Study on the Microgrid Pluripotent Complementary and Comprehensive Economic Optimization Planning Based on Virtual Energy Theory

In this paper, a comprehensive economic allocation method based on virtual energy theory for microgrid complementary power supply is proposed, which is successfully applied to the multi energy complementary planning and comprehensive economic planning of an island microgrid. In order to evaluate and optimize the microgrid in the planning stage more objectively based on the characteristics of various combinations of renewable micro generation and flexible operation modes, a new method of the micro-grid comprehensive economic configuration and optimization of the distributed generation is proposed based on virtual energy theory. Firstly, the Micro-grid model structure based on the virtual energy theory is given. Secondly, the virtual energy conversion model of different distributed power generations is also studied. The micro-grid pluripotent complementary and comprehensive economic optimization model (PCCEP) based on virtual energy theory are reached too. Combined with a practical example, a PCCEP method of the micro-grid based on the Homer software platform is discussed. The case study indicates that the method can avoid the uncertainty of the decision-making factors and the subjectivity of expert evaluation. It is shown that the new method is a practical solution for microgrid comprehensive economic optimization planning.


Introduction
Micro grid [1], [2], [3] is a small power distribution system consisting of distributed power supply device, energy storing device, energy conversion device, load, protection and monitoring device, etc. Micro grid has become an effective form of renewable energy utilization and has great application prospects [4], [5]. Because the characteristics of distributed power supply vary greatly, the multi-energy complementation, comprehensive economic allocation and evaluation of the micro-grid has become a technical problem to be solved urgently. In this paper, a comprehensive economic allocation method based on virtual energy theory for microgrid complementary power supply is proposed. And the new method has been successfully applied to the multi energy complementary planning and comprehensive economic planning of an island micro grid.
The remainder of this paper is organized as follows. In the next section, we propose and study the transformation model of distributed generation based on virtual energy theory. The comprehensive economic planning of microgrid based on virtual energy theory is introduced in detail. Section IV reports the simulation and analysis of an example. The conclusions drawn from this paper are given in Section V.

TRANSFORMATION model of distributed generation based on virtual energy theory
The microgrid structure scheme based on the virtual energy theory is shown in Fig.1. The power generation and storage process of distributed generation requires corresponding power control devices [1], [6]. The cost efficiency factor (  ) of the distributed generation is constructed according to the power generation and storage characteristics. The conversion model of power generation from PV, wind power generation and Diesel engine power generation to virtual energy pool can be established. The volume conversion formula of distributed generation (such as wind energy, photovoltaic, gas turbine power generation, etc.) into virtual energy pool is put forward. Using these formulas and models, the transformation and comprehensive evaluation model from the distributed generation to virtual energy pool is put forward, which can be used for the comprehensive evaluation of energy efficiency of distributed generation in micro grid. By optimizing the volume of the total virtual energy pool, the multi energy complementary optimization of microgrid can be realized. This model can solve the problem of lacking effective evaluation and management model in microgrid planning.

Conversion formula from distributed generation to virtual energy
First point: Conversion formula from photovoltaic power to virtual energy. Referring to the model in Fig. 1 Where, the meaning of 2  is the same as those in formula (1), is the total power generation of wind power generation, p C is the efficiency of the Wind turbine generation, A is the effective page area of the fan,   is the air density, and V is the wind speed.
Third point: Conversion formula from gas turbine power generation to virtual energy. Referring to the model in Fig. 1, the conversion formula from gas turbine power generation to the virtual energy pool can be constructed. The concrete form is shown in Formula (3).
Where, the meaning of 3 Referring to the model in Fig. 1, the conversion formula from other forms of distributed energy generation to the virtual energy pool can be constructed. The total power consumption of the other types of distributed power generation devices that flow through the power control devices is x W Pdt   . The conversion formula from other forms of distributed energy generation to the virtual energy pool is shown in formula (3). 9.8 Where, the meaning of x as those in formula (1).

Distributed microgrid generation model based on virtual energy theory
After calculating the volume of the virtual energy pool corresponding to the distributed power generation, the total capacity of the virtual power pool corresponding to the whole micro grid power generation can be calculated according to formula (5).
The total capacity of the virtual energy pool corresponding to the microgrid is determined by the type, quantity and power generation of the distributed generation mode used in the system. And the total capacity of virtual energy pool and the capacity of each virtual energy pool are normalized. Then, the per-unit total capacity of virtual energy pool can be determined with the formula (6).
The nominal energy capacity of the rated energy in the microgrid virtual energy pool can be shown in formula (6). Where, is the capacity factor of the virtual energy pool of photovoltaic power generation.

