技术领域technical field
本发明涉及一种微电网能量调度优化方法，涉及微电网能量管理技术领域，适用于微电网能量管理的日前经济调度。The invention relates to a micro-grid energy scheduling optimization method, relates to the technical field of micro-grid energy management, and is suitable for day-ahead economic scheduling of micro-grid energy management.
背景技术Background technique
随着能源危机加剧和环境污染日益严重，微电网作为一种新型能源网络化供应和管理技术，受到越来越多的关注。随着微电网工程的不断发展完善，微电网能量管理系统成为微电网研究的重点。As the energy crisis intensifies and environmental pollution becomes more and more serious, microgrid, as a new energy network supply and management technology, has received more and more attention. With the continuous development and improvement of microgrid engineering, microgrid energy management system has become the focus of microgrid research.
微电网能量管理系统是针对微电网的智能控制系统，其能优化分布式电源的功率分配，并使系统经济安全的运行。作为微电网能量管理的核心部分，其调度优化模型的设计对微电网能量管理的效果起着至关重要的作用。如何选择算法快速准确的求解考虑各种复杂约束条件的调度优化问题均给该课题带来较大的挑战。解决这一系列问题将提高能量的利用效率，减少微电网系统的运营成本，具有较好的社会效益。The microgrid energy management system is an intelligent control system for microgrids, which can optimize the power distribution of distributed power sources and make the system operate economically and safely. As the core part of microgrid energy management, the design of its scheduling optimization model plays a vital role in the effect of microgrid energy management. How to choose an algorithm to quickly and accurately solve the scheduling optimization problem considering various complex constraints has brought great challenges to this subject. Solving this series of problems will improve energy utilization efficiency, reduce the operating cost of the microgrid system, and have good social benefits.
现阶段，一些微电网调度优化的研究建立了线性的调度模型，但是这种模型过于粗糙，不利于深入研究。当前大多数模型建立的是非线性优化模型，一般采用粒子群算法、蚁群算法、遗传算法等智能优化算法求解。随着规模的增加，智能算法往往不容易收敛并且求解速度较慢。同时，当前模型一般只适用于少数微电网算例，模型的适用性有待提高。At this stage, some microgrid scheduling optimization studies have established a linear scheduling model, but this model is too rough to be conducive to in-depth research. At present, most models are established as nonlinear optimization models, which are generally solved by intelligent optimization algorithms such as particle swarm algorithm, ant colony algorithm, and genetic algorithm. As the scale increases, smart algorithms tend not to converge easily and are slower to solve. At the same time, the current model is generally only suitable for a small number of microgrid calculation examples, and the applicability of the model needs to be improved.
发明内容Contents of the invention
为了克服现有技术下的上述缺陷，本发明的目的在于提供一种基于能量管理的孤网状态下的微电网系统调度方法，该方法预测更加精准，在保证较高求解精度的基础上还能获得较快的求解速度。In order to overcome the above-mentioned defects in the prior art, the purpose of the present invention is to provide a microgrid system scheduling method based on energy management in the isolated grid state. Get a faster solution speed.
本发明的技术方案是：Technical scheme of the present invention is:
一种基于能量管理的孤网状态下的微电网系统调度方法，其以微电网孤网运行成本最小为控制目标，在孤网长期运行状态下，保持储能装置剩余电量的正常波动范围，通过对下列孤网长期运行目标函数的运行成本最小化求解获得相应的长期运行调度参数，保证微电网系统内一级负荷的用电需求：A microgrid system scheduling method based on energy management in the isolated grid state, which takes the minimum operating cost of the isolated grid as the control goal, and maintains the normal fluctuation range of the remaining power of the energy storage device under the long-term operation of the isolated grid, through The operation cost minimization of the following long-term operation objective function of the isolated grid is solved to obtain the corresponding long-term operation scheduling parameters to ensure the power demand of the primary load in the microgrid system:
孤网长期运行目标函数为：The long-term running objective function of the isolated network is:
在孤网短期运行的情况下，先将储能装置作为电源投入供电，保证所有负荷不断电直到电源电量不足，在出现电源电量不足的情形后，通过对下列孤网短期运行目标函数的运行成本最小化求解获得相应的短期运行调度参数：In the case of short-term operation of the isolated grid, the energy storage device is first put into power supply as a power supply to ensure that all loads are continuously powered until the power supply is insufficient. After the power supply is insufficient, the operating cost of the following short-term operation objective function of the isolated grid The minimization solution obtains the corresponding short-run scheduling parameters:
孤网短期运行目标函数为：The short-term operation objective function of the isolated network is:
其中，C为微电网系统的运行成本，Pi(t)为可控电源i的t时刻的输出功率，Ci(Pi(t))为可控电源i的t时刻的燃料成本，所述可控电源包括微型燃气轮机、柴油机和燃料电池；Among them, C is the operating cost of the microgrid system, Pi (t) is the output power of controllable power source i at time t, and Ci (Pi (t)) is the fuel cost of controllable power source i at time t, so The controllable power sources include micro gas turbines, diesel engines and fuel cells;
Si(t)为可控电源i的t时刻的启动成本；Uopen,i(t)为可控电源i的t时刻0、1开机决策变量，0表示非启动(保持原启停状态或者停机)，1表示启动；Si (t) is the start-up cost of controllable power supply i at time t; Uopen,i (t) is the start-up decision variable of controllable power supply i at time 0 and 1, 0 means non-start (keep the original start-stop state or stop), 1 means start;
COM,i(Pi(t))、CDP,i(Pi(t))、Ce(Pi(t))分别为可控电源i的t时刻的维护成本、折旧成本和环境成本；COM,i (Pi (t)), CDP,i (Pi (t)), Ce (Pi (t)) are the maintenance cost, depreciation cost and environmental cost;
Pk(t)为蓄电池k的t时刻的输出功率，COM,k(Pk(t))、CDP,k(Pk(t))分别为蓄电池k的t时刻的维护成本和折旧成本；Pk (t) is the output power of battery k at time t, COM,k (Pk (t)), CDP,k (Pk (t)) are the maintenance cost and depreciation of battery k at time t cost;
CPUNISH,k为蓄电池周期不平衡罚函数；CPUNISH, k is the penalty function of battery cycle unbalance;
Pl(t)为t时刻负荷节点l的负荷，Cl(Pl(t))为负荷节点l的切负荷费用函数，Ul(t)为负荷节点l的切负荷0、1决策变量，0表示不切负荷，1表示切负荷；Pl (t) is the load of load node l at time t, Cl (Pl (t)) is the load shedding cost function of load node l, Ul (t) is the load shedding 0, 1 decision variable of load node l , 0 means no load shedding, 1 means load shedding;
为一级负荷l1的切负荷0、1决策变量，0表示不切负荷，1表示切负荷，为一级负荷节点l1的罚函数的系数； It is the decision variable of load shedding 0 and 1 for primary load l1 , 0 means no load shedding, 1 means load shedding, is the coefficient of the penalty function of the first-level load node l1 ;
al为负荷节点l的罚函数的系数；al is the coefficient of penalty function of load node l;
I为可控电源的数量；K为蓄电池的数量；T、T’为一个调度周期内的优化时段数，L为可中断负荷的负荷节点个数，L1表示一级负荷节点的个数。I is the number of controllable power sources; K is the number of batteries; T, T' are the number of optimization periods in a dispatch cycle, L is the number of load nodes that can interrupt the load, and L1 is the number of primary load nodes.
本发明的有益效果为：The beneficial effects of the present invention are:
由于对微电网的各类电源分别进行了分类归纳，对不同类型的电源分别建模，由此可以将微电网大部分电源纳入日前经济调度模型中，且新增元件也很简单，因此具有很强的适用性。Since the various power sources of the microgrid are classified and summarized, and different types of power sources are modeled separately, most of the power sources of the microgrid can be included in the day-ahead economic dispatch model, and the new components are also very simple, so it has great potential. Strong applicability.
