CN108683211A - A kind of virtual power plant combined optimization method and model considering distributed generation resource fluctuation - Google Patents

Abstract

The invention discloses a combination optimization method and model of a virtual power plant considering the fluctuation of distributed power sources. Predict the output of distributed power sources; calculate the matching degree between the predicted output of each distributed power source and the power generation plan of the virtual power plant according to the matching degree calculation formula; incorporate each distributed power source when the minimum matching degree is obtained into the virtual power plant; judge Whether the virtual power plant formed by distributed power meets the dispatching needs of the system; form a virtual power plant. The invention adopts a multi-stage nonlinear integer programming method to establish a combination optimization model of a virtual power plant in an uncertain situation. The model takes the operating state of distributed resources as a decision variable, and aims at minimizing the matching degree. Govern virtual power plant portfolio decisions.

Description

A combined optimization method for virtual power plants considering the fluctuation of distributed power sources and its Model

technical field

The invention relates to a power system optimization method, in particular to a combined optimization method and model of a virtual power plant considering the fluctuation of distributed power sources.

Background technique

In recent years, virtual power plant, as a new form of distributed power generation development, has attracted more and more attention from the academic circles. The virtual power plant under the background of the smart grid has been expanded and extended. It can be understood that the distributed power supply, controllable load and energy storage system in the distribution network are combined as a special power plant to participate in the grid operation through the distributed power management system. In this way, the contradiction between smart grid and distributed power can be well coordinated, and the value and benefits brought by distributed energy to the power grid and users can be fully explored; unlike microgrids, virtual power plants cannot operate in isolation, and distributed power in virtual power plants The actual operating network needs to be procured through an external grid. The introduction of the concept of virtual power plants makes it possible for distributed power sources to be put into grid operation on a large scale, and can also provide services for the management of transmission systems.

With the rapid development of the user-side power supply, the virtual power plant technology that effectively utilizes the user-side distributed power supply will be widely used. At present, the domestic research on the user-side virtual power plant has just started. Mechanistic research on access and operation characteristics is still lacking, so this method has very important practical significance.

The combination method of the existing virtual power plant is divided according to the region, and all the distributed resources in the region are aggregated and expressed, and the obtained single virtual power plant model representing the region does not consider the active optimization problem. In some scenarios It cannot meet the requirements of the economical operation of the power system.

Contents of the invention

Purpose of the invention: In order to optimize the combination of distributed power sources inside the virtual power plant, the present invention considers the volatility of distributed power sources, establishes an optimal combination model of virtual power plants, and provides a combination optimization method for virtual power plants that considers the volatility of distributed power sources and models.

Technical solution: the present invention provides a virtual power plant combination optimization method considering the volatility of distributed power sources, including the following steps:

(1) According to the geographical location, environmental conditions, resource distribution, etc. of different regions, a variety of factors are comprehensively considered to choose distributed power sources reasonably, and predict the output of each distributed power source;

(2) According to the calculation formula of matching degree, calculate the matching degree of each distributed power generation's predicted output and the power generation and consumption plan of the virtual power plant;

(3) Incorporate each distributed power source when the minimum matching degree is obtained into the virtual power plant;

(4) Judging whether the virtual power plant formed by distributed power meets the scheduling needs of the system;

(5) Form a virtual power plant.

Further, in the step (2), the matching degree of the predicted output of each distributed power source and the power generation and consumption plan of the virtual power plant is calculated by the following formula:

The specific formula for calculating the matching degree in the period t is:

In the formula, S _t represents the matching degree of time period t, N is the type of distributed power generation, i is the number of distributed power generation, P _i ^t is the predicted output of distributed power generation i in time period t, and P _e is the power generation and consumption plan of the virtual power plant ;

The matching degree of distributed power generation based on renewable resources in the entire scheduling period (1,2,...t,...T) is:

In the formula, S represents the matching degree of the entire scheduling cycle, T is the time period selected by the scheduling cycle, t is the sequence number of the time period, N is the type of distributed power supply, i is the number of distributed power supply, P _i ^t is the distributed power supply i Forecast output in period t, P _e is the planned electricity generation and consumption of virtual power plant.

