consumption can be reduced at specific times, such as air conditioners and other household electrical equipment.

4. The energy storage capacity optimization model considering the transferable load characteristics of the rural distribution network area according to claim 1 is characterized by: In step 2, among the transferable load characteristics of the rural area, the transferable load can be more flexibly scheduled in terms of energy consumption time and load size changes compared to other types of flexible loads, achieving better coordinated optimization of both the source and the load, improving the comprehensive benefits of the new energy system, and increasing the absorption of renewable energy such as photovoltaics; the following is based on the energy consumption characteristics of the transferable load equipment, determining the transfer range of the transferable load and the constraints on the increase in electricity prices, establishing a load membership index, and scheduling the transferable load with the goal of minimizing the peak-to-valley difference of the load, reducing the source-load imbalance, and improving system safety; Constrain the electricity price and realize the optimal dispatch of transferable load, denoted by P t tran is the load that can be transferred at a certain moment. The transfer period of the transferable load is from 6 pm to 6 am. 0-1 is used to indicate whether the transferable load is transferred during a certain period, and β is used to mark it; (1) Transferable load power range In the formula, is the minimum value of the transferable load equipment power, The maximum value of the transferable load equipment power; (2) Calculate the membership degree for 24 hours a day The difference between the load power of each hour and the average power in a day is selected as the membership degree μ i ; Where P L (t) is the load power of the substation; (3) Transferable load dispatch constraints The operation and scheduling of transferable loads will have an impact on the user's life. Therefore, a penalty function based on transferable load scheduling is proposed to constrain it. Where f is the coefficient of the penalty function; (4) Constraints on electricity price fluctuations Through electricity price regulation, transferable load scheduling is achieved, with electricity prices increased during peak load periods and reduced during valley load periods; In the formula, p f is the average electricity price, i.e. the electricity price during normal periods, Δp i The adjustment range of peak and valley electricity prices; (5) Objective function The product of the membership degree of the corresponding time period and the power of the corresponding time period; By establishing a membership index that correlates load changes and electricity price fluctuations, the goal of optimizing the load curve with the objective function is achieved.

5. The energy storage capacity optimization model considering the transferable load characteristics of the rural distribution network area according to claim 1 is characterized by: In step three, in the objective function, the charging and discharging period and the charging and discharging power are used as decision variables, and the constraints such as the area voltage and the charging and discharging power are considered to solve the maximum photovoltaic absorption rate as the goal, so as to achieve a reasonable division of the charging and discharging strategy within 24 hours and maximize the photovoltaic absorption rate; the objective function is the photovoltaic absorption rate within the area within 24 hours: P SC (t)=min{P PV (t),P out (t)} (9) Where: R sel is the local consumption rate of new energy; P PV (t) is the photovoltaic power output of the new energy in the area, P L-a (t) is the new energy output consumed by the load in real time in the distribution network of the substation; P bat-charge (t) is the charging power of the energy storage during time period t, and P bat-charge (t)>0.

6. The energy storage capacity optimization model considering the transferable load characteristics of the rural distribution network area according to claim 1 is characterized by: In step three, the constraints include: (1) Substation voltage constraints The substation area must maintain a single-phase power supply voltage deviation of +7% and -10% for 0.22kv, and a power supply voltage deviation of ±7% of the rated voltage for 10kv and below, to ensure the power quality of BESS operation; U i,min ≤U i ≤U i,max i=1,2,...,n (10) U i,min , U i,max are the upper and lower limits of the voltage at node i; (2) Energy storage system operation constraints In order to ensure the safe operation of energy storage, the energy storage charging and discharging power is guaranteed not to exceed the charging and discharging power of the battery; -P bess ≤P bat (t)≤P bess (11) Among them, P bat (t) is the charging and discharging power of the energy storage system at each time period, P bess The maximum allowable charging and discharging power of energy storage. Positive values ​​represent discharging, and negative values ​​represent charging. Charging process: Discharge process: Among them, SOC(t+Δt) is the remaining power level at time t+Δt and t, that is, the state of charge; ε is the loss rate of the remaining power of BESS per hour, referred to as the self-discharge rate, in % / h; α and β are the charging and discharging efficiencies of BESS, respectively; E e is the capacity of BESS; t is the simulation time; Δt is the sampling interval, i.e. the time period during simulation, which is 1h here; (3) Energy storage system charge state constraints To avoid the impact of excessive charging and discharging of the battery on its service life, the state of charge is limited to a certain range to ensure that the battery is neither fully charged nor fully discharged; Where S min , S max are the minimum and maximum states of charge of the energy storage, respectively; S(0) and S(24) are the states of charge of the energy storage system at time periods 0 and 24, respectively; (4) Dischargeable Capacity Constraints In order to make the solution applicable in actual situations, to avoid the situation where the continuous discharge period is greater than the remaining power in the energy storage, or the continuous charge amount is greater than the energy storage content, certain constraints are imposed on the charge and discharge amount; AND dis,i ≤E soc,i (15) Among them, E dis,i is the dischargeable capacity in the i-th continuous discharge time period; E soc,i is the remaining power of BESS before discharge in the i-th continuous discharge time period; (5) Energy conservation constraints for charging and discharging within a daily cycle The substation must maintain a balance between the real-time power it exchanges with users and the real-time power it exchanges with the distribution network and new energy sources; P PV (t)+P grid (t)+P bat (t)=P L (t) (16) P PV (t) is the photovoltaic power output of the new energy in the area, P grid (t) is the power flowing from one node to the next node in the distribution network, P bat (t) is the charging and discharging power of the energy storage during period t, P L (t) is the load power of the substation.

7. The energy storage capacity optimization model considering the transferable load characteristics of the rural distribution network area according to claim 1 is characterized by: In step 3, the transferable load scheduling is a typical integer quadratic programming problem. The transferable load scheduling is solved by the CPLEX solver in the Matlab toolbox to obtain the load curve after the transferable load scheduling, so as to reduce the negative impact of the access to the distribution network system in the substation area. The source-load-storage collaborative planning model of BESS is solved by PSO intelligent optimization algorithm. PSO algorithm has good global search performance and fast convergence speed. It can quickly solve the energy storage capacity and power configuration when the photovoltaic absorption rate in the platform area is the largest. PV (t) and load forecast power demand P0(t) are known quantities. The power of energy storage in the model is P bess and capacity E bess is the decision variable to be optimized; the specific operation process of target optimization scheduling is as follows: (1) Data initialization: input the system structure of the distribution network in the substation, the photovoltaic output curve, the load curve after flexible scheduling, etc., and initialize the particle population at the same time; (2) Input each system variable and use it as a particle individual to calculate the particle fitness under each constraint condition; (3) Taking individual fitness as the input of the optimization model, the offspring particle group is obtained through particle movement; (4) Calculate the individual optimal value based on the objective function and compare it with the group optimal value. If it is better than the group optimal value, set the individual optimal value as the group optimal value. If not, continue to iterate towards the individual optimal value. (5) Return to step 3 until the fitness of all particles is the same as the optimal value of the group or the maximum number of iterations is reached, and the final optimization scheduling result is output.

Images