由于模型中综合考虑了设备折旧费用、燃料费用、维护费用、启停费用、环保费用等设备运行成本及切负荷成本、线路损耗成本等，同时满足了功率平衡、备用、爬坡率、出力上下限、最小启停时间、蓄电池容量上下限、蓄电池充放电功率上下限、蓄电池周期充放电总能量等约束条件，比现有技术下的调度方法更贴近微电网的实际需求，由此建立起的微电网日前经济调度模型也更为完善。Since the model comprehensively considers equipment depreciation costs, fuel costs, maintenance costs, start-up costs, environmental protection costs and other equipment operating costs, load shedding costs, line loss costs, etc., it also satisfies power balance, backup, ramp rate, and output. The lower limit, the minimum start-stop time, the upper and lower limits of battery capacity, the upper and lower limits of battery charge and discharge power, and the total energy of battery cycle charge and discharge are closer to the actual needs of the microgrid than the dispatching method under the existing technology. The economic scheduling model of the microgrid is also more perfect.
由于模型求解过程中采用了按时间分层求解的优化求解方法，可根据需要设置不同时间尺度，获得很短时间范围(自由设定，例如可以为10min或者15min等)的求解结果，因此可以显著提高调度的精度。Since the optimal solution method of layered solution by time is used in the model solution process, different time scales can be set according to the needs, and the solution results in a very short time range (freely set, such as 10min or 15min, etc.) can be obtained, so it can be significantly Improve scheduling accuracy.
由于引入了采用前推回代法的三相潮流计算方法，可以在三相不平衡的情况下计算微电网的潮流，得出微电网的线损，计算出微电网平均线损系数。Due to the introduction of the three-phase power flow calculation method using the forward push-back method, the power flow of the micro-grid can be calculated in the case of three-phase imbalance, the line loss of the micro-grid can be obtained, and the average line loss coefficient of the micro-grid can be calculated.
由于应用了分段线性化的处理方法将非线性问题转化为线性问题，并引入了0、1决策变量，将约束中的二次约束转换为一次约束，使微电网日前经济调度模型更易于求解。Due to the application of the piecewise linearization processing method to transform the nonlinear problem into a linear problem, and the introduction of 0, 1 decision variables, the secondary constraints in the constraints are converted into primary constraints, making the microgrid day-ahead economic dispatch model easier to solve .
在应用混合整数线性规划算法对微电网日前经济调度模型进行求解的基础上，加入按时间分层优化的方法，很好地同时保证了计算的准确性和快速性。On the basis of applying the mixed integer linear programming algorithm to solve the microgrid's day-ahead economic dispatch model, adding the method of hierarchical optimization according to time, it can well ensure the accuracy and rapidity of the calculation at the same time.
附图说明Description of drawings
图1是费用曲线线性化示意图；Figure 1 is a schematic diagram of the linearization of the cost curve;
图2是配电网前推回代法流程图；Fig. 2 is a flow chart of the distribution network push-back method;
图3是分层优化的时间周期对应关系；Fig. 3 is the time cycle corresponding relationship of hierarchical optimization;
图4是分层优化流程图；Fig. 4 is a hierarchical optimization flow chart;
图5是微电网结构图；Fig. 5 is a microgrid structural diagram;
图6是微电网典型日负荷曲线；Figure 6 is a typical daily load curve of a microgrid;
图7是光伏出力典型场景；Figure 7 is a typical scenario of photovoltaic output;
图8是预测误差5％时，光伏预测出力曲线；Fig. 8 is the photovoltaic forecast output curve when the forecast error is 5%;
图9是孤网长期稳定运行时计算得到的理论切负荷结果；Figure 9 is the theoretical load shedding results calculated during the long-term stable operation of the isolated grid;
图10是孤网长期稳定运行时计算得到的理论调度结果；Figure 10 is the theoretical scheduling result calculated during the long-term stable operation of the isolated network;
图11是孤网长期稳定运行时实际运行的切负荷结果；Figure 11 is the load shedding result of the actual operation of the isolated grid during long-term stable operation;
图12是孤网长期稳定运行时实际运行的调度结果；Figure 12 is the scheduling result of the actual operation when the isolated network runs stably for a long time;
图13是孤网短期运行时计算得到的理论切负荷结果；Figure 13 is the theoretical load shedding results calculated during the short-term operation of the isolated grid;
图14是孤网短期运行时计算得到的理论调度结果；Figure 14 is the theoretical scheduling result calculated during the short-term operation of the isolated network;
图15是孤网短期运行时实际运行的切负荷结果；Figure 15 is the load shedding result of the actual operation of the isolated grid during short-term operation;
图16是孤网短期运行时实际运行的调度结果；Figure 16 is the scheduling result of the actual operation during the short-term operation of the isolated network;
图17是采用现有调度方法的实际切负荷结果；Figure 17 is the actual load shedding result using the existing scheduling method;
图18是采用现有调度方法的实际调度结果。Fig. 18 is the actual scheduling result using the existing scheduling method.
具体实施方式Detailed ways
本发明涉及一种基于能量管理的孤网状态下的微电网系统调度方法，其以微电网孤网运行成本最小为控制目标，通过对本发明所构建的孤网长期运行目标函数进行运行成本最小化求解并获得相应的长期运行调度参数，保证微电网系统内一级负荷的用电需求，通过对本发明所构建的孤网短期运行目标函数进行运行成本最小化求解并获得相应的短期运行调度参数，保证微电网系统内所有用电设备的用电需求。The present invention relates to a microgrid system scheduling method based on energy management in the isolated grid state, which takes the minimum operating cost of the isolated grid of the microgrid as the control goal, and minimizes the operating cost through the long-term operation objective function of the isolated grid constructed by the present invention Solve and obtain the corresponding long-term operation scheduling parameters to ensure the power demand of the first-level load in the microgrid system, and obtain the corresponding short-term operation scheduling parameters by minimizing the operating cost of the isolated grid short-term operation objective function constructed by the present invention, Guarantee the electricity demand of all electrical equipment in the microgrid system.
所述目标函数的构建直至求解包括以下步骤：The construction of the objective function until the solution comprises the following steps:
步骤一：建立微电网中不同电源、储能装置的调度模型；Step 1: Establish a scheduling model for different power sources and energy storage devices in the microgrid;
步骤二：分别构建孤网长期稳定运行和孤网短期运行的目标函数；Step 2: Construct the objective functions of the long-term stable operation of the isolated network and the short-term operation of the isolated network respectively;
步骤三：对各模型进行线性化处理，使求解问题转化为混合整数线性规划问题；Step 3: Linearize each model to convert the solution problem into a mixed integer linear programming problem;
步骤四：应用前推回代法，计算微电网的三相潮流，得到线路损耗的系数；Step 4: Calculate the three-phase power flow of the microgrid by applying the forward push-back method, and obtain the coefficient of line loss;
步骤五：建立按时间分层的微电网调度优化模型；Step 5: Establish a time-layered microgrid scheduling optimization model;
步骤六：根据目标函数及对应的约束条件进行优化求解。Step 6: Optimizing and solving according to the objective function and corresponding constraints.
下面是对上述各步骤的详细介绍。The following is a detailed description of each of the above steps.
步骤一：建立微电网不同电源、储能装置的调度模型。微电网的电源主要包括传统的完全可控机组(包括微型燃气轮机、燃料电池、柴油机等常规机组(或称为可控电源))、新能源机组(包括风电、光伏等)，以及储能装置(包括各类蓄电池、超级电容等)。对上述不同类型的电源分别建模，由此可以将微电网大部分电源纳入调度优化模型中。每组中相应设备的数量可能是单数也可能是多数，目标函数中以相应设备组为单位，但名称上不限制是否有“组”字样。它们各自的模型如下：Step 1: Establish a scheduling model for different power sources and energy storage devices in the microgrid. The power supply of the microgrid mainly includes traditional fully controllable units (including micro gas turbines, fuel cells, diesel engines and other conventional units (or called controllable power sources)), new energy units (including wind power, photovoltaics, etc.), and energy storage devices ( Including all kinds of batteries, super capacitors, etc.). The above-mentioned different types of power sources are modeled separately, so that most of the power sources of the microgrid can be included in the scheduling optimization model. The number of corresponding equipment in each group may be singular or multiple, and the target function is based on the corresponding equipment group, but there is no restriction on whether there is a word "group" in the name. Their respective models are as follows:
1)风机模型：1) Fan model:
风机的出力与风速的关系可以由式(1)近似表示：The relationship between fan output and wind speed can be approximated by formula (1):
其中，v为风速，vci、vco、vr分别为切入风速、切出风速和额定风速；Prated_WT为风机额定功率。Among them, v is the wind speed, vci , vco , and vr are the cut-in wind speed, cut-out wind speed and rated wind speed respectively; Prated_WT is the rated power of the fan.
