A distributed energy storage operation optimization method for supply and demand balance

Abstract

The present invention belongs to the field of electric power, and discloses a distributed energy storage operation optimization method for supply and demand balance, including: step 1, defining the active power balance relationship within the power grid; step 2, estimating the load active power online; step 3, calculating the power actually stored and extracted by each battery energy storage system; step 4, considering relevant power constraints, constructing an optimization problem model of a distributed battery energy storage system; step 5, solving the optimization problem of step 4, and designing a coordinated operation optimization algorithm for a distributed battery energy storage system. The present invention proposes a distributed method to control the battery energy storage system to operate in a charging mode when the renewable power generation is greater than the demand, and to operate in a discharging mode when the renewable power generation is less than the demand, thereby improving the overall utilization efficiency of renewable energy power generation.

Description

Technical Field

[0001] The present invention relates to the field of electric power, and in particular to a method for optimizing distributed operation of energy storage for supply and demand balance. Background Art

[0002] Research and development of smart distribution networks can promote the large-scale access of distributed power sources and renewable energy, and achieve highly reliable supply of different energy sources. As abundant and clean energy sources, wind and solar energy have become important alternatives for renewable electricity supply. However, wind and solar energy have intermittent power supply problems, which poses important challenges to the operation and control of smart distribution networks. In addition, when the power generation of renewable energy is not equal to the total load demand, it will cause an imbalance between supply and demand. In short, intermittent renewable energy generation and frequently changing load demand will put tremendous pressure on the supply and demand balance of smart distribution networks.

[0003] There are three main approaches to address the above problems: (1) Adjust supply-side management to reduce the active power generation of renewable energy generators (REGs) or set REGs to unloading mode. However, this approach reduces the utilization efficiency of renewable energy and is ineffective when the maximum power generation of renewable energy is less than the load demand. (2) Adjust demand-side management to adjust load demand based on available power generation. This approach requires users to respond quickly to changes in energy prices or demand changes, which may cause user dissatisfaction. (3) Install and operate energy storage systems (ESS), such as batteries. ESS can collect and store excess energy that is not used for a period of time in some way, and can play an important role in energy distribution. As a buffer, ESS can reduce or even eliminate the problem of uneven power distribution. When power generation is greater than demand, ESS can store excess power resources. During peak power consumption, ESS compensates for insufficient power by releasing energy. Energy management based on ESS is crucial to the control and operation of smart distribution networks. If the management configuration of ESS is unreasonable, it will cause energy waste and increase system costs. However, the management configuration of ESS is closely related to the load level, while the actual load level of users is not always consistent with the preset load, which brings challenges to the design of distributed control and optimization methods of ESS.

[0004] In order to solve the above problems, the present invention aims at the coordinated control problem of multiple battery energy storage systems (BESS) and proposes a new distributed energy storage system control scheme, which can maintain the balance of active power when the user load measurement is inaccurate and minimize the total active power loss related to the charging / discharging efficiency of the battery energy storage system. Summary of the invention

[0005] The purpose of the present invention is to provide a distributed energy storage operation optimization method for supply and demand balance to solve the problems raised in the above background technology.

[0006] To achieve the above object, the present invention provides the following technical solutions:

[0007] A distributed energy storage operation optimization method for supply and demand balance, comprising:

[0008] Step 1, defining the active power balance relationship in the power grid;

[0009] Step 2, online estimation of load active power;

[0010] Step 3, calculating the power actually stored and extracted by each battery energy storage system;

[0011] Step 4: Considering relevant power constraints, construct an optimization problem model of the distributed battery energy storage system;

[0012] Step 5, solve the optimization problem in step 4 and design an optimization algorithm for the coordinated operation of the distributed battery energy storage system.