Further, the objective function of the combined optimization model of each distributed power supply during the minimum matching degree in the step (3) is:

Among them, T is the time period selected by the scheduling cycle; t is the number of the time period; N is the type of distributed power; i is the type number of distributed power; P _i ^t is the predicted output of distributed power _i at time t; power generation and consumption plan of the power plant; u _i represents the 0-1 variable of the state of the distributed generation, u _i = ₁ indicates that the distributed Power generation unit of a power plant.

Further, according to the demand, choose the day-ahead, monthly, quarterly or annual scheduling as the virtual power plant planning scheduling cycle, and the corresponding S is the daily matching degree, monthly matching degree, quarterly matching degree or annual matching degree.

Further, the scheduling in the step (4) needs to be a constraint condition, including:

(41) Power generation plan constraints:

in, Respectively represent the charge and discharge power of the battery at time t; are the charge and discharge efficiencies of the battery, respectively; is the load demand of the virtual power plant at time t, is the power generation plan submitted by the virtual power plant to the grid; u _i represents the 0-1 variable of the state of the distributed power generation, u _i = 1 means that the distributed power source i is the power generation unit of the virtual power plant, and u _i = 0 means that the distributed power source i Not as a power generation unit of a virtual power plant; P _i ^t is the predicted output of distributed power i at time t;

(42) Controllable distributed power output upper and lower limit constraints:

in, Indicates the minimum output of the controllable distributed power supply, Indicates the maximum output of the controllable distributed power supply;

(43) The upper and lower limits of battery charging and discharging power:

SOC _min ≤ SOC ^t ≤ SOC _max (9);

in, Respectively, the charging and discharging power of the battery at time t; Respectively, the minimum charge and discharge power of the battery; are the maximum charging and discharging power of the battery; SOC ^t is the storage capacity of the battery at time t, SOC _min is the minimum storage capacity of the battery, and SOC _max is the maximum storage capacity of the battery;

If the scheduling requirements are met, select the above-mentioned distributed power sources, and perform step (5) to form a virtual power plant; if the scheduling requirements are not met, return to step (3) to adjust the selected distributed power sources until the scheduling requirements are met.

The present invention also provides a combined optimization model of a virtual power plant considering the volatility of distributed power sources, the model includes an objective function and constraint conditions, and the objective function is:

Among them, T is the time period selected by the scheduling cycle; t is the number of the time period; N is the type of distributed power; i is the type number of distributed power; P _i ^t is the predicted output of distributed power _i at time t; power generation and consumption plan of the power plant; u _i represents the 0-1 variable of the state of the distributed generation, u _i = ₁ indicates that the distributed power generation units of power plants;

The constraints include:

(a) Generation plan constraints:

in, Respectively represent the charge and discharge power of the battery at time t; are the charge and discharge efficiencies of the battery, respectively; is the load demand of the virtual power plant at time t, is the power generation plan submitted by the virtual power plant to the grid; u _i represents the 0-1 variable of the state of the distributed power generation, u _i = 1 means that the distributed power source i is the power generation unit of the virtual power plant, and u _i = 0 means that the distributed power source i Not as a power generation unit of a virtual power plant; P _i ^t is the predicted output of distributed power i at time t;

(b) Controllable distributed power output upper and lower limit constraints:

in, Indicates the minimum output of the controllable distributed power supply, Indicates the maximum output of the controllable distributed power supply;

(c) The upper and lower limits of battery charging and discharging power:

SOC _min ≤ SOC ^t ≤ SOC _max (9);

in, Respectively, the charging and discharging power of the battery at time t; Respectively, the minimum charge and discharge power of the battery; are the maximum charging and discharging power of the battery; SOC ^t is the storage capacity of the battery at time t, SOC _min is the minimum storage capacity of the battery, and SOC _max is the maximum storage capacity of the battery.

Beneficial effects: Compared with the prior art, the present invention adopts the multi-stage nonlinear integer programming method to establish the combined optimization model of the virtual power plant under uncertain conditions. The goal is to real-time control the virtual power plant combination decision-making according to the system power generation plan. According to the geographical location, environmental conditions and resource distribution of different regions, a variety of factors are comprehensively considered to choose distributed power sources reasonably, and make full use of local resources to take advantage of distributed power generation. Reasonable selection of the type of distributed power generation can fully develop and utilize various available scattered energy sources, improve the utilization efficiency of sources, and reduce investment costs.