2)光伏发电模型：2) Photovoltaic power generation model:
在标准测试条件下，光伏发电的出力可以由式(2)表示：Under standard test conditions, the output of photovoltaic power generation can be expressed by formula (2):
其中，GSTC、TSTC分别为标准测试条件中的太阳辐射照度和电池温度；Gc为工作点的太阳辐射照度；Tc为工作点的电池温度；PSTC为标准测试条件下的光伏额定发电功率；kTC为功率温度系数。Among them, GSTC and TSTC are the solar irradiance and battery temperature in the standard test conditions; Gc is the solar irradiance at the working point; Tc is the battery temperature at the working point; PSTC is the photovoltaic rating under the standard test conditions Generating power; kTC is power temperature coefficient.
3)微型燃气轮机模型：3) Micro gas turbine model:
微型燃气轮机的燃料成本函数如下：The fuel cost function of a micro gas turbine is as follows:
其中Cnl是天然气价格；Ccal为天然气的热值；PMT(t)是微型燃气轮机t时刻的输出功率；ηMT(t)是微型燃气轮机t时刻的发电效率。Where Cnl is the price of natural gas; Ccal is the calorific value of natural gas; PMT (t) is the output power of the micro gas turbine at time t; ηMT (t) is the power generation efficiency of the micro gas turbine at time t.
4)燃料电池模型：4) Fuel cell model:
燃料电池发电的燃料成本函数与微型燃气轮机的相似，成本函数如下：The fuel cost function of fuel cell power generation is similar to that of micro gas turbine, and the cost function is as follows:
其中，Cnl是天然气价格；Ccal为天然气的热值；PFC(t)是t时刻燃料电池的输出功率；ηFC(t)是时刻t的燃料电池发电效率。Among them, Cnl is the price of natural gas; Ccal is the calorific value of natural gas; PFC (t) is the output power of the fuel cell at time t; ηFC (t) is the power generation efficiency of the fuel cell at time t.
5)柴油机模型：5) Diesel engine model:
柴油发电机的燃料成本与传统火力发电类似，可以表示为柴油发电机的输出功率的二次多项式，其燃料成本可以用下式表示：The fuel cost of diesel generators is similar to that of traditional thermal power generation, which can be expressed as a quadratic polynomial of the output power of diesel generators, and its fuel cost can be expressed by the following formula:
其中，a、b、c分别为费用函数的系数，可以由厂家提供或者拟合得到；PDG(t)是柴油发电机t时刻的输出功率。Among them, a, b, c are the coefficients of the cost function, which can be provided by the manufacturer or obtained by fitting;PDG (t) is the output power of the diesel generator at time t.
6)蓄电池模型：6) Battery model:
(1)蓄电池在t时刻的剩余能量跟其前一个时刻即t-Δt时刻的剩余能量以及其从t-Δt时刻到t时刻的充放电量有关，在充放电的过程中，蓄电池的剩余能量计算公式为：(1) The remaining energy of the battery at time t is related to the remaining energy at the previous time, that is, the time t-Δt, and the amount of charge and discharge from time t-Δt to time t. In the process of charging and discharging, the remaining energy of the battery The calculation formula is:
E(t)＝E(t-1)(1-σ)-Pdh(t)Δt/ηdh+Pch(t)Δtηch (6)E(t)＝E(t-1)(1-σ)-Pdh (t)Δt/ηdh +Pch (t)Δtηch (6)
其中E(t)、E(t-1)分别为t、t-1时刻末的蓄电池剩余能量(kWh)；Pch(t)、Pdh(t)分别为蓄电池的充电功率和放电功率；ηch、ηdh分别是蓄电池的充电效率和放电效率；σ是蓄电池的自放电率。Where E(t), E(t-1) are the remaining energy (kWh) of the battery at the end of time t and t-1, respectively; Pch (t), Pdh (t) are the charging power and discharging power of the battery, respectively; ηch and ηdh are the charge efficiency and discharge efficiency of the battery respectively; σ is the self-discharge rate of the battery.
(2)蓄电池的剩余电量有约束，超过约束就可能会导致的电池的损坏，蓄电池的剩余电量约束如下：(2) There are restrictions on the remaining power of the battery. Exceeding the limit may cause damage to the battery. The remaining power of the battery is restricted as follows:
其中SOC、分别是蓄电池最小和最大容许剩余电量，SOC(t)是蓄电池t时刻的剩余电量。Among them,SOC , are the minimum and maximum allowable remaining power of the battery, respectively, and SOC(t) is the remaining power of the battery at time t.
(3)蓄电池的充放电功率同样有限制，超过限制也会造成损坏，充放电功率的上下限约束表达如下：(3) The charging and discharging power of the battery is also limited, and damage will be caused if it exceeds the limit. The upper and lower limits of the charging and discharging power are expressed as follows:
其中，分别表示蓄电池的最大充电功率和最大放电功率。in, Respectively represent the maximum charging power and maximum discharging power of the battery.
(4)蓄电池的寿命模型(4) Battery life model
蓄电池的寿命一般是由厂家给出的不同放电深度下的循环次数来表示，蓄电池总的循环次数函数如下：The life of the battery is generally expressed by the number of cycles given by the manufacturer under different discharge depths. The function of the total number of cycles of the battery is as follows:
其中，DOD是蓄电池的放电深度；a1-a5是由生产厂家提供的参数。Among them, DOD is the discharge depth of the battery; a1-a5 is the parameter provided by the manufacturer.
但是，蓄电池通常会频繁的充放电，不能保证每次的放电深度都是一样的，给计算循环次数带来难度。However, batteries are usually charged and discharged frequently, and the depth of discharge cannot be guaranteed to be the same every time, which makes it difficult to calculate the number of cycles.
吞吐量法是针对蓄电池运行中循环次数难以统计而提出的简单易行的寿命预测方法，其寿命的计算公式如下：Throughput method is a simple and easy life prediction method proposed for the difficulty of counting the number of cycles in battery operation. The calculation formula of its life is as follows:
其中，Ethroughput为蓄电池寿命内总的能量吞吐量；Erated为蓄电池的额定容量；n为对蓄电池的不同放电深度测试的个数。Among them, Ethroughput is the total energy throughput within the battery life; Erated is the rated capacity of the battery; n is the number of different discharge depth tests on the battery.
其中，不同放电深度的循环次数和对应的总能量吞吐量由厂家给出。Among them, the number of cycles of different discharge depths and the corresponding total energy throughput are given by the manufacturer.
假设蓄电池总的能量吞吐量为一定值，则寿命损耗系数公式如下：Assuming that the total energy throughput of the battery is a certain value, the formula for the life loss coefficient is as follows:
其中，Eloss为计算周期内蓄电池的能量吞吐量，如式(12)所示：Among them, Eloss is the energy throughput of the battery in the calculation period, as shown in formula (12):
其中，P(t)是蓄电池t时刻的充电或放电功率；Among them, P(t) is the charging or discharging power of the battery at time t;
吞吐量法可以粗略的计算出蓄电池的使用寿命。The throughput method can roughly calculate the service life of the battery.
步骤二：分别构建孤网长期稳定运行和孤网短期运行时的目标函数和相应的约束条件。Step 2: Construct the objective functions and corresponding constraints for the long-term stable operation of the isolated network and the short-term operation of the isolated network.
模式一：孤网长期稳定运行：Mode 1: Long-term stable operation of the isolated network:
以费用最少为目标，综合考虑系统和各微源的约束条件，建立在孤网长期稳定运行条件下的微电网日前经济调度模型。Aiming at the least cost, comprehensively considering the constraints of the system and each micro-source, a day-ahead economic scheduling model of the microgrid under the condition of long-term stable operation of the isolated grid is established.