Economic benefit calculation of distributed generation based on virtual energy
Micro grid consists of photovoltaic, wind power, diesel power generation, water power, various energy storage and other forms of energy [7], [8], [9]. The distributed generation can build the micro grid under certain resource input. Based on the model in the previous chapters, the micro grid of the virtual energy pool can be evaluated economically [10]. Thus, the comprehensive economic evaluation of distributed generation can be realized. During the operation period of the micro grid, all kinds of distributed generation equipment with negative economic benefit coefficient are put into operation, and the corresponding energy resources(F) are put into operation too. Considering the environment, pollution and other factors, the corresponding energy resources can have positive or negative economic benefit coefficient [11], [12]. the corresponding energy resources(F) is sent to the distributed generation equipment respectively and convert to virtual energy [13], [14]. So, the comprehensive energy efficiency characteristics of microgrid based on virtual energy can be expressed as the relationship between F and V in microgrid operation. If F is input and V is output, it can be called the comprehensive benefit input output characteristic of microgrid system. First point: Comprehensive benefit characteristics of micro grid generation unit. The relationship between the economic benefits (F)and virtual energy(V)is extremely complex. In general, F is not only a function of V, but also related to the changing factors of V. This paper only discusses the static relation of F as input and V as output. And the comprehensive benefit characteristics of microgrid generation can be expressed as: Where, V is the virtual power output of microgrid power generation, the unit is m 3 . F is the comprehensive benefit cost of the distributed generation of microgrid, which can be calculated with the economic cost per hour (the cost is related to the market factors). It is an efficient way to measure the cost of consumption by consuming economic costs, which is widely used in the economic field to measure the operating efficiency of the system [15], [16]. In fact, if the cost of the actual economic benefits generated by the distributed generation unit is used to measure the input cost, the cost of economic benefits will be more effectively reflected, and this method will be more reasonable, which will be discussed in detail in another paper.
Second point: The comprehensive benefit factor of microgrid generation unit. The cost efficiency factor [17] described in the preceding section is a multivariable function, and its expression is, G is the government support index for energy storage devices, and j L is the annual utilization index of energy storage device. A static calculation method of partial benefit index is given below. 1) Energy saving benefit calculation. The energysaving benefits of microgrid include two aspects: ○ 1The value of fossil energy saved by renewable energy power generation. ○ 2The value of heat and cold energy provided by the waste heat generated by the waste heat of the micro gas turbine under the mode of cooling, heat and power cogeneration. Taking the gas turbine as an example, the energy saving efficiency is calculated by: Where, k coal is the average coal consumption of power generation unit for thermal power unit and the value is 357 g/kwh, P coal is Unit price for coal and the value is 700 ￥/t, E i is the annual power generation for Type I distributed generation, E 3 , x 3 are the annual generating capacity and installation number of micro gas turbine (i=3) respectively, p he , p co are thermal energy price and cold energy price respectively and the prices are 51.4 ￥/GJ and 79.6 ￥/GJ.
2)Environmental benefit calculation. The mitigation benefits of micro grids can be measured by the environmental benefits of the same amount of electricity produced by the microgrid [15], [18], and the environmental benefits are generally represented by the amount of pollutants such as SO 2 , NOx, CO 2 , CO and fly ash which are less discharged by distributed renewable energy power generation than that by the traditional coal-fired power generation. The calculation formula of environmental benefit is: Where, V j is the environmental value of item j pollutants, M is the type of pollutant, δ c,j is the quantity of j pollutant discharged by coal fired unit producing a certain amount of electric energy, δ i,j is the number of j pollutants emitted by i power generation units producing a certain amount of electric energy, j=1 ～ 5 represents SO 2 , NOx, CO 2 , CO and ash, respectively.
3) Reliability benefit calculation. When the power grid fails, the microgrid can run autonomously to ensure the power supply to the local load, which will improve the reliability of the power supply [19], [20]. Microgrid can improve power supply reliability and reduce power outage. Power outage will bring huge economic losses. The reliability benefit of microgrid is the reduced economic losses caused by microgrid. The formula for calculating reliability benefit is formula (12).
Where, EAR I is the outage loss evaluation rate, And it is

Comprehensive economic planning of microgrid based on virtual energy theory
The economic optimization of the micro grid distributed generation is to allocate the active power of power generation to the power generation units of the micro grid, making the total operating cost of the micro grid the lowest and the economic benefit the highest. There are two kinds of integrated economic planning methods in microgrid with multiple distributed generation systems. (1) The basic load benefit method, First, the distribution power generation is arranged according to the operating efficiency of a running point (such as the highest economic benefit operation point). Then, the load is added to the distributed generation with high benefit, and then the analogy is made. (2) The optimal point load method, first, the distributed generation is queuing according to the optimal economic benefits. Then, the distributed generation is allocated to the optimal operating point rather than the full load of a single distributed generation. The two methods are only based on a single factor to optimize the allocation of distributed generation, So, the comprehensive economy and benefit of microgrid cannot be accurately evaluated with those methods. In the next section, a comprehensive economic planning method for microgrid based on virtual energy theory is studied.