[0013] Furthermore, the step 1 comprises:

[0014] For smart distribution grids that include battery energy storage systems, energy storage devices can be regarded as elastic loads to adjust the balance between power supply and consumption. As shown in the following formula, the active power balance in the grid can be expressed as:

[0015]

[0016] in, , , Represents the collection of distributed generators, loads and battery energy storage systems, Indicates The power generated by the generator, represents the power consumed by the jth load, is the charging / discharging power of the ith battery energy storage system in positive / negative conditions.

[0017] Furthermore, the step 2 comprises:

[0018] Step 2.1, using the measurements with complex noise provided by the SCADA system and the phasor measurement unit to achieve robust estimation of the node voltage amplitude and phase angle;

[0019] Step 2.2, using the power flow analysis method, we know that when the voltage amplitude and phase angle of the nodes at both ends of the branch are known, the active power of the branch can be calculated as follows:

[0020]

[0021] in, represents the active power of the branch between load node j and node m, and are the series conductance and susceptance, , and are the voltage phase angle and amplitude at node j, and are the voltage phase angle and amplitude at node m respectively;

[0022] Step 2.3, calculate the estimated total power consumed by load node j , as shown below:

[0023] .

[0024] Furthermore, the step 3 comprises:

[0025] Step 3.1, transmission loss is modeled by multiplying the load active power by a factor of 0.05;

[0026] Step 3.2: The total net active power required by the battery energy storage system in the grid Calculated as:

[0027]

[0028] if is positive, the battery energy storage system will be controlled to charge. If the value is negative, the battery energy storage system will be controlled in the discharge mode. In order to ensure the balance of supply and demand in the power grid, it is necessary to design a suitable dispatching strategy to share the energy between distributed ESS. , which can be achieved by controlling the battery energy storage system The discharge / charge power is achieved by:

[0029]

[0030] Step 3.3: Due to the internal resistance of the battery, power loss will occur when the battery energy storage system is charged / discharged. and They represent the actual stored power and extracted power of the ith battery energy storage system, respectively, and are calculated as follows:

[0031]

[0032]

[0033] in, and Represent discharge and discharge efficiency respectively.

[0034] Furthermore, the step 4 comprises:

[0035] The total real power input into the battery energy storage system can be calculated as:

[0036]

[0037] The charging rate and the SoC of the battery energy storage system will affect the charging efficiency. The charging efficiency and charging rate are linearly related, and the formula is:

[0038]

[0039] in and are two constants representing the inverse of the intercept and slope of the linear relationship;

[0040] By substituting formula (9) into formula (8), the objective function becomes:

[0041]

[0042] Where Min is the minimization function;

[0043] A distributed battery energy storage system collaborative control strategy is designed to maximize the total input power of the battery energy storage system by reducing power loss. Considering the relevant constraints, the following optimization problem model is obtained:

[0044] (11)

[0045] in, and Represent the lower and upper charging limits respectively.

[0046] Furthermore, the step 5 comprises:

[0047] Step 5.1: Use the Lagrangian algorithm to solve the optimization problem obtained in step 4. According to formulas (10) and (11), the Lagrangian function L can be obtained as follows:

[0048]

[0049] in, is the Lagrange multiplier;

[0050] Step 5.2, calculate the dual function, which is the Lagrangian function in The minimum value on :

[0051]

[0052] Where inf means finding the infimum of a set, and ,satisfy:

[0053]

[0054] (14)

[0055] Step 5.3, Lagrange multipliers Update along the negative gradient direction, the update equation is:

[0056]

[0057] In the formula, represents the update step size, represents the partial derivative of the function D with respect to v, and N represents the total number of nodes in the network;

[0058] Step 5.4, consider the complex communication environment where there is delay in node communication, assuming that node With Node The communication delay at the kth iteration is , without loss of generality, the delay satisfies:

[0059] (16)

[0060] In the formula, represents the upper bound of the corresponding delay;

[0061] Step 5.5: Each battery energy storage system control unit can be abstracted as an intelligent agent with computing power and weak communication capabilities. Represented as a graph consisting of a set of vertices and a set of edges ; Representation Node Can receive nodes If there is a path from i to j, the corresponding two vertices are called connected; the set of nodes that can send information to node j is called the input neighbor node set of the node. ,Right now ;

