INTRODUCTION
Due to the limited power supply range of a single microgrid and the frequent access and withdrawal of distributed generators during operation, the reliability and economy of a single micro-grid are not high [1]. In order to solve the shortcomings of a single microgrid, multiple microgrids interconnect to form a complementary microgrid system, which becomes an effective way to improve the utilization of the distributed energy. The microgrid cluster system composed of multiple microgrids can realize the control and management of supply and demand balance in microgrids by utilizing the self-governing characteristics of a single microgrid and the coordination and complementarity characteristics of multiple microgrids. Therefore, the study of coordinated control among multiple microgrids is an important issue in power development.
Cooperative control of microgrid group is an important means to realize a reliable and economic operation of the microgrid [2]. Fridans et al. [3] construct a multi-microgrid system consisting of multiple renewable energy internets and engines to maximize the use of renewable energy microgrid to construct an economic dispatching model for multi-microgrid, and use traditional particle swarm optimization algorithm to optimize the solution. Zhao et al. [4] establishes a two-tier distributed dispatching model, the upper layer realizes the dispatching at the group control level of microgrid, and the lower layer realizes the allocation of resources within the microgrid. The economic operation of the system is solved by parameter transfer between the two layers, and the alternating direction multiplier method is used to optimize the solution. Nikmehr et al. [5] determined that the power optimization model of microgrid group is established with the power fluctuation entropy as the objective, and the model is solved by the quantum particle swarm optimization algorithm. Finally, the effectiveness of the proposed control method is verified by simulation. The above literature is to optimize the control of the economic costs caused by the power of multiple microgrids. However, from the perspective of complementarity between microgrids with different load types, there are few studies on coordinated optimal dispatching of microgrid cluster systems. Therefore, the innovative work of this study is to optimize the dispatching of microgrid group composed of different load types. Based on the fixed defect of mutation operator in traditional genetic algorithm (GA), an adaptive strategy is introduced to improve the performance of the algorithm, and then an improved genetic algorithm (IGA)-based optimal dispatching method for microgrid is proposed. From the point of view of power supply economy and environmental protection, the optimal dispatching model of the microgrid cluster system is aimed at minimizing the operation and maintenance costs and environmental compensation costs. The IGA is used to solve the optimal dispatching model of the multi-microgrid system and find out the optimal dispatching scheme of the system.
MULTI-MICROGRID DISPATCHING SCENARIO AND COOPERATIVE OPERATION CONDITION ANALYSIS
Multi-microgrid structure
In this article, the multi-microgrid system shown in Figure 1 is taken as the research object. The multi-microgrid system consists of three different types of microgrids: industrial, residential and commercial. The power interaction between the microgrids can be achieved through the corresponding common connection points (PCC1, PCC2, PCC3) and the whole multi-microgrid system can interact with the main network through the group-level common connection points PCC0. Within each microgrid, photovoltaic, wind, storage battery and diesel engine are used as power sources to supply energy to the power loads within their respective ranges. In order to maximize the utilization efficiency of renewable energy, wind power and photovoltaic power supply are given priority in each microgrid area. When renewable energy generation cannot meet the demand of load power in the region, it will be discharged by batteries or purchasing power from nearby micro-grids or large grids at a high price. On the contrary, when renewable energy generation power in microgrid exceeds the demand of load for power, the remaining power will be stored in the energy storage equipment, or sold to nearby smart microgrids or large power grids at a low price.