Objective function
In this paper, the comprehensive economic planning method of microgrid power is described as follows: By using of virtual energy pool theory, the cost, limit capacity and load characteristics of various alternative distributed power sources need to be referred and further studied. The wind, solar radiation, diesel and load are used as input data. The maximum net benefit of microgrid is the optimization objective. Considering the reliability of microgrid, the adequacy of islanding operation and the environmental constraints such as carbon emissions, the installation number of various alternative power sources is optimized. The objective function of the comprehensive benefit factor is as follows.
Where,     years. E i is the annual power output for the i distributed generation. A is the discount rate, and the value is 6% in this paper. i  is the cost correction factor, which is related to the environmental protection intensity, technical maturity and government support of the type i distributed generation. The correction coefficients of PV and wind energy ranges from 0.5 to 0.8. The correction coefficient of diesel power generation ranges from 1.0 to 1.5. The correction coefficient of hydraulic generator ranges from 0.7 to 0.8. The correction coefficient of thermal power generation ranges from 1.2 to 1.5.
Second point: Comprehensive benefits of distributed generation. This part calculates the energy saving benefits, emission reduction benefits, loss reduction benefits and reliability benefits of distributed generation and Microgrid respectively. The summation of the four benefits can obtain the comprehensive benefits of microgrid power generation configuration. The concrete formula is: Where,  Second point: Calculation constraint of battery charging and discharging. Excessive battery charge and discharge rate will reduce the service life of the battery. So, the charge and discharge capacity of the battery cannot exceed 20% of its available capacity per hour. At a certain t time point, when the sum of the generated energy of the fan and the PV array is greater than the load, the battery pack is charged. The battery constraint satisfies the following formula. Third point: Calculation constraint of power supply reliability for islanding operation of microgrid. Loss of load probability ( LOLP ) is a common index to characterize the reliability of microgrid system, and its magnitude is the ratio of the load demand when the system cannot meet to the total load demand. In order to ensure the generation adequacy of islanding operation of microgrid, LOLP should satisfy the following formula. ,0

Constraint conditions
is the maximum allowable outage probability of microgrid running throughout the year, , Total t E is the total power output for the microgrid in t hours, t dE is the discharge capacity of battery in t hours, When the battery is charged, the value is 0.
Fourth point: Calculation constraint of pollutant emission. With the global warming, the study of reducing emissions (CO 2 , CO, SO 2 , PM particles, hydrogen and oxygen compounds, nitrogen oxides, etc.) and development of green environmental protection economy has become the current research hot spot. In this paper, the pollutant emission constraint is added to the microgrid power optimization allocation model to evaluate and limit the pollutant emission of micro grid power generation. Pollutants will be directly or indirectly discharged from fans, photovoltaic arrays, micro gas turbines, and battery power generation, the constraint conditions of micro grid pollutant emission can be described as follows: Where, max e is the maximum allowable value of annual pollutant discharge for microgrid. The connection of distributed power will affect the node voltage and line current in microgrid. In this paper, the influence of node voltage constraint and line current constraint is considered.

Economical optimal allocation method of microgrid distributed generation
Genetic algorithm is a random search optimization algorithm simulating biological evolution. It has strong robustness and adaptability and has been applied to study the power system planning problems. In this paper, an adaptive genetic algorithm based on binary encoding is put forward. Each component of the chromosome is the number of distributed generators in the microgrid. And the binary coding length of each component depends on the single machine capacity of distributed generation corresponding to the component, the peak load demand within the microgrid and the encoding accuracy. Based on the improved genetic operator method proposed by the literature [16] and Taguchi algorithm, the crossover operation is optimized, which will improve the convergence and global performance of the algorithm. Considering the influence of the constraints, the penalty function is constructed to transform the constrained problem into the unconstrained problem in this paper. Photovoltaic, wind, micro gas turbines and other power generation constraints have been discussed in the microgrid configuration optimization model based on the virtual energy theory. Other constraints are contained in the fitness function in the form of penalty function.
Where, M is the number of constraints,   pi  is the i constraint penalty item, i  is the corresponding penalty factor, and it is a larger positive number.

Simulation and analysis of an example
In this section, an example of optimal allocation of distributed generation in island microgrid is given to verify new method of microgrid pluripotent complementary and comprehensive economic optimization planning based on virtual energy theory. This island (Xiachuan island) is in Guangdong province of China. The photovoltaic, diesel generator, wind power, and tidal energy of the island are very rich. The island power supply system connects with land power grid through submarine cable. Submarine cables have a maintenance period of 3 months a year. The island power demand is stable. Based on the above constraints, the specific experimental process is as follows.

Island load forecasting
Annual load of Xiachuan Island is basically consistent, and the annual power supply and electricity sales are basically consistent too. According to the electric load from 2010 to 2014, the electrical load of 2015 year can be obtained based on the artificial neural network method, as shown in Figure 2(unit: kwh). It can be shown that September is the peak period of annual electricity consumption, and January is the troughs period of annual electricity consumption. Annual load of Xiachuan Island is basically consistent, and the annual power supply and electricity sales are basically consistent too. According to the electric load from 2010 to 2014, the electrical load of 2015 year can be obtained based on the artificial neural network method, as shown in Figure 2(unit: kwh). It can be shown that September is the peak period of annual electricity consumption, and January is the troughs period of annual electricity consumption.
The island load is mainly based on Residents' electricity and commercial electricity. Daily load curve of typical household electricity consumption is selected as the daily load curve of electricity consumption in Xiachuan Island in this paper. The largest power load in September and the minimum electricity consumption in January from 2010 to 2014 were analyzed. The daily load curve of January and September can be predicted based on support vector machine algorithm, as shown in Figure 3.

Sep Jan
Daily Load Curve for Jen. and Sep. According to the analysis of the load data in Figure 3, the maximum daily load is about 10 megawatts, while in January it is only about 3 megawatts. The load has great seasonal and temporal characteristics.

Model constraint
First point: Wind power generation constraints. The annual average wind speed in Xiachuan Island is up to 4.85m/s. The annual effective wind speed time length can reach more than 60% of the total wind speed time. The effective wind energy value varies greatly in each month, the high value period is from October to February next year, and the low value period is from April to August. Daily power generation peak is 21:30 to 6:30, power generation trough at 9:30 to 15:30, as shown in figure 4.
Second point: Photovoltaic power generation constraints. The annual sunshine hours in Xiachuan Island can reach about 2006 hours, and the annual solar radiation energy reached 137.2kwh/m 2 . The annual solar radiation is the smallest in February, the largest in July. The shortest average sunshine duration is from early February to mid-April. The average daily sunshine is only 3.3 hours. The longest average sunshine duration is in July. The average daily sunshine is only7.6 hours, as shown in Fig. 5. Third point: Other boundary conditions.1 ) Diesel can get adequate supply in Xiachuan Island.2) Distributed wind power generation can be completely consumed by the energy storage system and cable and under local load.3) The system simulation period is 25 years.4) The cost of fan and PV includes the cost of auxiliary equipment such as converter.5) Line loss is not taken into account.

Simulation analysis
According to the basic configuration of the first phase of the intelligent microgrid, the micro grid configuration model is set up in HOMER software, the number of wind power equipment running, and the selection range of distributed generation type are also set up. By inputting the above data, the simulation model can be built in the main interface of HOMER, as shown in Fig.  6.
In the Homer software platform, based on the model of Fig. 6 and the calculation model in the previous section, the simulation is carried out. 480 kinds of matching schemes are calculated, and 4 typical schemes are selected by HOMER software. In the HOMER simulation scheme, the optimal scheme is scheme 1, and then the second, third, fourth optimal schemes are the second, third, fourth scheme. The basic configuration, capital, power generation and emission of the four schemes are analyzed in the next section. 1) Capacity distribution comparison of microgrid distributed generation.
The system configuration comparison of the 4 schemes calculated by HOMER simulation software is shown in table 1.
2) The system Investment cost comparison The system Investment cost comparison of the 4 schemes calculated by HOMER simulation software is shown in Table 2.
3) Comparison of emission data The system comparison of emission data of the 4 schemes calculated by HOMER simulation software is shown in Table 3. According to HOMER simulation software calculation, it is the optimal scheme that can meet the relevant requirements of emissions, power generation, investment maintenance costs and other factors.
Based on the virtual energy pool and HOMER simulation software, the modeling and simulation of the intelligent configuration of intelligent micro grid in Xiachuan Island are carried out. The comparison results between simulation cases have certain reference value for the microgrid pluripotent complementary and comprehensive economic optimization planning, which can optimize intelligent microgrid distributed power generation networking scheme.

Conclusion
A microgrid structure model based on virtual energy pool is proposed in this paper. A comprehensive benefit planning and optimization method for microgrid is proposed, which can optimize and evaluate the economic efficiency of microgrid construction and operation. An example of economic planning for distributed energy resources of an island microgrid is given. The simulation results are analyzed in detail. It is proved that this method is objective, accurate, credible and operable. It is an effective method in microgrid planning process. This method also has a high reference value for the economic evaluation of micro grid and other aspects of smart grid.