[0062] The node update rate is set to:

[0063]

[0064] In the formula,

[0065] (18)

[0066] in, Represents the state information of node j; represents the weight coefficient; represents the upper bound of the communication delay; represents the delay; r is the time step of the delay, from 0 to the maximum delay ; represents an indicator function used to indicate whether communication with a corresponding delay has occurred;

[0067] Step 5.6, each node is set to contain two state information and , in order to obtain global information , the node initial value is set to , the status update rate is as follows:

[0068]

[0069]

[0070] make , then when hour, The limit satisfies the following formula:

[0071] .

[0072] Compared with the prior art, the present invention has the following beneficial effects:

[0073] 1) When the renewable power generation is greater than the total load demand, and when the renewable power generation is less than the total load demand, a distributed method is proposed to control the battery energy storage system to operate in charging mode when the renewable power generation is greater than the demand, and to operate in discharging mode when the renewable power generation is less than the demand, thereby improving the overall utilization efficiency of renewable energy power generation.

[0074] 2) The present invention considers the charging efficiency in the objective function and proposes a distributed control algorithm for the battery energy storage system. The algorithm can minimize the active power loss related to the charging efficiency while maintaining the balance between active power supply and demand and the supply and demand balance of the power grid. BRIEF DESCRIPTION OF THE DRAWINGS

[0075] Figure 1 The present invention is a flow chart of a distributed energy storage operation optimization method for supply and demand balance.

[0076] Figure 2 Flowchart of the optimization algorithm for the coordinated operation of distributed battery energy storage systems.

[0077] Figure 3 The following is an example diagram of network communication topology.

[0078] Figure 4 Schematic diagram of the convergence of the Lagrange multiplier v as the number of iterations increases.

[0079] Figure 5 Schematic diagram of the average consistency results of the battery energy storage system.

[0080] Figure 6 Schematic diagram of the total charging power of all battery energy storage systems.

[0081] Figure 7 Schematic diagram of the charging power for each battery energy storage system unit.

[0082] Figure 8 Schematic diagram of the total charging power of all battery energy storage systems under charging rate constraints.

[0083] Fig. 9 Schematic diagram of the charging power of each battery energy storage system unit under the constraint of charging rate. DETAILED DESCRIPTION

[0084] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0085] This invention aims at the supply and demand balance problem in the smart distribution network, considers the scenario where the user end is equipped with a multi-battery energy storage system, and proposes a new distributed energy storage system optimization operation scheme based on the robust estimation method, thereby achieving the distributed supply and demand balance of the distribution network. Figure 1 As shown, the specific steps include:

[0086] Step 1: Define the active power balance relationship in the power grid, including:

[0087] For smart distribution grids that include battery energy storage systems, energy storage devices can be regarded as elastic loads to adjust the balance between power supply and consumption. As shown in the following formula, the active power balance in the grid can be expressed as:

[0088]

[0089] in, , , Represents the collection of distributed generation (DG) units, loads and battery energy storage systems, Indicates The power generated by a generator is usually known. represents the power consumed by the jth load, is the charging / discharging power of the ith battery energy storage system in positive / negative conditions.

[0090] Step 2, online estimation of load active power, including:

[0091] Step 2.1, using the measurements with complex noise provided by the SCADA system and the phasor measurement unit (PMU), a robust estimation of the node voltage amplitude and phase angle is achieved.

[0092] Step 2.2, using the power flow analysis method, we know that when the voltage amplitude and phase angle of the nodes at both ends of the branch are known, the active power of the branch can be calculated as follows:

[0093] (2)

[0094] middle, represents the active power of the branch between load node j and node m, and are the series conductance and susceptance, , and are the voltage phase angle and amplitude at node j, and are the voltage phase angle and amplitude at node m respectively.