目标函数：Objective function:
其中，C为微电网系统的运行成本，Pi(t)为可控电源i的t时刻的输出功率，Ci(Pi(t))为可控电源燃料成本函数，包括微型燃气轮机、柴油机和燃料电池；Si(t)为可控电源启动成本；Uopen,i(t)：t时刻可控电源启动的0、1决策变量，0表示非启动(保持原启停状态或者停机)，1表示启动；COM,i(Pi(t))、CDP,i(Pi(t))、Ce(Pi(t))分别表示可控机组的维护费用、折旧成本和环境成本；Pk(t)为蓄电池k的t时刻的输出功率，COM,k(Pk(t))、CDP,k(Pk(t))分别表示储能装置的维护费用、折旧成本；CPUNISH,k为蓄电池周期不平衡罚函数；Pl(t)为t时刻负荷节点l的负荷，Cl(Pl(t))为负荷节点l的切负荷费用函数，Ul(t)为负荷节点l的切负荷0、1决策变量，0表示不切负荷，1表示切负荷；为一级负荷l1的切负荷0、1决策变量，0表示不切负荷，1表示切负荷，为一级负荷节点l1的罚函数的系数，相应惩罚项迫使相同的一级负荷前一个时段切负荷成本比后一个时段高，此惩罚项系数的调节可改变模型对时段前后的重视程度，若设为0，则时段前后没有区别；al为负荷节点l的罚函数的系数，针对所有负荷，相应的惩罚项迫使相同等级的负荷前一个时段切负荷成本比后一个时段高，此惩罚项系数的调节可改变模型对时段前后的重视程度，若设为0，则时段前后没有区别；I为可控电源的数量；K为蓄电池的数量；T、T’为一个调度周期内的优化时段数，L为可中断负荷的负荷节点个数，L1表示一级负荷节点的个数。Among them, C is the operating cost of the microgrid system, Pi (t) is the output power of the controllable power source i at time t, and Ci (Pi (t)) is the fuel cost function of the controllable power source, including micro gas turbines, diesel engines and fuel cell; Si (t) is the start-up cost of the controllable power supply; Uopen,i (t): 0, 1 decision variable for the start-up of the controllable power supply at time t, 0 means non-start (keep the original start-stop state or stop) , 1 means start; COM,i (Pi (t)), CDP,i (Pi (t)), Ce (Pi (t)) represent the maintenance cost, depreciation cost and Environmental cost; Pk (t) is the output power of battery k at time t, COM,k (Pk (t)), CDP,k (Pk (t)) represent the maintenance cost of the energy storage device, Depreciation cost; CPUNISH,k is the battery cycle unbalance penalty function; Pl (t) is the load of load node l at time t, Cl (Pl (t)) is the load shedding cost function of load node l, Ul (t) is the load shedding 0, 1 decision variable of load node l, 0 means no load shedding, 1 means load shedding; It is the decision variable of load shedding 0 and 1 for primary load l1 , 0 means no load shedding, 1 means load shedding, is the coefficient of the penalty function of the primary load node l1. The corresponding penalty item forces the same primary load to have a higher load shedding cost in the previous period than in the subsequent period. The adjustment of the penalty item coefficient can change the model’s emphasis on the period before and after the period. If it is set to 0, there is no difference before and after the time period; al is the coefficient of the penalty function of the load node l. For all loads, the corresponding penalty item forces the load of the same level to have a higher load shedding cost in the previous period than in the latter period. The adjustment of the item coefficient can change the model’s emphasis on the period before and after the period. If it is set to 0, there will be no difference between the period before and after the period; I is the number of controllable power sources; K is the number of batteries; T and T’ are the optimization within a scheduling cycle The number of time periods, L is the number of load nodes that can interrupt the load, and L1 is the number of first-level load nodes.
式中的可控机组维护费用、折旧成本和环境成本的函数见式(14)-(16)：The functions of the controllable unit maintenance cost, depreciation cost and environmental cost in the formula are shown in formula (14)-(16):
COM,i(Pi(t))＝KOM,i·Pi(t)·Δt (14)COM,i (Pi (t))=KOM,i Pi (t) Δt (14)
其中，Caz,i为第i个可控电源的单位容量安装成本的现值(本文费用都折算到现值)；ki为第i个微源的容量因数；ni为第i个微源的使用年限；KOM,i为第i个微源的单位电量运行维护成本系数；Vj为第j项污染物的单位排放费用；Qij为第i个微源单位电量的第j项污染物排放量；m为污染物的种类；Δt为一个时段的时长。Among them, Caz,i is the present value of the unit capacity installation cost of the i-th controllable power source (the cost in this paper is converted to the present value); ki is the capacity factor of the i-th micro-source; ni is the i-th micro-source The service life of the source; KOM,i is the operation and maintenance cost coefficient of the unit electricity of the i-th micro-source; Vj is the unit discharge cost of the j-th pollutant; Qij is the j-th item of the i-th micro-source unit electricity Pollutant emission; m is the type of pollutant; Δt is the duration of a period.
式中的储能装置的维护费用、折旧成本函数见式(17)、(18)：The maintenance cost and depreciation cost functions of the energy storage device in the formula are shown in formulas (17) and (18):
COM,k(Pk(t))＝KOM,k·|Pk(t)|·Δt (17)COM,k (Pk (t))=KOM,k ·|Pk (t)|·Δt (17)
CDP,k(Pk(t))＝(Erated,k·CE,k+Prated,k·CP,k)gLloss,k (18)CDP,k (Pk (t))＝(Erated,k CE,k +Prated,k CP,k )gLloss,k (18)
其中，KOM,k为第k个蓄电池组的单位电量运行维护成本系数；Erated,k、Prated,k分别为第k个蓄电池组的额定容量和额定功率，CE,k、CP,k分别为蓄电池组k的单位容量和单位功率安装成本的现值；Lloss,k为蓄电池组k的寿命损耗系数。Among them, KOM,k is the unit power operation and maintenance cost coefficient of the kth battery pack; Erated,k , Prated,k are the rated capacity and rated power of the kth battery pack respectively, CE,k , CP ,k are the present value of unit capacity and unit power installation cost of battery pack k respectively; Lloss,k is the life loss coefficient of battery pack k.
蓄电池的SOC周期不平衡罚函数如下：The SOC period imbalance penalty function of the battery is as follows:
式中，bk为罚函数的系数，其值根据需要比一级负荷的惩罚系数Cl_cut稍小。SOCk(l)、SOCk(T)分别表示蓄电池k的SOC初始值和一个调度周期后的值，ΔSOCk为程序设置的一个调度周期后SOCk的增加量，如果蓄电池SOCk状态良好，可以将ΔSOCk设为0。In the formula, bk is the coefficient of the penalty function, and its value is slightly smaller than the penalty coefficient Cl_cut of the primary load as required. SOCk (l) and SOCk (T) represent the initial value of SOC of battery k and the value after one scheduling period respectively, ΔSOCk is the increment of SOCk after one scheduling period set by the program, if the state of battery SOCk is good, ΔSOCk may be set to zero.
切负荷费用函数如下：The load shedding cost function is as follows:
Cl(Pl(t))＝Cl_cut·Pl(t) (20)Cl (Pl (t)) = Cl_cut Pl (t) (20)
式中Pl(t)表示t时间负荷节点l的负荷；Cl_cut表示切负荷节点l时单位电能的经济损失惩罚系数。In the formula, Pl (t) represents the load of load node l at time t; Cl_cut represents the economic loss penalty coefficient of unit electric energy at load shedding node l.
约束条件Restrictions
1)功率平衡和备用约束1) Power balance and backup constraints
(1)功率平衡约束：(1) Power balance constraints:
其中，Pi(t)、Pk(t)、Pw(t)、Pp(t)分别表示可控机组出力、蓄电池组出力、风电出力和太阳能发电出力；Pd(t)为微电网的负荷需求；M为微电网支路总数；Ploss(t)为微电网的线损；Pm(t)为流过线路m的功率；αm为线路m的线损系数，由潮流计算估算得到。Among them, Pi (t), Pk (t), Pw (t), and Pp (t) respectively represent the output of controllable units, battery packs, wind power, and solar power; Pd (t) is micro The load demand of the power grid; M is the total number of microgrid branches; Ploss (t) is the line loss of the microgrid; Pm (t) is the power flowing through the line m; αm is the line loss coefficient of the line m, determined by the power flow Calculated to estimate.