[IMAGE OMITTED. SEE PDF]
Analysis of cooperative operation conditions of the microgrid
Due to the cooperation of multiple microgrids, certain conditions must be met: Each sub-microgrid participating in operation is a heterogeneous individual establishing an active contract that has a common purpose [4]. The microgrid cluster system, which is composed of industrial, residential and commercial microgrids, also meets these three conditions. Firstly, because the main loads in the three sub-microgrids are industrial load, residential load and commercial load, they have obvious differences in power demand. Because industrial loads are mainly produced by various industries, their electricity demand generally does not change much, and there is no obvious peak-valley difference. Resident load is mainly the resident users, whose load demand is closely related to residents’ living habits. Usually, 8:00–9:00 a.m. and 20:00–23:00 p.m. are peak periods of electricity consumption. Therefore, different electricity pricing policies have an important impact on their electricity consumption behaviour. Commercial loads are usually large shopping malls or shopping malls, and their peak load demand is generally around 11:00 in the daytime to 23:00 in the evening, and belong to rigid demand, the difference of power demand is relatively large. Therefore, they satisfy the first heterogeneous condition of multi-microgrid cooperative operation. The goal of each microgrid operation is to improve the reliability of power supply and the economy of microgrid operation. Therefore, they have the same goal to meet the second common objective condition of multi-microgrid cooperative operation. Similarly, in order to achieve good complementarity among the microgrids, the three parties need to sign a transaction contract to ensure that each participating subject actively participates in the operation of other microgrids in the cluster system under the condition of meeting their own needs. At the same time, on establishing a standard exchange price in this way, the third initiative condition of the multi-microgrid cooperative operation is satisfied. Therefore, the multi-microgrid system established in this chapter meets the conditions of cooperative operation.
SYSTEM MODEL
Optimization objective function
In order to realize the complementarity between microgrids in a microgrid cluster, it is necessary to meet the balance of supply and demand within a single microgrid as well as the balance of supply and demand among all microgrids in the whole microgrid cluster system. At the same time, it ensures the economic and reliable operation of the smart microgrid in the system [5]. Therefore, this article takes the operation and maintenance cost and environmental governance cost of the microgrid cluster system as the objective, and establishes the optimization objective function.
-
objective function f1 : f1 is the operation and maintenance cost of smart microgrid. Including fuel cost Cfuel,i, operation and maintenance cost Com,i, main network power exchange cost Cgrid,i, and between the microgrids for each smart microgrid Cex,i. Therefore, it can be described as follows [6]:
In the equation, M1 is the number of microgrids constituting the microgrid cluster system. CMG,i is the operation and maintenance cost of the i-th microgrid. Considering that controllable power sources such as diesel engines need fossil fuels to generate electricity, the fuel cost is [7]:
In the model, CDE,i is the fuel cost of diesel engine in microgrid i. a, b and c are the fuel cost coefficients of diesel engines, respectively. Because each microgrid uses the same fuel, their fuel cost coefficients are the same. PDE,i is the actual generating power of diesel engine in microgrid i.
After a period of operation, the internal distributed generation of each microgrid should be maintained regularly to ensure the normal and stable operation of the microgrid. Therefore, its operation and maintenance cost Com,i can be expressed as [8]:
When multiple microgrids complement each other in power and still cannot meet the power demand of the load in the microgrid group system, power needs to be purchased from the main network; However, when the power generated by distributed power sources in a plurality of microgrid systems is larger than the power demand of surplus loads in the microgrid system, in order to avoid power waste, surplus power needs Cgrid,i is as follows [9]:
Due to the different load types of multiple microgrid that make up the microgrid cluster system, the power supply-demand in each microgrid is also different. In order to realize the stability and economy of power supply for microgrid systems, multiple microgrid systems need to reach an active state so that they will contract in advance. Each sub-microgrid participating in the operation of the microgrid group must actively participate in the power interaction between the microgrid and at the same time establish a unified interactive electricity price. However, when multiple microgrids are operating in a complementary way, it is inevitable that they need to interact with the neighbouring microgrid according to their own operating conditions. Therefore, the cost of power exchange between multiple neighboring microgrid can be established as follows Cex,i [10]:
In the equation, Pex,ik and cik are the power and cost of the interaction between the i-th microgrid and the k-th microgrid, respectively.
-
Objective function f2 : f2 is the environmental compensation cost of the system and consists of the diesel engine and the main network [11]:
-
This chapter regards the operation and maintenance cost f1 and the environmental compensation cost f2 of the microgrid group as equally important objectives. It can be determined that the economic dispatching target f of the microgrid F is [12]:
Constraints
In order to ensure the stable and reliable operation of microgrid, microgrid should meet the following constraints.