[0095] Step 2.3, calculate the estimated total power consumed by load node j , as shown below:

[0096] .

[0097] Step 3, calculate the actual stored and extracted power of each battery energy storage system, including:

[0098] In step 3.1, considering that the load power may deviate from the expected load level due to the user's autonomous electricity consumption behavior and be affected by the sensor measurement accuracy, the transmission loss can be modeled by multiplying the load active power by a coefficient of 0.05.

[0099] Step 3.2, due to the intermittent nature of renewable energy generation, if they are controlled in MPPT (Maximum PowerPoint Tracking, MPPT) mode, they cannot be considered dispatchable. Therefore, the total net active power required by the battery energy storage system in the grid is It can be calculated as:

[0100]

[0101] if is positive, the battery energy storage system will be controlled to charge. If the value is negative, the battery energy storage system will be controlled in the discharge mode. In order to ensure the balance of supply and demand in the power grid, it is necessary to design a suitable dispatching strategy to share the distributed ESS. , which can be achieved by controlling the battery energy storage system The discharge / charge power is achieved by:

[0102]

[0103] Step 3.3: Due to the internal resistance of the battery, power loss will occur when the battery energy storage system is charged / discharged. and They represent the actual stored power and extracted power of the ith battery energy storage system, respectively, and are calculated as follows:

[0104] (6)

[0105] (7)

[0106] in, and Represent the charging and discharging efficiencies respectively.

[0107] Step 4: Considering the relevant power constraints, construct the optimization problem model of the distributed battery energy storage system, including:

[0108] For the sake of simplicity, the present invention only discusses the case where the battery energy storage system needs to operate in a charging mode, and the designed algorithm can also be applied to the case where the battery energy storage system operates in a discharging mode.

[0109] The total real power input into the battery energy storage system can be calculated as:

[0110]

[0111] The charging rate and the SoC (State of Charge, SoC) of the battery energy storage system will affect the charging efficiency. The charging efficiency and charging rate are linearly related, and the formula is:

[0112] (9)

[0113] in and are two constants representing the inverse of the intercept and slope of this linear relationship.

[0114] By substituting formula (9) into formula (8), the objective function becomes:

[0115]

[0116] Where Min is the minimization function.

[0117] Considering the economic issues of power grid operation, the power loss caused by charging the battery energy storage system should be minimized. The present invention can be implemented by controlling the charging reference value of the battery energy storage system to maximize equation (8). Therefore, the present invention will design a distributed battery energy storage system collaborative control strategy to maximize the total input power of the battery energy storage system by reducing power losses. Taking into account the relevant constraints, the following optimization problem model can be obtained:

[0118] (11)

[0119] in, and Represent the lower and upper charging limits respectively.

[0120] Step 5, solve the optimization problem in step 4, and the distributed battery energy storage system coordinated operation optimization algorithm:

[0121] Step 5.1: Use the Lagrangian algorithm to solve the above optimization problem. According to formulas (10) and (11), the Lagrangian function L can be obtained:

[0122]

[0123] in is the Lagrange multiplier.

[0124] Step 5.2, calculate the dual function, which is the Lagrangian function in The minimum value on :

[0125]

[0126] Where inf means finding the infimum of a set, and ,satisfy:

[0127]

[0128] (14)

[0129] Step 5.3, Lagrange multipliers Update along the negative gradient direction, the update equation is:

[0130]

[0131] In the formula, represents the update step size, represents the partial derivative of the function D with respect to v, and N represents the total number of nodes in the network. As can be seen from the above formula, the update of the Lagrange multiplier v needs to utilize global information and global information .

[0132] Step 5.4: The present invention considers the complex communication environment where there is a delay in node communication. Assume that the node With Node The communication delay at the kth iteration is , without loss of generality, the delay satisfies:

[0133] (16)

[0134] In the formula, Indicates the upper bound of the corresponding delay.