Description of drawings

Fig. 1 is a flow chart of the method of the present invention.

Detailed ways

The technical solutions of the present invention will be described in detail below in conjunction with the accompanying drawings and specific embodiments.

Based on the concept of distributed power supply matching degree, the present invention establishes a distributed power supply combination optimization model. In the virtual power plant planning and scheduling period, the model aims to minimize the matching variance of distributed power generation, satisfy certain constraints, and obtain the optimal combination scheme from the combination of distributed power generation of different types and capacities. The matching degree defined by the method of the present invention is the output matching degree of the distributed power supply, that is, the matching degree between the predicted curve of the distributed power supply and the actual output curve.

Combined with the concept of variance in mathematics, the method of the present invention proposes the concept of distributed power output matching degree, and provides a specific calculation formula, which is a quantified characterization value when selecting a suitable distributed power source to form a virtual power plant.

In a distributed system based on distributed power sources based on renewable energy, it generally contains clean distributed power sources such as wind power and photovoltaics with strong volatility; it will also inevitably contain other distributed power sources with relatively higher output stability such as micro gas turbines etc. to obtain better quality electric energy. Therefore, there will be a certain difference between the total output curve and the total expected expectation curve of the distributed power system with strong output fluctuations, and the present invention uses the concept of matching degree, combined with the predicted output of distributed power, so as to To determine the optimal combination of virtual power plants.

As shown in Figure 1, a virtual power plant combination optimization method considering the fluctuation of distributed power sources of the present invention includes the following steps:

(1) According to the geographical location, environmental conditions, resource distribution and other factors in different regions, the distributed power supply is reasonably selected, and the output of each distributed power supply is predicted in combination with the existing power generation forecast data and power generation probability model;

(2) According to the calculation formula of matching degree, calculate the matching degree of each distributed power generation's predicted output and the power generation and consumption plan of the virtual power plant;

The specific formula for calculating the matching degree in the period t is:

In the formula, S _t represents the matching degree of time period t, N is the type of distributed power generation, i is the number of distributed power generation, P _i ^t is the predicted output of distributed power generation i in time period t, and P _e is the power generation and consumption plan of the virtual power plant .

The closer the power generation-power consumption capacity is, that is, the closer S _t is to zero, the smaller the value of the matching degree, indicating that it contains less uncertain information. The smaller the value of the matching degree, the better the power generation-consumption characteristics match, and it is suitable to be combined as a partner.

The matching degree of distributed power generation based on renewable resources in the entire scheduling period (1,2,...t,...T) is:

In the formula, S represents the matching degree of the entire scheduling cycle, T is the time period selected by the scheduling cycle, t is the sequence number of the time period, N is the type of distributed power supply, i is the number of distributed power supply, P _i ^t is the distributed power supply i Forecast output in period t, P _e is the planned electricity generation and consumption of virtual power plant.

(3) According to the above-mentioned concept of distributed power matching degree, a distributed power generation combination optimization model is established, and each distributed power source that can obtain the minimum matching degree is included in the virtual power plant;

The objective function of this model is:

Among them, T is the time period selected by the scheduling cycle; t is the number of the time period; N is the type of distributed power; i is the type number of distributed power; P _i ^t is the predicted output of distributed power _i at time t; power generation and consumption plan of the power plant; u _i represents the 0-1 variable of the state of the distributed generation, u _i = ₁ indicates that the distributed Power generation unit of a power plant. S is the matching variance of distributed power generation. The smaller S is, the more the actual power generation curve matches the expected power generation curve, and the higher the similarity.

According to the demand, you can choose day-ahead, monthly, quarterly or annual scheduling as the planning and scheduling cycle of the virtual power plant. If it is day-ahead scheduling, use hours as the scale, T is taken as 24 hours, and the corresponding S is defined as the daily matching degree, that is, the distributed power supply The degree of matching between the sum of daily power generation and the daily expected power generation curve is similar for other dispatch cycles. According to the characteristics of each distributed power source, when selecting the type of distributed power source in a virtual power plant, at least one dispatchable distributed power source should be included to improve the overall stability of the system.

(4) Judging whether the virtual power plant formed by the above-mentioned distributed power sources meets the scheduling needs of the system, that is, the constraints;

Among them, the constraints include:

(41) Power generation plan constraints:

in, Respectively represent the charge and discharge power of the battery at time t; are the charge and discharge efficiencies of the battery, respectively; is the load demand of the virtual power plant at time t, is the power generation plan submitted by the virtual power plant to the grid; u _i represents the 0-1 variable of the state of the distributed power generation, u _i = 1 means that the distributed power source i is the power generation unit of the virtual power plant, and u _i = 0 means that the distributed power source i Not as a power generation unit of a virtual power plant; P _i ^t is the predicted output of distributed power i at time t.

(42) Controllable distributed power output upper and lower limit constraints:

in, Indicates the minimum output of the controllable distributed power supply, Indicates the maximum output of the controllable distributed power supply.

(43) The upper and lower limits of battery charging and discharging power:

SOC _min ≤ SOC ^t ≤ SOC _max (9);

in, Respectively, the charging and discharging power of the battery at time t; Respectively, the minimum charge and discharge power of the battery; are the maximum charging and discharging power of the battery; SOC ^t is the storage capacity of the battery at time t, SOC _min is the minimum storage capacity of the battery, and SOC _max is the maximum storage capacity of the battery.

If the dispatching requirements are met, select the above-mentioned distributed power sources, and perform step (5) to form a virtual power plant; power until the scheduling needs are met.

(5) Form a virtual power plant.

A virtual power plant combination method that can obtain the best match between distributed power output and forecast. The optimization model is established with the goal of minimizing the matching degree of distributed power sources in the scheduling cycle of the virtual power plant planning day-ahead, and meets the constraints of power generation planning, the upper and lower limits of controllable distributed power output, and the upper and lower limits of battery charge and discharge power. , The optimal combination scheme is obtained from the combination of distributed power sources with different capacities.

Based on the fluctuation analysis of distributed power generation, the matching degree calculation and analysis of the prediction curve and actual output curve of each distributed power generation in the virtual power plant are carried out to obtain the matching degree.

The day-ahead scheduling is selected as the planning and scheduling cycle of the virtual power plant, and the daily matching variance is defined on the scale of hours, that is, the matching degree between the daily power generation sum of the distributed power generation and the daily expected power generation curve. Obtain the optimal combination scheme from the combination of distributed power sources of different types and capacities. According to the characteristics of each distributed power source, when selecting the type of distributed power source in a virtual power plant, at least one dispatchable distributed power source should be included to improve the overall stability of the system.

Power generation plan constraints: The power generation plan submitted by the virtual power plant to the grid must be equal to the charging and discharging output of distributed power sources and batteries other than the internal load demand of the virtual power plant.

Controllable distributed power output upper and lower limit constraints: controllable distributed power output must be between the upper and lower limits of output.

Battery charge and discharge power upper and lower limit constraints: It is characterized in that the battery charge and discharge power must be bounded by the upper and lower limits of the charge and discharge power, and the storage capacity of the battery must also be bounded between the upper and lower limits of the storage capacity.

Claims (6)

1. A virtual power plant combined optimization method considering the volatility of distributed power sources, is characterized in that, comprising the following steps: (1) According to the geographical location, environmental conditions and resource distribution of different regions, a variety of factors are considered to reasonably select distributed power sources, and the output of each distributed power source is predicted; (2) According to the calculation formula of matching degree, calculate the matching degree of each distributed power generation's predicted output and the power generation and consumption plan of the virtual power plant; (3) Incorporate each distributed power source when the minimum matching degree is obtained into the virtual power plant; (4) Judging whether the virtual power plant formed by distributed power meets the scheduling needs of the system; (5) Form a virtual power plant. 2. A kind of virtual power plant combination optimization method considering the fluctuation of distributed power sources according to claim 1, characterized in that, in the step (2), each distributed power source predicts output and virtual power plant power generation plan The matching degree of is calculated by the following formula: The specific formula for calculating the matching degree in the period t is: In the formula, S _t represents the matching degree of time period t, N is the type of distributed power generation, i is the number of distributed power generation, P _i ^t is the predicted output of distributed power generation i in time period t, and P _e is the power generation and consumption plan of the virtual power plant ; The matching degree of distributed power generation based on renewable resources in the entire scheduling period (1,2,...t,...T) is: In the formula, S represents the matching degree of the entire scheduling cycle, T is the time period selected by the scheduling cycle, t is the sequence number of the time period, N is the type of distributed power supply, i is the number of distributed power supply, P _i ^t is the distributed power supply i Forecast output in period t, P _e is the planned electricity generation and consumption of virtual power plant. 3. a kind of virtual power plant combination optimization method considering the fluctuation of distributed power supply according to claim 1, is characterized in that, the target of the combination optimization model of each distributed power supply during minimum matching degree in described step (3) The function is: Among them, T is the time period selected by the scheduling cycle; t is the number of the time period; N is the type of distributed power; i is the type number of distributed power; P _i ^t is the predicted output of distributed power _i at time t; power generation and consumption plan of the power plant; u _i represents the 0-1 variable of the state of the distributed generation, u _i = ₁ indicates that the distributed Power generation unit of a power plant. 4. A virtual power plant combination optimization method considering the volatility of distributed power sources according to claim 3, characterized in that: according to demand, select day-ahead, monthly, quarterly or annual scheduling as the virtual power plant planning and scheduling cycle, corresponding S is daily matching degree, monthly matching degree, quarterly matching degree or annual matching degree. 5. A kind of virtual power plant combination optimization method considering the volatility of distributed power sources according to claim 1, characterized in that, the scheduling in the step (4) needs to be a constraint condition, including: (41) Power generation plan constraints: in, Respectively represent the charge and discharge power of the battery at time t; are the charge and discharge efficiencies of the battery, respectively; is the load demand of the virtual power plant at time t, is the power generation plan submitted by the virtual power plant to the grid; u _i represents the 0-1 variable of the state of the distributed power generation, u _i = 1 means that the distributed power source i is the power generation unit of the virtual power plant, and u _i = 0 means that the distributed power source i Not as a power generation unit of a virtual power plant; P _i ^t is the predicted output of distributed power i at time t; (42) Controllable distributed power output upper and lower limit constraints: in, Indicates the minimum output of the controllable distributed power supply, Indicates the maximum output of the controllable distributed power supply; (43) The upper and lower limits of battery charging and discharging power: SOC _min ≤ SOC ^t ≤ SOC _max (9); in, Respectively, the charging and discharging power of the battery at time t; Respectively, the minimum charge and discharge power of the battery; are the maximum charging and discharging power of the battery; SOC ^t is the storage capacity of the battery at time t, SOC _min is the minimum storage capacity of the battery, and SOC _max is the maximum storage capacity of the battery; If the scheduling requirements are met, select the above-mentioned distributed power sources, and perform step (5) to form a virtual power plant; if the scheduling requirements are not met, return to step (3) to adjust the selected distributed power sources until the scheduling requirements are met. 6. A combined optimization model of a virtual power plant considering the volatility of distributed power sources, characterized in that the model includes an objective function and constraints, and the objective function is: Among them, T is the time period selected by the scheduling cycle; t is the number of the time period; N is the type of distributed power; i is the type number of distributed power; P _i ^t is the predicted output of distributed power _i at time t; power generation and consumption plan of the power plant; u _i represents the 0-1 variable of the state of the distributed generation, u _i = ₁ indicates that the distributed power generation units of power plants; The constraints include: (a) Generation plan constraints: in, Respectively represent the charge and discharge power of the battery at time t; are the charge and discharge efficiencies of the battery, respectively; is the load demand of the virtual power plant at time t, is the power generation plan submitted by the virtual power plant to the grid; u _i represents the 0-1 variable of the state of the distributed power generation, u _i = 1 means that the distributed power source i is the power generation unit of the virtual power plant, and u _i = 0 means that the distributed power source i Not as a power generation unit of a virtual power plant; P _i ^t is the predicted output of distributed power i at time t; (b) Controllable distributed power output upper and lower limit constraints: in, Indicates the minimum output of the controllable distributed power supply, Indicates the maximum output of the controllable distributed power supply; (c) The upper and lower limits of battery charging and discharging power: SOC _min ≤ SOC ^t ≤ SOC _max (9); in, Respectively, the charging and discharging power of the battery at time t; Respectively, the minimum charge and discharge power of the battery; are the maximum charging and discharging power of the battery; SOC ^t is the storage capacity of the battery at time t, SOC _min is the minimum storage capacity of the battery, and SOC _max is the maximum storage capacity of the battery.