-
Supply and demand balance constraints within a single microgrid: The sum of all power supplies in a single microgrid at any time (including distributed power supply, large grid or other microgrid) must be consistent with the power demand of the load in the microgrid.
-
The upper and lower limits of the output power of the distributed generation are constrained. In order to ensure the reliable operation of each distributed power source, its actual output power must be guaranteed to work within an appropriate range.
-
Energy storage state of charge constraints: Energy storage equipment should keep its state of charge in a reasonable interval during power exchange to maximize the service life of energy storage equipment.
-
Power exchange constraints with the main network: When the microgrid group system exchanges power with the large power grid, the power exchange conditions between the large power grid and the microgrid system must be satisfied.
-
Power exchange constraints between multiple microgrid: When carrying out power exchange among multiple microgrid, it must conform to the power range allowed by the connecting line, and then carry out reasonable power exchange.
MULTI-MICROGRID SCHEDULING STRATEGY BASED ON IMPROVED GENETIC ALGORITHM
GA is a heuristic algorithm based on the random reconnaissance method to find the optimal solution from the coded solution space. It is based on the evolution of natural selection law of biosphere [13]. Its core operation consists of selection, crossover and mutation [7]. The algorithm is simple and universal, has good global search capability and robustness, and is widely used in nonlinear, multi-objective and combinatorial optimization fields.
-
Selection: The operation of selecting the best individual and removing the inferior individual from the algorithm population is called selection. The goal of selection is to generate the best individuals from the algorithm population and pass them on to the next generation. The selection operation is evaluated based on some optimization criteria, and the roulette wheel selection method is currently the simplest and most commonly used selection method. In this method, the more excellent the individuals are in the population, the greater is the probability of being selected. Suppose the population size is N, and the fitness value of the individual i is fit(i) then the individual’s selection probability pi is [8,14]:
-
Cross: Cross operation is a key step in GA. The so-called crossover is an operation to generate new individuals by reorganizing the information of two father generations [15]. Through crossover operation, the population diversity of the GA is maintained, which greatly improves the global search capability of the GA.
Improved mutation operation
Mutation operation is an important step of GA. In the process of its operation, it is necessary to select the appropriate mutation operator. The size of the selected value determines the optimization range of the algorithm and further determines the optimization effect of the algorithm. Due to the fact that the mutation operator of the traditional GA generally selects a constant, it cannot be adjusted adaptively with the number of iterations to balance the global optimization ability and convergence speed ability of the algorithm. Therefore, in view of this situation, a strategy of adaptive adjustment of mutation operator is proposed. The mutation operator has a larger value in the early stage of the algorithm, which increases the search range of the algorithm and improves the global optimization capability of the algorithm. However, as the iteration progresses, the mutation operator gradually decreases making the algorithm search in a small range and improving the convergence ability of the algorithm. It is defined as follows [11]:
Algorithm flow
The execution steps of the algorithm and its specific flow chart are shown in Figure 2.
-
Initialization: In the process of microgrid optimization, parameters such as the size n of the algorithm population, the maximum iteration number c of algorithm search and the crossover probability Pc are initialized.
-
Calculate fitness: Calculating the fitness value of individual population. The fitness value corresponding to each individual is calculated and judging whether the iteration times reach a preset value, if so, outputting a result, otherwise, converting to Step (3) in continuing to execute downwards.
-
Select the best individual: According to the calculated individual fitness, individuals with high fitness value are selected.
-
Cross operation: According to the given crossover factor and crossover strategy, new individuals are generated.
-
Adaptive mutation factor is generated: According to Equation (15), an adaptive mutation factor is generated.
-
Variation operation: According to the mutation factors generated in step (5), the mutation operation is performed.
-
The new generation of individuals generated by cross operation and mutation operation returns to step (2) to enter the next cycle.