[0135] Step 5.5: Each battery energy storage system control unit can be abstracted as an intelligent agent with computing power and weak communication capabilities. Represented as a graph consisting of a set of vertices and a set of edges ; Representation Node Can receive nodes If there is a path from i to j, the corresponding two vertices are called connected; the set of nodes that can send information to node j is called the input neighbor node set of the node. ,Right now .

[0136] The node update rate is set to:

[0137]

[0138] In the formula,

[0139] (18)

[0140] in, Represents the state information of node j; represents the weight coefficient; represents the upper bound of the communication delay; represents the delay; r is the time step of the delay, from 0 to the maximum delay ; Represents an indicator function used to indicate whether communication with a corresponding delay has occurred.

[0141] It can be seen from the above formula that when node j is updated, all received neighbor nodes are processed, including the neighbor node information that arrives with a delay. In addition, the update of node information does not require the knowledge of the upper limit of the delay.

[0142] Step 5.6, each node is set to contain two state information and , in order to obtain global information , the node initial value is set to , the status update rate is as follows:

[0143]

[0144]

[0145] Then , then when hour, The limit satisfies the following formula:

[0146] .

[0147] The flow chart of the coordinated operation optimization algorithm of the distributed battery energy storage system proposed in the present invention is as follows: Figure 2 First, the load node voltage amplitude and phase angle are estimated, and then the actual load power is estimated using formulas (2) and (3).

[0148] In order to verify the effectiveness of the proposed energy storage distributed operation strategy optimization scheme for supply and demand balance, the present invention uses a 5-storage energy storage system to test the proposed consistent distributed optimization strategy. Assuming that the microgrid operates in an isolated island state, the communication network topology diagram is as follows: Figure 3 As shown, the communication delay takes a random number with an upper bound of 3, and the system parameters are shown in Table 1:

[0149] Table 1 Parameters of each energy storage unit

[0150]

[0151] The convergence of the Lagrnge multiplier v is as follows Figure 4 The average consistency result of the battery energy storage system is shown in Figure 5 Verified in.

[0152] The power generation of the microgrid is set to a constant value, and the load is set to a constant value, but its measurement contains non-Gaussian noise. Each energy storage unit exchanges information with adjacent energy storage units and updates local information. The step size of the energy storage unit in the distributed iteration process is 10-3 , and no charging rate constraints are imposed on the battery energy storage system. The test results are as follows Figure 6 and Figure 7 The test results show that the total charging power converges to 150KW, which is the local net power, that is, the sum of the total net power in the microgrid. Each energy storage unit also converges to a fixed constant value, verifying the effectiveness of the algorithm.

[0153] In the former case, the charging power of the battery energy storage system exceeds the upper limit, which may lead to the existence of an infeasible solution. Therefore, in order to meet the actual application scenario, we impose the following restrictions on the charging power:

[0154] (twenty two)

[0155] The simulation results are as follows Figure 8 and Fig. 9 As shown. Figure 8 and Fig. 9 In contrast, when the charging rate is constrained, their final convergence values ​​are the same. The difference is that when there is a constraint, each iteration is within the actual situation constraint, which is more practical in actual applications.

[0156] Although embodiments of the present invention have been shown and described, it will be appreciated by those skilled in the art that various changes, modifications, substitutions and variations may be made to the embodiments without departing from the principles and spirit of the present invention, and that the scope of the present invention is defined by the appended claims and their equivalents.