(2)备用约束：(2) Alternate constraints:
其中，Ui(t)为可控电源i的运行状态，0表示停运，1表示运行；为可控电源i的输出功率上限，Δup,i为可控电源i的最大向上爬坡率，Ek(t)分别为蓄电池剩余能量的上限、蓄电池k在t时刻的剩余能量，ηdh,k为蓄电池k的放电效率，为蓄电池k的最大放电功率，Rs(t)为微电网的旋转备用需求。此约束中，不等式左边为系统旋转备用的值，分别为可控机组、蓄电池的旋转备用值，Rs(t)为系统所需旋转备用。Among them, Ui (t) is the running state of the controllable power supply i, 0 means outage, 1 means running; is the upper limit of the output power of the controllable power supply i, Δup,i is the maximum upward ramp rate of the controllable power supply i, Ek (t) is the upper limit of the remaining energy of the battery and the remaining energy of the battery k at time t, respectively,ηdh,k is the discharge efficiency of the battery k, is the maximum discharge power of battery k, and Rs (t) is the spinning reserve requirement of the microgrid. In this constraint, the left side of the inequality is the value of the spinning reserve of the system, which are the spinning reserve values of the controllable unit and the battery respectively, and Rs (t) is the spinning reserve required by the system.
2)与可控机组相关的约束2) Constraints related to controllable units
(3)可控电源的输出功率约束：(3) The output power constraint of the controllable power supply:
其中，Pi、分别表示可控机组i的出力下限和出力上限。Among them,Pi , Respectively represent the output lower limit and output upper limit of the controllable unit i.
(4)运行状态变量与启停变量约束：(4) Operating state variables and start-stop variable constraints:
Ui(t)-Ui(t-1)＝Uopen,i(t-1)-Uclose,i(t-1) (25)Ui (t)-Ui (t-1)＝Uopen,i (t-1)-Uclose,i (t-1) (25)
Uopen,i(t)+Uclose,i(t)≤1 (26)Uopen,i (t)+Uclose,i (t)≤1 (26)
其中，Ui(t)表示可控机组i的运行状态，0表示停运、1表示运行；Uopen,i(t)为可控电源i的t时刻0、1开机决策变量，0表示非启动(保持原启停状态或者停机)，1表示启动。Uclose,i(t)为可控电源i的t时刻0、1关机决策变量，0表示非停机(保持原启停状态或者开机)，1表示停机操作。Among them, Ui (t) represents the operating state of the controllable unit i, 0 means outage, 1 means running; Uopen,i (t) is the decision variable of the controllable power supply i at time 0 and 1, and 0 means not Start (maintain the original start-stop state or stop), 1 means start. Uclose,i (t) is the shutdown decision variable of the controllable power supply i at time t 0, 1, 0 means non-shutdown (maintain the original start-stop state or start-up), 1 means shutdown operation.
(5)最小停运时间、运行时间约束：(5) Minimum downtime and running time constraints:
其中，MOTi为可控机组i的最小开机持续时间；MDTi为可控机组i的最小关机持续时间；T为调度时间周期。Among them, MOTi is the minimum startup duration of controllable unit i; MDTi is the minimum shutdown duration of controllable unit i; T is the scheduling time period.
(6)爬坡率约束：(6) Gradient rate constraint:
-Δdown,i≤Pi(t)-Pi(t-1)≤Δup,i (29)-Δdown,i ≤Pi (t)-Pi (t-1)≤Δup,i (29)
式中Δdown,i、Δup,i分别表示可控机组i的最大向下和向上爬坡率。whereΔdown,i and Δup,i represent the maximum downward and upward ramp rates of controllable unit i, respectively.
3)与蓄电池组相关的约束：3) Constraints related to the battery pack:
(7)荷电状态约束：(7) State of charge constraints:
其中，SOCk(t)为蓄电池k的t时刻的剩余电量，SOCk、分别表示蓄电池组k的剩余电量的最小和最大值。Among them, SOCk (t) is the remaining power of battery k at time t,SOCk , Respectively represent the minimum and maximum value of the remaining power of battery pack k.
(8)满足功率约束：(8) Satisfy the power constraint:
其中分别表示蓄电池组k的最大充电功率和最大放电功率。in Respectively represent the maximum charging power and maximum discharging power of battery pack k.
(9)周期储能平衡约束：(9) Periodic energy storage balance constraints:
SOCk(1)+ΔSOCk＝SOCk(T)+ΔSOCpunish,k (32)SOCk (1)+ΔSOCk ＝SOCk (T)+ΔSOCpunish,k (32)
SOCk(l)、SOCk(T)分别为蓄电池k的SOC初始值和一个调度周期后的值。ΔSOCk为程序设置的一个调度周期后SOCk的增加量，如果蓄电池SOCk状态良好，可以将ΔSOCk设为0，ΔSOCpunish,k表示一个调度周期后蓄电池计划外的SOCk的减少量。此约束与模型中蓄电池的SOC周期不平衡罚函数结合，可以确保蓄电池在保证一级负荷供电的情况下，调度周期的始末时刻蓄电池组k的剩余电量尽量满足设定值的。SOCk (l) and SOCk (T) are the initial value of SOC of battery k and the value after one scheduling period, respectively. ΔSOCk is the increase of SOCk after a scheduling cycle set by the program. If the battery SOCk is in good condition, ΔSOCk can be set to 0, and ΔSOCpunish,k represents the unplanned reduction of SOCk of the battery after a scheduling cycle. This constraint, combined with the SOC period imbalance penalty function of the battery in the model, can ensure that the remaining power of the battery pack k at the beginning and end of the dispatch cycle can meet the set value as much as possible while ensuring the power supply of the primary load.
(10)同一蓄电池组的充放电互斥约束：(10) The charge and discharge mutual exclusion constraints of the same battery pack:
Udh,k(t)+Uch,k(t)≤1 (33)Udh,k (t)+Uch,k (t)≤1 (33)
式中Udh,k(t)为蓄电池k的放电状态0、1决策变量，0表示非放电状态(充电或者不充不放)，1表示放电状态，Uch,k(t)为蓄电池k的充电状态0、1决策变量，0表示非充电状态(放电或者不充不放)，1表示充电状态。In the formula, Udh,k (t) is the decision variable of the discharge state 0 and 1 of the battery k, 0 means the non-discharge state (charging or not charging), 1 means the discharge state, and Uch,k (t) is the battery k The charging state 0, 1 decision variable, 0 means non-charging state (discharging or not charging), 1 means charging state.
(11)不同蓄电池组的充放电互斥约束：(11) Mutual exclusion constraints of charging and discharging of different battery packs:
Uch,k(t)≤K×(1-edh,k(t)) (35)Uch,k (t)≤K×(1-edh,k (t)) (35)
Udh,k(t)≤K×(1-ech,k(t)) (37)Udh,k (t)≤K×(1-ech,k (t)) (37)
其中edh,k(t)、ech,k(t)为辅助0、1变量；edh,k(t)为0表示无法放电，1表示无法充电，ech,k(t)为0表示无法充电，1表示无法放电，K为蓄电池组的个数。此约束表示当一个蓄电池放电时其它蓄电池组不能充电，反之亦然。Among them, edh,k (t), ech,k (t) are auxiliary 0 and 1 variables; edh,k (t) is 0, which means it cannot be discharged, and 1 means that it cannot be charged, and ech,k (t) is 0 Indicates that it cannot be charged, 1 indicates that it cannot be discharged, and K is the number of battery packs. This constraint means that when one battery is discharged, the other battery pack cannot be charged, and vice versa.
模式二：孤网短期运行Mode 2: Short-term operation of the isolated network
以费用最少为目标，综合考虑系统和各微源的约束条件，建立在孤网短期运行条件下的微电网日前经济调度模型。与孤网长期稳定运行不同的是，孤网短期运行尽量考虑孤网短时间内运行的最优结果。Aiming at the least cost, comprehensively considering the constraints of the system and each micro-source, a day-ahead economic scheduling model of the microgrid under the condition of short-term operation of the isolated grid is established. Different from the long-term stable operation of the isolated network, the short-term operation of the isolated network tries to consider the optimal result of the short-term operation of the isolated network.