[IMAGE OMITTED. SEE PDF]
SIMULATION RESULTS AND ANALYSIS
Experimental scene and related data
The multi-microgrid system in this study is shown in Figure 1. It consists of three microgrid (residential microgrid, industrial microgrid and commercial microgrid). The experimental data in the microgrid are from references [16,17] and [2]. Their respective renewable energy and load output curves are shown in Figures 3, 4 and 5.
[IMAGE OMITTED. SEE PDF]
[IMAGE OMITTED. SEE PDF]
[IMAGE OMITTED. SEE PDF]
Because, in the system model established in this study, not only the operation and maintenance costs of distributed power sources within a single microgrid, but also the environmental compensation costs required for the output of distributed power sources are considered. At the same time, the whole multi-microgrid system not only needs power interaction with the main network, but also can realize power coordination between the microgrid. Therefore, the operating parameters of the distributed power sources in each microgrid and the interaction costs of the main network are shown in Table 1 and Figure 6, respectively. However, the environmental compensation coefficient of each pollution source and the maintenance coefficient of the distributed power source are the same as those in the document [2]. The interaction cost between the microgrid is 0.45 CNY/h.
TABLE 1 Distributed power supply parameters
Type | Pmin(Kw) | Pmax(Kw) |
BA1 | −30 | 30 |
DE1 | 0 | 50 |
BA2 | −30 | 30 |
DE2 | 0 | 50 |
BA3 | −30 | 30 |
DE3 | 0 | 50 |
Grid | −80 | 80 |
[IMAGE OMITTED. SEE PDF]
Experimental analysis of multi-microgrid dispatching
In order to verify the effectiveness of the multi-microgrid scheduling strategy proposed in this article, the parameters of the algorithm are set as follows: The number of iterations is 1000, the crossover operator is 0.2, and the population size is 288. Considering the peak-valley electricity price of the main grid and the power interaction price between the microgrid, the proposed IGA scheduling strategy is used to optimize the scheduling situation of distributed power sources in each microgrid in the microgrid group. The optimization results are shown in Figures 7–9.
[IMAGE OMITTED. SEE PDF]
[IMAGE OMITTED. SEE PDF]
[IMAGE OMITTED. SEE PDF]
As can be seen from Figure 7, the power supply from 1:00 to 6:00 is mainly provided by storage batteries and large power grids, and the diesel engine does not work because he wind power generation does not generate photovoltaic power at this time, and the output power of wind power generation cannot meet the power demand of industrial load in the microgrid. At the same time, as the electricity price of the large power grid is at a low ebb at this moment, it is lower than the transaction price of the complementary power between the micro power grids and the power generation cost of diesel engines. Therefore, at the moment only electricity is purchased from the main network. At about 7:00–9:00 h, the microgrid interacts with other microgrids in power. At about 10:00–17:00 hours, with the photovoltaic power generation work, the wind and light output in the microgrid is greater than the load demand, and one part of the surplus electric energy is stored in the storage battery and the other part is sold to the main grid. From 18:00 to 23:00 hours, as the sun goes down, the output of renewable energy is once again less than the demand for industrial power in the microgrid. At this moment, the storage battery is discharged, the diesel engine is started again to supplement power, and electricity is purchased from the grid. Figure 8 shows that between 1:00 and 6:00 h, the power output is mainly carried out by the large power grid and the storage battery, because the electricity price of the main power grid is relatively low at this time. Around 7:00–9:00 h, a small amount of power shortage can be obtained from other microgrid, thus saving the power purchased from the large grid. At about 10:00–18:00 h, with the photovoltaic power generation work, when the wind and light power generation power in the microgrid is greater than the load power supply demand, the remaining power will be sold to the large grid. From 19:00 to 24:00 h, the output of renewable energy in the microgrid is less than the demand of the residents. At this moment, the diesel engine is started again to supplement the power, and at the same time, electricity is purchased from the large grid. Since commercial activities are mainly concentrated from 11:00 to 23:00 h during the day, the output of renewable energy at 1:00–11:00 h in the microgrid can meet the load demand, and there is still some surplus electricity. Some electricity is stored in storage batteries as spare parts and some are sold to large power grids at low prices. At about 7:00–9:00 h, some of the electricity will be added to other microgrids. As the sun goes down, at about 15:00 hours, when the renewable energy in the microgrid cannot meet the electricity demand of commercial activities, the diesel engine will be started to supplement and electricity will be purchased from the main grid. In order to verify the superiority of the algorithm proposed in this chapter in the multi-microgrid optimal scheduling problem, the original GA and the IGA proposed in this study are used to optimize the objective function established in this article, respectively. The optimization process of the two methods is shown in Figure 10.