Claims

1. A distributed energy storage operation optimization method for supply and demand balance, characterized in that: include: Step 1: Define the active power balance relationship in the power grid, including: For smart distribution grids that include battery energy storage systems, energy storage devices can be regarded as elastic loads to adjust the balance between power supply and consumption. As shown in the following formula, the active power balance in the grid can be expressed as: in, , , Represents the collection of distributed generators, loads and battery energy storage systems, Indicates The power generated by the generator, represents the power consumed by the jth load, is the charging / discharging power of the ith battery energy storage system in positive / negative conditions; Step 2, online estimation of load active power, including: Step 2.1, using the measurements with complex noise provided by the SCADA system and the phasor measurement unit to achieve robust estimation of the node voltage amplitude and phase angle; Step 2.2, using the power flow analysis method, we know that when the voltage amplitude and phase angle of the nodes at both ends of the branch are known, the active power of the branch can be calculated as follows: in, represents the active power of the branch between load node j and node m, and are the series conductance and susceptance, , and are the voltage phase angle and amplitude at node j, and are the voltage phase angle and amplitude at node m respectively; Step 2.3, calculate the estimated total power consumed by load node j , as shown below: ； Step 3, calculate the actual stored and extracted power of each battery energy storage system, including: Step 3.1, transmission loss is modeled by multiplying the load active power by a factor of 0.05; Step 3.2: The total net active power required by the battery energy storage system in the grid Calculated as: if is positive, the battery energy storage system will be controlled to charge. If the battery energy storage system is negative, it will be controlled in the discharge mode. In order to ensure the balance of supply and demand in the power grid, it is necessary to design a suitable dispatching strategy to share the distributed ESS. , which can be achieved by controlling the battery energy storage system The discharge / charge power is achieved by: Step 3.3: Due to the internal resistance of the battery, power loss will occur when the battery energy storage system is charged / discharged. and They represent the actual stored power and extracted power of the ith battery energy storage system, respectively, and are calculated as follows: in, and represent discharge and discharge efficiency respectively; Step 4: Considering the relevant power constraints, construct the optimization problem model of the distributed battery energy storage system, including: The total real power input into the battery energy storage system can be calculated as: The charging rate and the SoC of the battery energy storage system will affect the charging efficiency. The charging efficiency and charging rate are linearly related, and the formula is: in and are two constants representing the inverse of the intercept and slope of the linear relationship; By substituting formula (9) into formula (8), the objective function becomes: Where Min is the minimization function; A distributed battery energy storage system collaborative control strategy is designed to maximize the total input power of the battery energy storage system by reducing power loss. Considering the relevant constraints, the following optimization problem model is obtained: （11） in, and Represent the lower and upper limits of charging respectively; Step 5, solve the optimization problem in step 4 and design an optimization algorithm for coordinated operation of distributed battery energy storage systems, including: Step 5.1: Use the Lagrangian algorithm to solve the optimization problem obtained in step 4. According to formulas (10) and (11), the Lagrangian function L can be obtained as follows: in, is the Lagrange multiplier; Step 5.2, calculate the dual function, which is the Lagrangian function in The minimum value on : Where inf means finding the infimum of a set, and ,satisfy: （14） Step 5.3, Lagrange multipliers Update along the negative gradient direction, the update equation is: ， In the formula, represents the update step size, represents the partial derivative of the function D with respect to v, and N represents the total number of nodes in the network; Step 5.4, consider the complex communication environment where there is delay in node communication, assuming that node With Node The communication delay at the kth iteration is , without loss of generality, the delay satisfies: （16） In the formula, represents the upper bound of the corresponding delay; Step 5.5: Each battery energy storage system control unit can be abstracted as an intelligent agent with computing power and weak communication capabilities. Represented as a graph consisting of a set of vertices and a set of edges ; Representation Node Can receive nodes If there is a path from i to j, then the corresponding two vertices are called connected states; the set of nodes that can send information to node j is called the input neighbor node set of the node ,Right now ; The node update rate is set to: In the formula, （18） in, Represents the state information of node j; represents the weight coefficient; represents the upper bound of the communication delay; represents the delay; r is the number of time steps of the delay, from 0 to the maximum delay ; represents an indicator function used to indicate whether communication with a corresponding delay has occurred; Step 5.6, each node is set to contain two state information and , in order to obtain global information , the node initial value is set to , the status update rate is as follows: make , then when hour, The limit satisfies the following formula: 。

Images