目标函数objective function
孤网运行下微电网经济调度模型的目标函数如下：The objective function of the microgrid economic dispatch model under isolated grid operation is as follows:
与孤网长期稳定运行方式不同的是，孤网短期运行时目标函数中没有蓄电池没有SOC周期不平衡惩罚项，而最后一项的范围扩大到所有等级的负荷。目标函数中T’表示孤网短期运行时的优化周期，一般比较小(1h)，这样可以保证只考虑短期优化，连续多个优化周期可以得到更长时间内的优化结果；L表示可中断负荷的负荷节点个数，Ul(t)表示切负荷的0、1决策变量，0表示不切负荷，1表示切负荷；Cl(Pl(t))表示负荷节点l的切负荷的费用函数；al表示罚函数的系数，针对所有负荷，相应的惩罚项迫使相同等级的负荷前一个时段切负荷成本比后一个时段高，此惩罚项系数的调节可改变模型对时段前后的重视程度，若设为0，则时段前后没有区别。Different from the long-term stable operation mode of the isolated grid, the short-term operation of the isolated grid does not include the storage battery and the SOC cycle unbalance penalty item in the objective function, and the scope of the last item is extended to all levels of load. In the objective function, T' represents the optimization cycle of the isolated grid in short-term operation, which is generally relatively small (1h), so that only short-term optimization can be considered, and multiple consecutive optimization cycles can obtain optimization results in a longer period of time; L represents the interruptible load Ul (t) represents the decision variable of 0 and 1 for load shedding, 0 means no load shedding, 1 means load shedding; Cl (Pl (t)) means the cost of load shedding for load node l function; al represents the coefficient of the penalty function. For all loads, the corresponding penalty item forces the same level of load to force the load shedding cost in the previous period to be higher than that in the subsequent period. The adjustment of the coefficient of this penalty item can change the model’s emphasis on the period before and after the period , if set to 0, there is no difference before and after the period.
切负荷费用函数如下：The load shedding cost function is as follows:
Cl(Pl(t))＝Cl_cut·Pl(t) (39)Cl (Pl (t)) = Cl_cut Pl (t) (39)
式中Pl(t)表示t时间负荷节点l的负荷；Cl_cut表示切负荷节点l时单位电能的经济损失惩罚系数。In the formula, Pl (t) represents the load of load node l at time t; Cl_cut represents the economic loss penalty coefficient of unit electric energy at load shedding node l.
约束条件Restrictions
孤网短期运行方式的约束条件大部分和孤网长期稳定运行方式相同，只是孤网短期运行方式下没有周期储能平衡约束。在此不再赘述。Most of the constraints of the short-term operation mode of the isolated grid are the same as the long-term stable operation mode of the isolated grid, except that there is no periodic energy storage balance constraint in the short-term operation mode of the isolated grid. I won't repeat them here.
步骤三：对模型进行线性化处理，使求解问题转化为混合整数线性规划问题。Step 3: Linearize the model to convert the solution problem into a mixed integer linear programming problem.
混合整数线性规划算法的计算时间相对其他算法较少，可以解决相对大型的问题。但是它需要将非线性的问题转化为线性问题来求解，在微电网的模型中有多个非线性的约束需要进行线性化处理。模型中的大部分非线性的约束，已经在约束条件里进行了分段线性化的处理，但是可控机组出力还需进行额外的处理。The calculation time of the mixed integer linear programming algorithm is less than other algorithms, and it can solve relatively large problems. However, it needs to transform the nonlinear problem into a linear problem to solve, and there are multiple nonlinear constraints in the microgrid model that need to be linearized. Most of the nonlinear constraints in the model have been dealt with piecewise linearization in the constraint conditions, but the output of the controllable unit still needs to be dealt with additionally.
可控机组出力的处理：Handling of controllable unit output:
可控机组的燃料费用曲线是非线性的，要在混合整数线性规划中考虑燃料费用，需要将费用曲线进行分段线性化，如图1所示，可控机组的燃料费用函数C(P)是一个非线性函数，假设将费用曲线分为n段线性组合，各个分段的分点为P1≤P2≤L≤Pn≤Pn+1，引入一个新的变量zk，可以将出力P和燃料费用的分段线性函数C′(P)分别表示为式(40)和式(41)：The fuel cost curve of the controllable unit is non-linear. To consider the fuel cost in the mixed integer linear programming, the cost curve needs to be linearized piecewise. As shown in Figure 1, the fuel cost function C(P) of the controllable unit is A non-linear function, assuming that the cost curve is divided into n segments of linear combination, the sub-points of each segment are P1 ≤ P2 ≤ L ≤ Pn ≤ Pn+1 , a new variable zk is introduced, and the output can be The piecewise linear function C′(P) of P and fuel cost is expressed as formula (40) and formula (41):
再次引入一个新的0、1变量ys，其中zs与ys满足下列约束：Introduce a new 0, 1 variable ys again, where zs and ys satisfy the following constraints:
zs与ys为辅助变量，此式用于保证对成本函数的分段的合理与唯一。zs and ys are auxiliary variables, and this formula is used to ensure the rationality and uniqueness of the segmentation of the cost function.
步骤四：应用前推回代法，计算微电网的三相潮流，得到线路损耗的系数。Step 4: Calculate the three-phase power flow of the microgrid by applying the forward push-back method, and obtain the coefficient of line loss.
前推回代法的原理过程是：一般给定配电网的始端电压和末端负荷，以馈线为计算基本单位。开始时由末端向始端推算，设全网电压都为额定电压，根据负荷功率由末端向始端逐段推导，仅计算各元件中的功率损耗而不计算电压，求得各条支路上的电流和功率损耗，并据此获得始端功率，这是前推过程；再根据给定的始端电压和求得的始端功率由始端向末端逐段算电压降落，求得各节点电压，这是回代过程；如此重复上述过程，直至各个节点的电压偏差满足容许条件为止。具体流程图见图2。The principle process of the push-forward method is: generally given the initial terminal voltage and terminal load of the distribution network, the feeder is the basic unit of calculation. At the beginning, it is calculated from the end to the beginning, and the voltage of the whole network is assumed to be the rated voltage. According to the load power, it is derived step by step from the end to the beginning. Only the power loss in each component is calculated without calculating the voltage, and the current and voltage on each branch are obtained. Power loss, and obtain the initial power accordingly, this is the forward process; then according to the given initial voltage and the obtained initial power, the voltage drop is calculated step by step from the initial to the end, and the voltage of each node is obtained, which is the back-substitution process ; Repeat the above process until the voltage deviation of each node meets the allowable conditions. See Figure 2 for the specific flow chart.
将潮流计算得到的每个支路的平均线损结果，代入模型的线损系数αm。Substitute the average line loss result of each branch obtained from the power flow calculation into the line loss coefficient αm of the model.
步骤五：对模型按时间进行分层求解。Step 5: Solve the model hierarchically according to time.
由于同一时刻下微电网中每个单元的出力关系都是耦合在一起的，尽量不对其进行分解，而微电网中的电源响应速度都比较快，所以可以将模型按时间分为上下两层优化进行求解。上层优化求解模型的时间间隔较大(一般取1h)，下层优化求解模型为上层优化求解模型的每个时间间隔内的优化。分层优化对应的时间关系如图3所示。Since the output relationship of each unit in the microgrid is coupled together at the same time, try not to decompose it, and the response speed of the power supply in the microgrid is relatively fast, so the model can be divided into upper and lower layers according to time for optimization Solve. The time interval of the upper-level optimization solution model is relatively large (generally 1h), and the lower-level optimization solution model is the optimization within each time interval of the upper-level optimization solution model. The time relationship corresponding to hierarchical optimization is shown in Figure 3.
将原始问题按照一定时间间隔分为上层优化问题和下层优化问题。首先求解上层优化问题，上层优化的问题的求解结果作为下层优化问题的初始条件，而后求解下层优化问题，所有下层优化问题都求解完后得出最终的优化结果。具体的分层优化流程图见图4。Divide the original problem into an upper-level optimization problem and a lower-level optimization problem according to a certain time interval. First solve the upper-level optimization problem, the solution result of the upper-level optimization problem is used as the initial condition of the lower-level optimization problem, and then solve the lower-level optimization problem, and the final optimization result is obtained after all the lower-level optimization problems are solved. The specific hierarchical optimization flow chart is shown in Figure 4.
