CN112101987A - Multi-microgrid random prediction control method - Google Patents

Abstract

A multi-microgrid random prediction control method is used for modeling the uncertainty of renewable energy power generation and electric heating load by combining probability constraint on electricity and gas purchase quantity in a system. And decomposing the joint probability constraint with coupling into M chance constraints by adopting a decomposition method, and replacing the chance constraints with deterministic constraints considering uncertain parameter statistical characteristics by adopting chance constraint prediction control. In the energy management control strategy, the influence of uncertain renewable energy sources and uncertain load in the system on the system is considered in a mode of joint probability constraint of system electricity sale and gas purchase. Meanwhile, as the permeability of renewable energy sources in the microgrid is higher and higher, the strategy can enable the behavior of the interconnected microgrid to be more predictable from the perspective of the main grid, so that the complexity of an energy management system is reduced.

Description

A Stochastic Predictive Control Method for Multiple Microgrids

technical field

The invention relates to a multi-microgrid random prediction control method.

Background technique

Model Predictive Control (MPC) is one of the most successful control strategies in the power system community. MPC can take into account the characteristics and operational constraints of the system, and consider future predictions of system behavior and closed-loop control strategies, making it an attractive control strategy for power system applications.

In the existing inventions about energy management strategies for multiple microgrids, there are the following problems: (1) the existing inventions are mainly aimed at microgrids with electricity as a single energy source, and less consideration is given to microgrids containing multiple energy forms (2) ) did not analyze the uncertainty of the renewable energy output and the thermal power load in the system and did not take them into account in the proposed algorithm.

Wind power generation and photovoltaic power generation have a high degree of uncertainty, which seriously affects the actual operation of the microgrid. At the same time, the electrical load and thermal load in the system also have certain uncertainties due to the influence of weather and human factors.

SUMMARY OF THE INVENTION

In order to overcome the deficiencies of the prior art, the present invention proposes an energy management method for coordinating multiple micro-grids on the basis of a stochastic predictive control framework. The operating cost of the network is minimal.

The technical scheme adopted by the present invention to solve its technical problems is:

A multi-microgrid random prediction control method, comprising the following steps:

1) Establish a multi-microgrid system framework

The operation and management of multi-microgrids are coordinated under the stochastic predictive control framework. The multi-microgrid system contains M heterogeneous microgrids. Equations (1)-(4) are the dynamic equations of each microgrid:

和

分别为t时刻微网i的购电和售电功率；P_WT,i(t)为t时刻微网i中风力发电功率；P_PV,i(t)为t时刻微网i中光伏发电功率；

为t时刻微网i中燃料电池发电功率；

为t时刻微网i中热泵热功率；η_HP,i微网i中热泵产热效率；

和

分别t时刻微网i的电负荷和热负荷；

为t时刻微网i中燃气锅炉热功率；

为t时刻微网i的购气量；LHV为天然气的低热值，取9.7kWh/Nm³；η_FC,i与η_GB,i分别为微网i中燃料电池发电效率和燃气锅炉热效率；where _zi (t) is the state of charge (SOC) of the energy storage system in microgrid i at time t; P _ESS,i (t) is the charging and discharging power of the energy storage system in microgrid i at time t; C _ESS,i is the charge-discharge coefficient of the energy storage system in the microgrid i;

and

are the purchase and sale power of microgrid i at time t, respectively; _PWT,i (t) is the wind power generation in microgrid i at time t; P _PV,i (t) is the photovoltaic power generation in microgrid i at time t;

is the power generated by the fuel cell in the microgrid i at time t;

is the heat pump heat power in the microgrid i at time t; η _HP,i is the heat production efficiency of the heat pump in the microgrid i;

and

are the electrical load and thermal load of microgrid i at time t, respectively;

is the thermal power of the gas boiler in the microgrid i at time t;

is the gas purchase volume of microgrid i at time t; LHV is the low calorific value of natural gas, which is taken as 9.7kWh/Nm ³ ; ηFC _,i and ηGB _,i are the power generation efficiency of the fuel cell and the thermal efficiency of the gas boiler in the microgrid i, respectively;

According to (2), it can be seen that in the case of power shortage, the microgrid can buy the required power from the main grid and sell energy to the grid in the case of power surplus, but simultaneous buying and selling is not allowed;

The total operating cost of each microgrid is shown in formula (5):

为微网i总运行成本；cost_i(t)为微网i购买电以及购气所花成本；

为微网i中设备运行成本；λ_pur、λ_sell和λ_gas分别为购买电价格以及购气价格；λ_ESS,i、λ_SOFC,i、λ_HP,i和λ_GB,i分别为微网i中各装置维护成本价格系数；where:

is the total operating cost of microgrid i; cost _i (t) is the cost of purchasing electricity and gas for microgrid i;

is the operating cost of equipment in the microgrid i; λ _pur , λ _sell and λ _gas are the purchase price of electricity and gas, respectively; λ _ESS,i , λ _SOFC,i , λ _HP,i and λ _GB,i are the microgrid, respectively The maintenance cost price coefficient of each device in i;

For the security of the system, the following constraints also need to be satisfied:

分别为微网i中储能系统荷电状态上下限；

分别为微网i中储能系统充放电功率上下限；

分别为微网i中燃料电池发电功率上下限；

分别为微网i中热泵热功率上下限；

分别为微网i中燃气锅炉热功率上下限；

为微网i购电功率上限；

为微网i卖电功率上限；

为微网i购气量上限；where:

are the upper and lower limits of the state of charge of the energy storage system in microgrid i, respectively;

are the upper and lower limits of the charging and discharging power of the energy storage system in the microgrid i, respectively;

are the upper and lower limits of fuel cell power generation in microgrid i, respectively;

are the upper and lower limits of heat pump thermal power in microgrid i, respectively;

are the upper and lower limits of the thermal power of the gas boiler in the microgrid i, respectively;

The upper limit of purchasing power for microgrid i;

Selling power cap for microgrid i;

It is the upper limit of gas purchase volume for microgrid i;

Considering the uncertainty of microgrid energy transaction, a joint probability constraint is carried out on microgrid energy transaction, where P is the probability of event occurrence; 1-ρ is a preset confidence level;

where M is the number of microgrids;

According to the joint constraint (14), the probability that the energy exchange constraints of all microgrids in the system satisfy the requirement is greater than 1-ρ, but at the same time, this constraint also makes the operation of each microgrid mutually coupled; a decomposition method is used to convert the random constraints into For M chance constraints and M new coupling constraints, such as equations (15) and (16), in a coordinated multi-microgrid system, the energy management center assigns risk parameters to each microgrid by minimizing the total cost of the system;

where: σ _i is the risk coefficient of each microgrid;

2) Chance-constrained predictive control

In MPC, the dynamic prediction model of the system is considered, a constrained optimization problem is solved within the set control range, and a series of optimal control actions are obtained; however, the system only executes the optimal control sequence obtained. The first sample ignores the remaining samples. In the next step, the control system performs the entire optimization process considering the latest information of the system state and parameters. This inherent feedback mechanism brings robustness to the algorithm, making it Be an appropriate tool for decision-making under conditions of uncertainty;

In CCMPC, in addition to deterministic constraints, the optimization problem can also contain some probability (chance) constraints such as equation (17). In this case, the most common solution strategy is to replace chance constraints with deterministic constraints that take into account the statistical properties of uncertain parameters;

P{x ^min ≤x(t)≤x ^max }≥1-ρ (17)

Assuming that the uncertain parameter x(t) obeys the Gaussian probability density function, that is, x(t)～N(m, a ² ), m is the expected value, and a ² is the variance, then formula (17) is transformed into formula (18) ( 19):

Using the property of cumulative distribution function, the deterministic constraint of Eq. (17) is obtained:

m≤x ^max -aφ ^-1 (1-ρ) (20)

m≥x ^min +aφ ^-1 (1-ρ) (21)

where: φ( ) represents the cumulative distribution function of a standard normal variable with zero mean and unit variance;

When the probability density function of the uncertain parameters is unknown, use the Chershev inequality to derive deterministic constraints such as equations (22), (23):

3) Energy management methods

其中

和

为两个变量的期望值，∑_μ,i和∑_β,i为两个变量的方差；Under the framework of CCMPC, the problem of coordinated operation and management of microgrids with joint constraints is studied. The intermittency of wind turbines and photovoltaic power generation and the variability of loads are used as different sources of uncertainty, and two new variables μ _i (t), β _i (t), assuming that μ _i (t), β (t) obey the normal distribution probability density function, namely

in

and

is the expected value of the two variables, ∑ _μ,i and ∑ _β,i are the variances of the two variables;

Taking into account the linear relationship of equations (2) to (4), the expectation and covariance of the power sold and purchased and the amount of natural gas purchased are expressed as:

和

为t时刻微网i购电功率和卖电功率期望值；

和

为t时刻微网i购电功率和卖电功率方差；

为t时刻微网i购气量期望值；

为t时刻微网i购气量方差；where:

and

is the expected value of purchasing power and selling power of microgrid i at time t;

and

is the variance of purchasing power and selling power of microgrid i at time t;

is the expected value of gas purchase volume of microgrid i at time t;

is the variance of gas purchase volume of microgrid i at time t;

Finally, the proposed coupled probability constrained microgrid energy management problem based on CCMPC is formulated as follows:

s.t.(1),(6)-(13),(16),(26)-(30)(32)

Where: H _p is the operating cycle of the system.

At the beginning of each sampling interval, the energy management center of the multi-microgrid system will collect the required information from each microgrid and solve the above optimization problem. The control sequence is then obtained and communicated with the microgrid local controller for implementation in the associated microgrid. The control sequence information sent by the energy management center includes energy storage charge and discharge power, fuel cell power generation power, gas boiler heat generation power, heat pump heat generation power, electricity purchases and sales from the main network, and gas purchases.

The present invention models the uncertainty of renewable energy power generation and electric heating load by constraining the combined probability of electricity and gas purchases in the system. The coupled joint probability constraints are decomposed into M chance constraints by the decomposition method, and the chance constraints are replaced by the deterministic constraints considering the statistical characteristics of uncertain parameters by the chance constraint predictive control.

Aiming at the micro-grid including multiple energy forms, the present invention researches on the cooperative operation management of the micro-grid under the framework of chance-constrained MPC (chance-constrained MPC, CCMPC). A joint probability constraint is proposed to ensure that the real-time power exchange between the microgrid and the main grid and the gas purchase from the gas company do not violate the expected operating range of the pre-set confidence level, so as to realize the economical optimization of the system operating cost.

First, the local controller in each microgrid will forecast the renewable energy and load demand in the cycle, and send the forecast information to the upper-layer microgrid energy management center. The energy management center calculates the system operation risk and obtains the power scheduling scheme according to formula (33), and sends the information to the local controllers in each microgrid. The local controller will execute the first sample of the optimal control sequence and update the renewable energy output and load demand information in the system in preparation for the next cycle. The invention is divided into three parts: (1) establishment of multi-microgrid system framework; (2) chance constraint prediction control method; (3) energy management strategy.

The beneficial effects of the present invention are mainly manifested in that: in the proposed energy management control strategy, in the form of a combined probability constraint on the amount of electricity sold and purchased by the system and the amount of gas purchased, the influence of the uncertainty of renewable energy in the system and the uncertainty of load on the system is considered. At the same time, due to the increasing penetration of renewable energy in microgrids, from the perspective of the main network, this strategy can make the behavior of interconnected microgrids more predictable, thereby reducing the complexity of the energy management system.

Considering the increasing penetration of renewable energy-based microgrids into the power system, this support strategy could make the behavior of connected microgrids more predictable from the main grid perspective, thereby reducing the complexity of EMSs.

Description of drawings

Figure 1 is a flow chart of a multi-microgrid stochastic predictive control method.

Detailed ways

The present invention will be further described below in conjunction with the accompanying drawings.

Referring to Fig. 1, a multi-microgrid stochastic predictive control method, comprising the following steps:

1) Establish a multi-microgrid system framework

The operation and management of multi-microgrids are coordinated under the stochastic predictive control framework. The multi-microgrid system contains M heterogeneous microgrids. Equations (1)-(4) are the dynamic equations of each microgrid:

和

分别为t时刻微网i的购电和售电功率；P_WT,i(t)为t时刻微网i中风力发电功率；P_PV,i(t)为t时刻微网i中光伏发电功率；

为t时刻微网i中燃料电池发电功率；

为t时刻微网i中热泵热功率；η_HP,i微网i中热泵产热效率；

和

分别t时刻微网i的电负荷和热负荷；

为t时刻微网i中燃气锅炉热功率；

为t时刻微网i的购气量；LHV为天然气的低热值，取9.7kWh/Nm³；η_FC,i与η_GB,i分别为微网i中燃料电池发电效率和燃气锅炉热效率；where _zi (t) is the state of charge (SOC) of the energy storage system in microgrid i at time t; P _ESS,i (t) is the charging and discharging power of the energy storage system in microgrid i at time t; C _ESS,i is the charge-discharge coefficient of the energy storage system in the microgrid i;

and

are the purchase and sale power of microgrid i at time t, respectively; _PWT,i (t) is the wind power generation in microgrid i at time t; P _PV,i (t) is the photovoltaic power generation in microgrid i at time t;

is the power generated by the fuel cell in the microgrid i at time t;

is the heat pump heat power in the microgrid i at time t; η _HP,i is the heat production efficiency of the heat pump in the microgrid i;

and

are the electrical load and thermal load of microgrid i at time t, respectively;

is the thermal power of the gas boiler in the microgrid i at time t;

is the gas purchase volume of microgrid i at time t; LHV is the low calorific value of natural gas, which is taken as 9.7kWh/Nm ³ ; ηFC _,i and ηGB _,i are the power generation efficiency of the fuel cell and the thermal efficiency of the gas boiler in the microgrid i, respectively;

According to (2), it can be seen that in the case of power shortage, the microgrid can buy the required power from the main grid and sell energy to the grid in the case of power surplus, but simultaneous buying and selling is not allowed;

The total operating cost of each microgrid is shown in formula (5):

为微网i总运行成本；cost_i(t)为微网i购买电以及购气所花成本；

为微网i中设备运行成本；λ_pur、λ_sell和λ_gas分别为购买电价格以及购气价格；λ_ESS,i、λ_SOFC,i、λ_HP,i和λ_GB,i分别为微网i中各装置维护成本价格系数；where:

is the total operating cost of microgrid i; cost _i (t) is the cost of purchasing electricity and gas for microgrid i;

is the operating cost of equipment in the microgrid i; λ _pur , λ _sell and λ _gas are the purchase price of electricity and gas, respectively; λ _ESS,i , λ _SOFC,i , λ _HP,i and λ _GB,i are the microgrid, respectively The maintenance cost price coefficient of each device in i;

For the security of the system, the following constraints also need to be satisfied:

分别为微网i中储能系统荷电状态上下限；

分别为微网i中储能系统充放电功率上下限；

分别为微网i中燃料电池发电功率上下限；

分别为微网i中热泵热功率上下限；

分别为微网i中燃气锅炉热功率上下限；

为微网i购电功率上限；

为微网i卖电功率上限；

为微网i购气量上限；where:

are the upper and lower limits of the state of charge of the energy storage system in microgrid i, respectively;

are the upper and lower limits of the charging and discharging power of the energy storage system in the microgrid i, respectively;

are the upper and lower limits of fuel cell power generation in microgrid i, respectively;

are the upper and lower limits of heat pump thermal power in microgrid i, respectively;

are the upper and lower limits of the thermal power of the gas boiler in the microgrid i, respectively;

The upper limit of purchasing power for microgrid i;

Selling power cap for microgrid i;

It is the upper limit of gas purchase volume for microgrid i;

Considering the uncertainty of microgrid energy transaction, a joint probability constraint is carried out on microgrid energy transaction, where P is the probability of event occurrence; 1-ρ is a preset confidence level;

where M is the number of microgrids;

According to the joint constraint (14), the probability that the energy exchange constraints of all microgrids in the system satisfy the requirement is greater than 1-ρ, but at the same time, this constraint also makes the operation of each microgrid mutually coupled; a decomposition method is used to convert the random constraints into For M chance constraints and M new coupling constraints, such as equations (15) and (16), in a coordinated multi-microgrid system, the energy management center assigns risk parameters to each microgrid by minimizing the total cost of the system;

where: σ _i is the risk coefficient of each microgrid;

2) Chance-constrained predictive control

In MPC, the dynamic prediction model of the system is considered, a constrained optimization problem is solved within the set control range, and a series of optimal control actions are obtained; however, the system only executes the optimal control sequence obtained. The first sample ignores the remaining samples. In the next step, the control system performs the entire optimization process considering the latest information of the system state and parameters. This inherent feedback mechanism brings robustness to the algorithm, making it Be an appropriate tool for decision-making under conditions of uncertainty;

In CCMPC, in addition to deterministic constraints, the optimization problem can also contain some probability (chance) constraints such as equation (17). In this case, the most common solution strategy is to replace chance constraints with deterministic constraints that take into account the statistical properties of uncertain parameters;

P{x ^min ≤x(t)≤x ^max }≥1-ρ (17)

Assuming that the uncertain parameter x(t) obeys the Gaussian probability density function, that is, x(t)～N(m, a ² ), m is the expected value, and a ² is the variance, then formula (17) is transformed into formula (18) ( 19):

Using the property of cumulative distribution function, the deterministic constraint of Eq. (17) is obtained:

m≤x ^max -aφ ^-1 (1-ρ) (20)

m≥x ^min +aφ ^-1 (1-ρ) (21)

where: φ( ) represents the cumulative distribution function of a standard normal variable with zero mean and unit variance;

When the probability density function of the uncertain parameters is unknown, use the Chershev inequality to derive deterministic constraints such as equations (22), (23):

3) Energy management methods

其中

和

为两个变量的期望值，∑_μ,i和∑_β,i为两个变量的方差；Under the framework of CCMPC, the problem of coordinated operation and management of microgrids with joint constraints is studied. The intermittency of wind turbines and photovoltaic power generation and the variability of loads are used as different sources of uncertainty, and two new variables μ _i (t), β _i (t), assuming that μ _i (t), β (t) obey the normal distribution probability density function, namely

in

and

is the expected value of the two variables, ∑ _μ,i and ∑ _β,i are the variances of the two variables;

Taking into account the linear relationship of equations (2) to (4), the expectation and covariance of the power sold and purchased and the amount of natural gas purchased are expressed as:

和

为t时刻微网i购电功率和卖电功率期望值；

和

为t时刻微网i购电功率和卖电功率方差；

为t时刻微网i购气量期望值；

为t时刻微网i购气量方差；where:

and

is the expected value of purchasing power and selling power of microgrid i at time t;

and

is the variance of purchasing power and selling power of microgrid i at time t;

is the expected value of gas purchase volume of microgrid i at time t;

is the variance of gas purchase volume of microgrid i at time t;

Finally, the proposed coupled probability constrained microgrid energy management problem based on CCMPC is formulated as follows:

s.t.(1),(6)-(13),(16),(26)-(30) (32)

Where: H _p is the operating cycle of the system.

At the beginning of each sampling interval, the energy management center of the multi-microgrid system will collect the required information from each microgrid and solve the above optimization problem. The control sequence is then obtained and communicated with the microgrid local controller for implementation in the associated microgrid. The control sequence information sent by the energy management center includes energy storage charge and discharge power, fuel cell power generation power, gas boiler heat generation power, heat pump heat generation power, electricity purchases and sales from the main network, and gas purchases.

This embodiment studies the collaborative operation management of the microgrid under the framework of CCMPC for a microgrid that includes multiple energy forms. A joint probability constraint is proposed to ensure that the real-time power exchange between the microgrid and the main grid and the amount of gas purchased from the gas company do not violate the expected operating range of the pre-set confidence level.

Claims (1)

1. a multi-microgrid random prediction control method, is characterized in that, described method comprises the following steps: 1) Establish a multi-microgrid system framework The operation and management of multi-microgrids are coordinated under the stochastic predictive control framework. The multi-microgrid system contains M heterogeneous microgrids. Equations (1)-(4) are the dynamic equations of each microgrid:

where _zi (t) is the state of charge (SOC) of the energy storage system in microgrid i at time t; P _ESS,i (t) is the charging and discharging power of the energy storage system in microgrid i at time t; C _ESS,i is the charge-discharge coefficient of the energy storage system in the microgrid i;

and

are the purchase and sale power of microgrid i at time t, respectively; _PWT,i (t) is the wind power generation in microgrid i at time t; P _PV,i (t) is the photovoltaic power generation in microgrid i at time t;

is the power generated by the fuel cell in the microgrid i at time t;

is the heat pump heat power in the microgrid i at time t; η _HP,i is the heat production efficiency of the heat pump in the microgrid i;

and

are the electrical load and thermal load of microgrid i at time t, respectively;

is the thermal power of the gas boiler in the microgrid i at time t;

is the gas purchase volume of microgrid i at time t; LHV is the low calorific value of natural gas, which is taken as 9.7kWh/Nm ³ ; ηFC _,i and ηGB _,i are the power generation efficiency of the fuel cell and the thermal efficiency of the gas boiler in the microgrid i, respectively; According to (2), it can be seen that in the case of power shortage, the microgrid can buy the required power from the main grid and sell energy to the grid in the case of power surplus, but simultaneous buying and selling is not allowed; The total operating cost of each microgrid is shown in formula (5):

where:

is the total operating cost of microgrid i; cost _i (t) is the cost of purchasing electricity and gas for microgrid i;

is the operating cost of equipment in the microgrid i; λ _pur , λ _sell and λ _gas are the purchase price of electricity and gas, respectively; λ _ESS,i , λ _SOFC,i , λ _HP,i and λ _GB,i are the microgrid, respectively The maintenance cost price coefficient of each device in i; For the security of the system, the following constraints also need to be satisfied:

where:

are the upper and lower limits of the state of charge of the energy storage system in microgrid i, respectively;

are the upper and lower limits of the charging and discharging power of the energy storage system in the microgrid i, respectively;

are the upper and lower limits of fuel cell power generation in microgrid i, respectively;

are the upper and lower limits of heat pump thermal power in microgrid i, respectively;

are the upper and lower limits of the thermal power of the gas boiler in the microgrid i, respectively;

The upper limit of purchasing power for microgrid i;

Selling power cap for microgrid i;

It is the upper limit of gas purchase volume for microgrid i; Considering the uncertainty of microgrid energy transaction, a joint probability constraint is carried out on microgrid energy transaction, where P is the probability of event occurrence; 1-ρ is a preset confidence level;

where M is the number of microgrids; According to the joint constraint (14), the probability that the energy exchange constraints of all microgrids in the system satisfy the requirement is greater than 1-ρ, but at the same time, this constraint also makes the operation of each microgrid mutually coupled; a decomposition method is used to convert the random constraints into For M chance constraints and M new coupling constraints, such as equations (15) and (16), in a coordinated multi-microgrid system, the energy management center assigns risk parameters to each microgrid by minimizing the total cost of the system;

where: σ _i is the risk coefficient of each microgrid; 2) Chance-constrained predictive control Replace chance constraints with deterministic constraints that consider the statistical properties of uncertain parameters; P{x ^min ≤x(t)≤x ^max }≥1-ρ (17) Assuming that the uncertain parameter x(t) obeys the Gaussian probability density function, that is, x(t)～N(m, a ² ), m is the expected value, and a ² is the variance, then formula (17) is transformed into formula (18) ( 19):

Using the property of cumulative distribution function, the deterministic constraint of Eq. (17) is obtained: m≤x ^max -aφ ^-1 (1-ρ) (20) m≥x ^min +aφ ^-1 (1-ρ) (21) where: φ( ) represents the cumulative distribution function of a standard normal variable with zero mean and unit variance; When the probability density function of the uncertain parameters is unknown, use the Chershev inequality to derive deterministic constraints such as equations (22), (23):

3) Energy management methods Under the framework of CCMPC, the problem of coordinated operation and management of microgrids with joint constraints is studied. The intermittency of wind turbines and photovoltaic power generation and the variability of loads are used as different sources of uncertainty, and two new variables μ _i (t), β _i (t), assuming that μ _i (t), β (t) obey the normal distribution probability density function, namely

in

and

is the expected value of the two variables, ∑ _μ,i and ∑ _β,i are the variances of the two variables;

Taking into account the linear relationship of equations (2) to (4), the expectation and covariance of the power sold and purchased and the amount of natural gas purchased are expressed as:

where:

and

is the expected value of purchasing power and selling power of microgrid i at time t;

and

is the variance of purchasing power and selling power of microgrid i at time t;

is the expected value of gas purchase volume of microgrid i at time t;

is the variance of gas purchase volume of microgrid i at time t; Finally, the proposed coupled probability constrained microgrid energy management problem based on CCMPC is formulated as follows:

s.t.(1),(6)-(13),(16),(26)-(30) (32)

Where: H _p is the operating cycle of the system.

Images