[IMAGE OMITTED. SEE PDF]
As can be seen from Figure 10, since the algorithm in this study improves the mutation operator on the basis of the original GA, the fitness value of the algorithm proposed in this study is smaller than that of the original GA when solving the same multi-microgrid optimal scheduling problem. At the same time, the original GA stopped searching after about 870 times. However, the algorithm proposed in this article found the optimal solution in 650 times, which is smaller than the original algorithm. Therefore, the IGA proposed in this article is easier to find the optimal solution of multi-microgrid optimization than the original algorithm, with a faster speed.
CONCLUSION
In this article, we first present our multi-microgrid scenario, and at the same time make a simple analysis of the cooperative operation conditions of this scenario. Then, according to the multi-microgrid scenario we have established, under the mechanism of peak-valley electricity price and inter-microgrid electricity price, with the objective of minimizing the operating cost and environmental protection cost of the whole multi-microgrid system, we have established our microgrid economic dispatching model under the condition of satisfying the relevant power supply constraints. In order to solve the optimal scheduling problem of this multi-microgrid system, an IGA is proposed. Finally, the scheduling algorithm is verified through simulation experiments. The results show that the method can effectively solve the optimal scheduling problem of the multi-microgrid and improve the power supply reliability and economy of the microgrid system.
Wu, J., et al.: Research on coordinated control technology of multi‐microgrid for large‐scale parks. Electr. Appliances Energy Effic. Manage. Technol. (25), 35–41 (2017)
He, L., Lu, H.: Study on optimal dispatch of multi‐microgrid considering economy. Power Gener. Technol. 39(5), 397–404 (2018)
Fridaus, U., Penangsang, O.: Economic dispatch of multi microgrid systems with renewable energy sources using particle swarm optimization. Int. Semin. Intell. Technol. Appl. 595–600 (2016)
Zhao, M., et al.: Characteristic analysis of microgrid group and demonstration engineering design. Power Grid Technol. 6(5), 125–138 (2015)
Nikmehr, N., Najafi‐Ravadanegh, S.: Probabilistic optimal power dispatch in multi‐microgrids using heuristic algorithms. 2014 Smart Grid Conference (SGC) 1–6 (2014)
Wang, J., et al.: Microgrid economic dispatch method considering electric vehicles. Power System Protection and Control 44(17), 111–117 (2016)
Liu, C., Yang, W.: Cloud computing multi‐objective task scheduling based on genetic particle swarm optimization. Comput. Technol. Dev. 27(2), 56–59 (2017)
Wang, X., Cao, L.: Genetic Algorithm‐Theory, Application and Software Implementation, Xi ’an, pp. 20–40. Xi ’an Jiaotong University Press (2002)
Mohamed, F.A., Koivo, H.N.: System modelling and online optimal management of microgrid using multiobjective. International conference on clean electrical power. IEEE, pp. 148–153 (2007)
Xiangding et al.: Electric vehicles participated in the study of optimal peak‐valley price of V2G. Chinese J. Electr. Eng. 33(31), 15–25 (2013)
Liu, L., Yan, Q.: Research on improvement and application of differential evolution algorithm. Jiangxi Sci. 36(4), 573–578 (2018)
Wang, L., Xu, J., Shangyong. Y.: Joint optimal dispatch of multi‐microgrid systems under time‐sharing tariff mechanism based on multi‐agent improved particle swarm optimization. J. Power Supply. 1–17 (2018)
Yang, Z.: Research on reactive power optimization of microgrid based on genetic algorithm. Integr. Circuit Appl. 35(12), 93–94 (2018)
Huang, J., Ai, J., Wan, J.: Aircraft aerodynamic parameter identification based on simulated annealing particle swarm optimization. Fudan J. (Natural Science Edition). 55(3), 336–341 (2016)
Huang, H., Lin, J., Wang, J.: Cloud resource scheduling algorithm based on improved hybrid genetic algorithm. Telev. Technol. 39(18), 36–41 (2015)
Wu, F., Zhang, H.: Optimal configuration of energy storage for distributed photovoltaic generation. 2017 China international electrical and energy conference (CIEEC2017), pp. 332–337 (2017)
Lu, Q., et al.: Gauss mutation differential evolution algorithm and its application. J. Eng. Des. 19(5), 372–378 (2012)
You have requested "on-the-fly" machine translation of selected content from our databases. This functionality is provided solely for your convenience and is in no way intended to replace human translation. Show full disclaimer
Neither ProQuest nor its licensors make any representations or warranties with respect to the translations. The translations are automatically generated "AS IS" and "AS AVAILABLE" and are not retained in our systems. PROQUEST AND ITS LICENSORS SPECIFICALLY DISCLAIM ANY AND ALL EXPRESS OR IMPLIED WARRANTIES, INCLUDING WITHOUT LIMITATION, ANY WARRANTIES FOR AVAILABILITY, ACCURACY, TIMELINESS, COMPLETENESS, NON-INFRINGMENT, MERCHANTABILITY OR FITNESS FOR A PARTICULAR PURPOSE. Your use of the translations is subject to all use restrictions contained in your Electronic Products License Agreement and by using the translation functionality you agree to forgo any and all claims against ProQuest or its licensors for your use of the translation functionality and any output derived there from. Hide full disclaimer
© 2021. This work is published under http://creativecommons.org/licenses/by-nc-nd/4.0/ (the "License"). Notwithstanding the ProQuest Terms and Conditions, you may use this content in accordance with the terms of the License.
Abstract
A multi‐microgrid economic dispatching strategy based on adaptive mutation genetic algorithm is proposed for multi‐microgrid systems with different load types and power demands. Based on the analysis of industrial, residential and commercial loads, considering the synergy and complementarity between multi‐microgrids, an optimal dispatching model of multi‐microgrids based on the minimum operation cost and environmental protection cost of multi‐microgrids is established from the point of view of environmental protection and economy. At the same time, an adaptive mutation genetic algorithm is proposed to optimize the system model and find the optimal economic dispatching scheme.
You have requested "on-the-fly" machine translation of selected content from our databases. This functionality is provided solely for your convenience and is in no way intended to replace human translation. Show full disclaimer
Neither ProQuest nor its licensors make any representations or warranties with respect to the translations. The translations are automatically generated "AS IS" and "AS AVAILABLE" and are not retained in our systems. PROQUEST AND ITS LICENSORS SPECIFICALLY DISCLAIM ANY AND ALL EXPRESS OR IMPLIED WARRANTIES, INCLUDING WITHOUT LIMITATION, ANY WARRANTIES FOR AVAILABILITY, ACCURACY, TIMELINESS, COMPLETENESS, NON-INFRINGMENT, MERCHANTABILITY OR FITNESS FOR A PARTICULAR PURPOSE. Your use of the translations is subject to all use restrictions contained in your Electronic Products License Agreement and by using the translation functionality you agree to forgo any and all claims against ProQuest or its licensors for your use of the translation functionality and any output derived there from. Hide full disclaimer
Details
1 School of Electronics and Information Engineering, Guangdong Ocean University, Guangdong, China
2 School of Automation, Hangzhou Dianzi University, Hangzhou, China
3 School of Logistics Engineering, Shanghai Maritime University, Shanghai, China