步骤六：选择合适的模式，进行计算。Step 6: Select the appropriate mode and perform the calculation.
所建模型有不同的微电网模式可供选择，根据微电网的特点和运行要求进行选择，修改模型中的参数可以达到各种不同的要求，也可以选择使用默认值进行计算。The built model has different microgrid models to choose from. The selection is made according to the characteristics and operation requirements of the microgrid. The parameters in the model can be modified to meet various requirements. You can also choose to use the default values for calculation.
图5所示为某项目的部分微电网结构示意图。微电网中有1个光伏发电电源，共计容量25kW；1组储能装置(蓄电池)，共计容量100kWh；3个负荷节点，假设负荷C6为一级负荷，负荷C7为二级负荷，负荷D5为三级负荷。Figure 5 shows a schematic diagram of a part of the microgrid structure of a project. There is one photovoltaic power source in the microgrid, with a total capacity of 25kW; one group of energy storage devices (battery batteries), with a total capacity of 100kWh; three load nodes, assuming that load C6 is the primary load, load C7 is the secondary load, and load D5 is tertiary load.
典型日负荷曲线：Typical daily load curve:
微电网中有3个负荷节点，其总负荷的日负荷曲线如图6所示。此典型日负荷曲线有三个负荷高峰，最大负荷为21.6kW，最小负荷为5.1kW。There are three load nodes in the microgrid, and the daily load curve of the total load is shown in Figure 6. This typical daily load curve has three load peaks, the maximum load is 21.6kW and the minimum load is 5.1kW.
光伏出力特性：Photovoltaic output characteristics:
图7给出了晴天光伏出力的典型场景。图中光伏容量为25kW。假设预测误差5％时的光伏预测出力如图8所示。Figure 7 shows a typical scenario of photovoltaic output in sunny days. The photovoltaic capacity in the figure is 25kW. Assuming that the forecast error is 5%, the photovoltaic forecast output is shown in Figure 8.
蓄电池基本参数：Battery basic parameters:
算例中的蓄电池基本参数如表1所示。The basic parameters of the battery in the example are shown in Table 1.
表1蓄电池基本参数Table 1 Basic parameters of battery
污染物排放征收费用：Pollutant discharge charges:
算例中的污染物排放征收费用见表2。See Table 2 for the pollutant discharge charges in the calculation example.
表2污染物排放征收费用Table 2 Pollutant Discharge Charges
切负荷费用：Load shedding fee:
不同等级负荷的切负荷费用见表3。See Table 3 for load shedding fees for different levels of load.
表3不同等级负荷的切负荷费用Table 3 Load Shedding Fees for Different Levels of Loads
测试结果：Test Results:
1)模式二：孤网长期稳定运行。测试内容如下：1) Mode 2: The isolated network operates stably for a long time. The test content is as follows:
场景：晴天；模式：孤网运行，孤网长期稳定运行；光伏预测误差5％。Scenario: sunny; mode: isolated grid operation, long-term stable operation of the isolated grid; photovoltaic forecast error 5%.
A)程序计算结果：A) Program calculation results:
模式二下，软件计算的理论切负荷结果如图9所示。In mode 2, the theoretical load shedding results calculated by the software are shown in Figure 9.
由图9可知，三级负荷D5全天24时段全部切掉，二级负荷C7只切晚间时段，一级负荷C6保证不断电。It can be seen from Figure 9 that the third-level load D5 is cut off at all 24 hours a day, the second-level load C7 is only cut off at night, and the first-level load C6 ensures uninterrupted power supply.
模式二下，软件计算的理论调度结果如图10所示。In mode 2, the theoretical scheduling results calculated by the software are shown in Figure 10.
由图10可知，切负荷切掉的功率、蓄电池充放电功率、光伏出力相加基本等于总负荷功率，即功率基本平衡。白天由光伏供电，多余的电量由蓄电池存储起来，晚间由蓄电池供电。It can be seen from Figure 10 that the sum of the power cut off by load shedding, battery charging and discharging power, and photovoltaic output is basically equal to the total load power, that is, the power is basically balanced. During the day, it is powered by photovoltaics, and the excess electricity is stored by the battery, and it is powered by the battery at night.
B)实际运行结果：B) Actual running results:
模式二下，实际运行的切负荷结果如图11所示。In mode 2, the load shedding results of the actual operation are shown in Figure 11.
图11中显示的实际切负荷情况与计算结果相同。The actual load shedding situation shown in Fig. 11 is the same as the calculated result.
模式二下，实际运行的调度结果如图12所示。In mode 2, the scheduling result of the actual operation is shown in Figure 12.
由图12可知，功率基本平衡，并且实际运行结果与程序计算结果的趋势相同：白天由光伏供电，多余的电量由蓄电池存储起来，晚间由蓄电池供电。It can be seen from Figure 12 that the power is basically balanced, and the trend of the actual operation results is the same as that of the program calculation results: during the day, it is powered by photovoltaics, the excess power is stored by the battery, and it is powered by the battery at night.
4)模式三：孤网短期运行。测试内容如下：4) Mode 3: short-term operation of the isolated network. The test content is as follows:
场景：晴天；模式：孤网运行，满足短期负荷需求；光伏预测误差5％。Scenario: sunny; mode: isolated grid operation to meet short-term load demand; photovoltaic forecast error 5%.
A)程序计算结果：A) Program calculation results:
模式三下，软件计算的理论切负荷结果如图13所示。In mode three, the theoretical load shedding results calculated by the software are shown in Figure 13.
由图13可知，从0点到3点，没有切负荷指令，保证所有负荷不断电。3点之后蓄电池电量放光，需要依次切除负荷。It can be seen from Figure 13 that from 0:00 to 3:00, there is no load shedding command to ensure that all loads are powered continuously. After 3 o'clock, the battery power is discharged, and the load needs to be cut off in sequence.
模式三下，软件计算的理论调度结果如图14所示。In mode three, the theoretical scheduling results calculated by the software are shown in Figure 14.
由图14可知，孤网运行时，满足短期负荷需求，一般用于并网模式下的短期停电，需要全力保证短期内负荷不断电，3小时后需要切负荷的时段，没有意义。It can be seen from Figure 14 that when the isolated grid is running, it meets the short-term load demand. It is generally used for short-term power outages in the grid-connected mode. It needs to make every effort to ensure that the load is uninterrupted in the short term. The period of load shedding after 3 hours is meaningless.
B)实际运行结果：B) Actual running results:
模式三下，实际运行的切负荷结果如图15所示。In mode three, the load shedding results of the actual operation are shown in Figure 15.
图15中的实际切负荷情况与程序计算结果相同。The actual load shedding situation in Figure 15 is the same as the program calculation result.
模式三下，实际运行的调度结果如图16所示。In mode three, the scheduling result of the actual operation is shown in Figure 16.
由图16可知，微电网实际运行结果与程序计算结果的趋势基本相同。It can be seen from Figure 16 that the trend of the actual operation results of the microgrid is basically the same as that of the program calculation results.
4)孤网原模式自动运行(即没有采用本发明的调度方法)时的测试内容如下：4) The test content when the original mode of the isolated network automatically runs (that is, does not adopt the scheduling method of the present invention) is as follows:
场景：晴天；模式：孤网运行，没有调度策略；Scenario: sunny; mode: isolated network operation, no scheduling strategy;
如图17所示，微电网孤网运行，只保留一级负荷，切掉其他负荷。As shown in Figure 17, the microgrid operates in an isolated network, only retaining the first-level load, and cutting off other loads.
实际调度结果如图18所示，晚间由蓄电池供电，白天由光伏供电，多余的电量由蓄电池储存，但是蓄电池在15点后充满电量，导致弃光现象出现。The actual dispatching results are shown in Figure 18. The battery is powered at night and the photovoltaic power is supplied during the day. The excess power is stored by the battery, but the battery is fully charged after 15:00, resulting in light abandonment.
测试结论：Test conclusion:
不同测试内容的实际运行成本对比如表4所示。The comparison of the actual running costs of different test contents is shown in Table 4.
表4微电网运行成本Table 4 Microgrid operating costs
1)对应同一种运行状态，在存在预测误差的情况下，程序计算结果和实际运行结果存在偏差，但是趋势基本相同。1) Corresponding to the same operating state, in the case of prediction errors, there is a deviation between the program calculation results and the actual operating results, but the trend is basically the same.
2)对比孤网长期稳定运行和孤网短期运行两种状态可知，孤网运行、负荷需求大时，微电网需要进行选择性的切负荷。在满足孤网长期稳定运行的策略下，微电网首先保证一级负荷的供电，切除部分时段的二级负荷，切除三级负荷，在一个调度周期后(一天)，蓄电池的剩余容量与调度前基本一致，使蓄电池有充足电量保证孤网长期稳定运行。在满足短期负荷需求的策略下，微电网保证短期内所有负荷不断电直到电源电量不足，此种情况适用于并网运行时的短时停电，进入孤网运行状态后会很快回到并网运行状态。在没有经济调度策略、自动运行时，微电网会只保证一级负荷的供电，将其它负荷切除。不同策略下的总费用依次为1821.6元、1795.05元和2450.91元。其中孤网长期稳定运行策略和满足短期负荷需求策略下的总费用比没有经济调度策略的总费用低很多，而满足短期负荷需求策略的总费用比孤网长期稳定运行策略低，这是因为满足短期负荷需求策略下将蓄电池的剩余容量放到很低的状态。2) Comparing the long-term stable operation of the isolated grid and the short-term operation of the isolated grid, it can be seen that when the isolated grid operates and the load demand is large, the microgrid needs to perform selective load shedding. Under the strategy of satisfying the long-term stable operation of the isolated grid, the microgrid first ensures the power supply of the primary load, removes the secondary load in some periods, and removes the tertiary load. After a dispatch cycle (one day), the remaining capacity of the battery is the same as that before dispatch. Basically the same, so that the battery has enough power to ensure the long-term stable operation of the isolated grid. Under the strategy of meeting the short-term load demand, the microgrid guarantees that all loads will be continuously powered in the short term until the power supply is insufficient. This situation is suitable for short-term power outages during grid-connected operation, and will soon return to grid-connected operation after entering the isolated grid operation state. Operating status. When there is no economic dispatch strategy and automatic operation, the microgrid will only guarantee the power supply of the primary load and cut off other loads. The total costs under different strategies are 1821.6 yuan, 1795.05 yuan and 2450.91 yuan. Among them, the total cost of the long-term stable operation strategy of the isolated network and the strategy of meeting the short-term load demand is much lower than the total cost of no economic dispatch strategy, and the total cost of the strategy of meeting the short-term load demand is lower than the long-term stable operation strategy of the isolated network. Under the short-term load demand strategy, the remaining capacity of the battery is placed in a very low state.
本发明所谓的几级负荷是指按照用户对供电连续性的要求，将供电负荷依次分为若干级，一级负荷是对供电连续性要求最高的负荷，从一级负荷往后对供电连续性依次降低。The so-called several-level load in the present invention refers to dividing the power supply load into several levels in turn according to the user's requirements for power supply continuity. The first-level load is the load with the highest requirement for power supply continuity. Decrease in turn.
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201510598528.8ACN105207207B (en) | 2015-09-18 | 2015-09-18 | Micro-grid system dispatching method under isolated network state based on energy management |
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201510598528.8ACN105207207B (en) | 2015-09-18 | 2015-09-18 | Micro-grid system dispatching method under isolated network state based on energy management |
| Publication Number | Publication Date |
|---|---|
| CN105207207A CN105207207A (en) | 2015-12-30 |
| CN105207207Btrue CN105207207B (en) | 2018-03-06 |
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201510598528.8AActiveCN105207207B (en) | 2015-09-18 | 2015-09-18 | Micro-grid system dispatching method under isolated network state based on energy management |
| Country | Link |
|---|---|
| CN (1) | CN105207207B (en) |
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106779180B (en)* | 2016-11-29 | 2020-04-21 | 国网陕西省电力公司电力科学研究院 | Optimization method for power system spinning reserve based on curve piecewise linearization |
| CN108537409B (en)* | 2018-03-15 | 2021-08-31 | 广东电网有限责任公司电网规划研究中心 | Industrial park power distribution network collaborative planning method considering multi-energy coupling characteristics |
| CN109214597B (en)* | 2018-10-25 | 2021-08-03 | 浙江工业大学 | An Optimal Planning Method for Microgrid Power and Operational Reserve Capacity |
| CN109255503B (en)* | 2018-11-09 | 2022-03-25 | 浙江大学 | Energy router replacement time optimal configuration method based on artificial firefly swarm optimization algorithm |
| CN110854904B (en)* | 2019-10-30 | 2021-05-11 | 深圳供电局有限公司 | Operation strategy planning method of energy storage device in power distribution system |
| TWI840223B (en)* | 2023-05-12 | 2024-04-21 | 遠東科技大學 | Charge and discharge control system and method thereof |
| CN118316015B (en)* | 2024-03-19 | 2024-12-03 | 深圳信息职业技术学院 | Microgrid energy dispatching method and related equipment for microgrid cluster |
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102289566A (en)* | 2011-07-08 | 2011-12-21 | 浙江大学 | Multiple-time-scale optimized energy dispatching method for micro power grid under independent operation mode |
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US8086355B1 (en)* | 2007-02-28 | 2011-12-27 | Global Embedded Technologies, Inc. | Method, a system, a computer-readable medium, and a power controlling apparatus for applying and distributing power |
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102289566A (en)* | 2011-07-08 | 2011-12-21 | 浙江大学 | Multiple-time-scale optimized energy dispatching method for micro power grid under independent operation mode |
| Publication number | Publication date |
|---|---|
| CN105207207A (en) | 2015-12-30 |
| Publication | Publication Date | Title |
|---|---|---|
| CN105207259B (en) | Micro-grid system dispatching method under based on energy management and net state | |
| CN105139147A (en) | Economic scheduling method for micro-grid system | |
| CN105207207B (en) | Micro-grid system dispatching method under isolated network state based on energy management | |
| Acevedo-Arenas et al. | MPC for optimal dispatch of an AC-linked hybrid PV/wind/biomass/H2 system incorporating demand response | |
| Logenthiran et al. | Short term generation scheduling of a microgrid | |
| CN104901338B (en) | A kind of island isolates microgrid energy control method | |
| CN103490410A (en) | Micro-grid planning and capacity allocation method based on multi-objective optimization | |
| CN106327006A (en) | Comprehensive benefit analysis-based micro-power-grid optimal configuration method | |
| CN108448632B (en) | Intraday rolling optimal scheduling method for AC/DC microgrid considering energy storage state-of-charge cycle | |
| CN105140941B (en) | Optimal scheduling method of microgrid considering variable depreciation cost of lithium battery and practical charging and discharging strategy | |
| CN108462198A (en) | A kind of microgrid Optimization Scheduling of providing multiple forms of energy to complement each other based on multi-agent technology | |
| Moghimi et al. | Battery energy storage cost and capacity optimization for university research center | |
| CN112270433A (en) | Micro-grid optimization method considering renewable energy uncertainty and user satisfaction | |
| Chen et al. | Energy storage sizing for dispatchability of wind farm | |
| CN110739684B (en) | Multi-time scale-based multi-AC/DC hybrid micro-grid optimal scheduling method | |
| CN114301081A (en) | Micro-grid optimization method considering energy storage life loss and demand response of storage battery | |
| Xie et al. | Optimal capacity configuration of hybrid energy storage for an isolated microgrid based on QPSO algorithm | |
| Wankhade et al. | Lithium-ion battery: Battery sizing with charge scheduling for load levelling, ramp rate control and peak shaving using metaheuristic algorithms | |
| CN115021406A (en) | Microgrid controller integrating economic model predictive control | |
| CN111224422A (en) | Reliability-based micro-grid distributed power supply configuration method and system | |
| Parthasarathy et al. | Optimal sizing of energy storage system and their impacts in hybrid microgrid environment | |
| CN108629445A (en) | The alternating current-direct current mixing microgrid Robust Scheduling method of meter and energy storage dynamic loss | |
| Pinceti et al. | Technical KPIs for microgrids | |
| Li et al. | Optimal operation of AC/DC hybrid microgrid under spot price mechanism | |
| CN108599148A (en) | The Robust Scheduling method of meter and alternating current-direct current microgrid reply Disaster Event elasticity capacity |
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |