Distributed energy storage optimization method, system and medium for distribution network based on two-layer model

Abstract

The present invention discloses a distributed energy storage optimization method, system and medium for a distribution network based on a double-layer model, which belongs to the technical field of distribution network optimization operation, including: simulating and obtaining the load output error of the node based on a random scenario method, modeling the light intensity using Beta distribution, and constructing a double-layer operation model; setting the minimum risk coefficient of the node voltage as the inner optimization target based on the maximum entropy criterion of random variables; determining the energy storage power output as the outer optimization target based on the energy efficiency function; obtaining the constraint data set of the double-layer model based on system constraints; considering the node voltage safety margin and the comprehensive operation benefit, and solving the double-layer model based on the superior and inferior solution distance method, obtaining the optimal value of the double-layer model, and formulating the optimal energy storage operation plan that meets the node voltage safety margin of the distribution network. Through the present invention, the safety and operation efficiency of the distribution network can be improved, the smoothing regulation effect of the distributed energy storage device can be brought into play, and the flexibility of the distribution network operation can be enhanced.

Description

Technical Field

[0001] The present invention belongs to the technical field of distribution network optimization operation, and particularly relates to a distribution network distributed energy storage optimization method, system and medium based on a double-layer model. Background Art

[0002] Renewable energy sources are gradually becoming a supporting source of power in the energy mix. Distributed photovoltaic power generation follows the principles of adapting to local conditions, being clean and efficient, decentralized, and utilizing nearby resources. This fully utilizes solar energy resources and can reduce fossil energy consumption. However, due to the randomness and intermittent nature of photovoltaic output, the integration of distributed photovoltaics into distribution networks can easily lead to voltage overshoots, increased curtailment rates, and increased network losses, which in turn impact the operational safety and efficiency of distribution networks.

[0003] To address the aforementioned issues associated with distributed PV integration, consideration can be given to deploying distributed energy storage devices within the distribution network. These devices offer rapid charge and discharge response and smooth output, enabling voltage support and other functions, promoting the integration of distributed PV and enhancing the flexibility of distribution network operations. However, the frequent charging and discharging of distributed energy storage devices is detrimental to the lifespan of the energy storage system and its subsequent maintenance. Therefore, it is necessary to explore optimized operational methods for distributed energy storage that are suitable for efficient regulation of the distribution network.

[0004] For example, Chinese patent application CN115800347A discloses a method and system for optimizing the configuration of energy storage in a distribution network based on a planning-operation two-layer model. This method is suitable for connecting to new urban distribution networks containing distributed renewable energy sources. Under the premise of considering the safety constraints of the distribution network flow, this method takes the lowest annual average operating cost as the goal, comprehensively considers cost indicators such as the construction cost of the energy storage throughout its life cycle and the operating cost of the distribution network, and proposes a two-layer model for energy storage system planning-operation. It also finds key coupling factors in the two-layer model and converts the two-layer model into a single-layer model. The solver is used to solve the problem, thereby obtaining an optimized configuration plan for energy storage capacity. This method realizes the organic and coordinated flow of energy between the distribution network, energy storage, distributed renewable energy, and loads, providing support for the distribution network load to respond to the peak and valley electricity prices of the power grid, alleviate the pressure on the power grid supply, and improve the operating efficiency of the distribution network.

[0005] However, the above-mentioned existing technologies do not take into account the losses caused by frequent operation of energy storage. At the same time, the uncertain output of loads and photovoltaics will increase the complexity of distribution network trends, thereby affecting the node voltage safety. Therefore, studying the optimized operation of distributed energy storage in the distribution network under uncertain loads and photovoltaic access can ensure the safe operation of the distribution network and improve operational efficiency and economic benefits. Summary of the Invention

[0006] To solve the above problems, the present invention provides a distributed energy storage optimization method, system and medium for distribution network based on a double-layer model to solve the problems in the prior art.

[0007] In order to achieve the above-mentioned object of the invention, the present invention proposes a distributed energy storage optimization method for distribution network based on a two-layer model, comprising:

[0008] Step S1: multiple nodes are set based on the transmission mode of the distribution network, the load output error of the nodes is simulated and obtained based on the random scenario method, and a two-layer model is established for the light intensity of the nodes based on the load output error and Beta distribution, wherein the two-layer model includes an inner model and an outer model;

[0009] Step S2: calculating the probability flow of the distribution network based on the semi-invariant method, generating a node voltage probability density function based on the probability flow according to the first criterion, calculating the risk coefficient of the node voltage based on the node voltage probability density function, and setting the minimum risk coefficient as the inner optimization target of the inner model;

[0010] Step S3: constructing an energy efficiency function for the distributed energy storage of the distribution network based on the acquired first variable value and second variable value, and calculating an outer optimization target of the outer model based on the energy efficiency function;

[0011] Step S4: setting multiple constraint conditions, and obtaining a constraint data set of the two-layer model based on the constraint conditions, wherein the constraint conditions include system power flow constraint, photovoltaic output constraint, energy storage state of charge and capacity constraint;

[0012] Step S5: The two-layer model obtains the node voltage safety margin and the operating benefit based on the node voltage probability density function and the energy efficiency function, and solves the two-layer model based on the superior-inferior solution distance method to obtain the optimal value of the two-layer model, and generates an energy storage operation plan based on the node voltage safety margin and the optimal value of the two-layer model.

[0013] Furthermore, establishing the two-layer model includes the following steps:

[0014] The random scenario method combines a hyperbolic distribution function and a probability density function and describes the load output error based on a first formula, which is: ,in, is the probability density function, π 、 ξ 、 μ and δ are the parameters of the hyperbolic distribution function, K λ ( ξ )for λ The modified Bessel function of the third kind of order, is the load output error of the node in the random scenario method;

[0015] The load output error and the Beta distribution are combined based on the second formula to generate the double-layer model to represent the photovoltaic output. , the second formula is: ,in, α and β are the parameters of the Beta distribution, P max is the maximum value of photovoltaic active power, is the photovoltaic active power;

[0016] Calculate the actual load demand of the i-th node in the distribution network at time t based on the third formula , the second formula is: ,in, 、 and are the user load, photovoltaic output and energy storage charging and discharging power of the i-th node at time t, is the load output error of the i-th node at time t in the random scenario method;

[0017] If the i-th node is not connected to a photovoltaic device or an energy storage device at time t, then or The value of is zero.

[0018] Furthermore, the semi-invariant method includes:

[0019] Obtain a power reference point of the distribution network, generate a Taylor first-order form of the AC power flow equation based on the power reference point, and obtain the fluctuation of node power based on the Taylor first-order form. The mathematical form of the AC power flow equation is: ,in, is a first-order semi-invariant, X 0 represents the expected value at the power reference point, is the power reference point, J 0 is the Jacobian matrix of the power reference point, W 0 is J The inverse matrix of 0, S Inject power fluctuation amount into the node;

[0020] Calculate the r-order semi-invariant based on the fourth formula , the fourth formula is: , where r is a positive integer, is a matrix W The e-th power of each element in 0, is the r-order semi-invariant of photovoltaic active power, is the r-order semi-invariant of load power;

[0021] Replacement of r-order semi-invariants based on conversion formula , the conversion formula is: , , where C is the probability combination operator symbol.

[0022] Furthermore, generating the node voltage probability density function includes:

[0023] The first criterion refers to the maximum entropy criterion of the random variable, and is set based on a mathematical model. The mathematical model is: ,in, is the entropy value, To satisfy the constraint symbols, is the node voltage probability density function, is the constraint function;

[0024] The optimal solution of the node voltage probability density function is output based on the mathematical model, and a Lagrangian function is constructed based on the maximum entropy value. The Lagrangian function is: ,in, λ 0 and λ r is the Lagrange multiplier, m is any positive integer;

[0025] The extreme value condition of the Lagrangian function is set based on the third formula, and the third formula is: ,exist When , the node voltage probability density function is: , where exp is the exponential function symbol.

[0026] Furthermore, obtaining the inner layer optimization target includes:

[0027] Obtain node voltage deviation, and establish the i-th node voltage comprehensive function based on the node voltage deviation , the voltage comprehensive function of the i-th node is: ,in, is the voltage deviation of the i-th node, and These are voltage deviation parameters. is the voltage amplitude of the i-th node;

[0028] The node voltage probability density function calculates the risk coefficient of the node voltage based on the fourth formula , the fourth formula is: , where the minimum risk coefficient is set as the inner optimization target , is the maximum amplitude of the voltage at the i-th node, is the minimum amplitude of the voltage at the i-th node.

[0029] Furthermore, calculating the outer optimization target includes:

[0030] The distributed energy storage determines the change value of the photovoltaic output, charges during the peak period of the photovoltaic output, and discharges during the valley period when the photovoltaic output is less than the load demand, and calculates the first variable value based on the fifth formula , the fifth formula is: ,in, and are the electricity prices during discharge and charging respectively, and are the energy storage discharge power and energy storage charging power respectively, and are the valley period and the peak period respectively, and are the discharge efficiency and charging efficiency of the distributed energy storage respectively;

[0031] The second variable value is calculated based on the sixth formula , the sixth formula is: ,in, C s and are the line loss cost and energy storage charging and discharging loss cost of the distributed energy storage, The system operation period, P s is the line loss power during the period;

[0032] The energy efficiency function outputs the maximum operating benefit based on the seventh formula , and determine the energy storage output power, the seventh formula is: , set the energy storage output power under the maximum operating benefit as the outer optimization target .

[0033] Furthermore, obtaining the constraint data set includes:

[0034] Setting the system power flow constraint and the node voltage safety margin as the first constraint of the inner optimization objective in the two-layer model, setting the photovoltaic output constraint, the energy storage state of charge, and the capacity constraint as the second constraint of the outer optimization objective in the two-layer model, and combining the first constraint and the second constraint to generate the constraint data set of the two-layer model;

[0035] The first constraint is: , ,in, and are the power source active power, photovoltaic active power, energy storage active power and load active power of the i-th node respectively, and are the power supply reactive power, photovoltaic reactive power, energy storage reactive power and load reactive power of the i-th node respectively, and are the voltage amplitudes of the i-th node and the j-th node respectively, is the voltage phase between the i-th node and the j-th node, and are the system conductance and system susceptance of the distribution network respectively;

[0036] The second constraint is: , ,in, and The minimum and maximum values ​​corresponding to the photovoltaic active power in the photovoltaic output constraint are respectively, and are the minimum and maximum values ​​corresponding to the photovoltaic reactive power in the photovoltaic output constraint, respectively. is the value corresponding to the energy storage charge state, is the value corresponding to the capacity constraint.

[0037] Furthermore, obtaining the optimal value of the two-layer model includes:

[0038] The inner model in the two-layer model transmits the photovoltaic output and the node voltage safety margin to the outer model based on the inner optimization target; the outer model formulates the energy storage operation plan based on the outer optimization target and the risk factor of the distributed energy storage, and provides the probabilistic power flow information to the inner model;

[0039] The maximum indices of the inner optimization objective and the outer optimization objective of the two-layer model are calculated based on the good-bad solution distance method. The good-bad solution distance method is: ,in, When is the inner model, When is the outer model, is the maximum size index of model n in the f-th solution, is the optimization target of model n in the f-th solution, is the minimum value of the optimization target of model n in all solutions, is the maximum value of the optimization target of model n among all the schemes, f is a positive integer from 1 to F, and F is the number of schemes;

[0040] The extremely large index is normalized based on the eighth formula to generate a standard index. The eighth formula is: ,in, is the standard index of model n in the f-th scheme;

[0041] Based on the ninth formula, the first distance between the standard index of model n in the f-th solution and the corresponding maximum value is calculated respectively. , the second distance to the corresponding minimum , the ninth formula is: , calculate the fit of all standard indicators based on the tenth formula , the tenth formula is: , extract the standard index corresponding to the maximum fit and set it as the optimal value of the two-layer model.

[0042] The present invention also provides a distribution network distributed energy storage optimization system based on a two-layer model, which is used to implement the above-mentioned distribution network distributed energy storage optimization method based on a two-layer model. The system mainly includes:

[0043] a node analysis module, which sets multiple nodes based on the transmission mode of the distribution network, simulates and obtains the load output error of the nodes based on a random scenario method, and establishes a two-layer model for the light intensity of the nodes based on the load output error and Beta distribution, wherein the two-layer model includes an inner model and an outer model;

[0044] an optimization setting module, calculating a probability flow of the distribution network based on a semi-invariant method, generating a node voltage probability density function based on the probability flow according to a first criterion, calculating a risk coefficient of the node voltage based on the node voltage probability density function, setting a minimum risk coefficient as an inner optimization target of the inner model, constructing an energy efficiency function of the distributed energy storage of the distribution network based on the obtained first variable value and second variable value, and calculating an outer optimization target of the outer model based on the energy efficiency function;

[0045] A model constraint module sets multiple constraint conditions and obtains a constraint data set of the two-layer model based on the constraint conditions, wherein the constraint conditions include system power flow constraint, photovoltaic output constraint, energy storage state of charge and capacity constraint;

[0046] A model output module, wherein the two-layer model obtains the node voltage safety margin and the operating benefit based on the node voltage probability density function and the energy efficiency function, solves the two-layer model based on the superior-inferior solution distance method to obtain the optimal value of the two-layer model, and generates an energy storage operation plan based on the node voltage safety margin and the optimal value of the two-layer model.

[0047] The present invention also provides a computer medium storing program instructions, wherein when the program instructions are executed, the device where the computer medium is located is controlled to execute the above-mentioned distribution network distributed energy storage optimization method based on the two-layer model.

[0048] Compared with the prior art, the beneficial effects of the present invention are at least as follows:

[0049] The present invention first sets a random scenario method to obtain the load output error. It can be considered that the uncertainty scenarios of load and photovoltaic will increase the complexity of the distribution network flow and affect the node voltage safety in the distribution network. Then, based on the uncertainty of photovoltaic output and the random volatility of load, the semi-invariant method is used to calculate the probability flow of the distribution network. Finally, the node voltage safety probability density function is generated through the maximum entropy criterion of random variables, and the safety margin index of the node voltage is obtained.

[0050] The present invention also combines the relationship between energy efficiency functions to determine the optimal energy efficiency function, and forms a two-layer model with the node voltage safety margin index. Based on the optimal value of the two-layer model, the optimal energy storage operation plan is generated to achieve safe operation and economic benefit optimization of distributed energy storage in the distribution network. BRIEF DESCRIPTION OF THE DRAWINGS

[0051] Figure 1 This is a flow chart of the distributed energy storage optimization method for distribution network based on a double-layer model of the present invention;

[0052] Figure 2 This is a structural diagram of the distributed energy storage optimization system for distribution network based on a double-layer model of the present invention;

[0053] Figure 3 This is a simulation model diagram of the distribution network including the distributed energy storage system in the present invention;

[0054] Figure 4 The output curves of the system load and a single photovoltaic unit in the simulation model of the present invention;

[0055] Figure 5 The voltage diagram of key nodes before optimization in the simulation model of the present invention;

[0056] Figure 6 is the distributed energy storage active power change curve in the simulation model of the present invention;

[0057] Figure 7 The optimized voltage diagram of key nodes in the simulation model of the present invention;

[0058] Figure 8 This is a diagram showing changes in system loss before and after optimization in the simulation model of the present invention. DETAILED DESCRIPTION

[0059] In order to make the purpose, technical solutions and advantages of the present invention more clearly understood, the present invention will be further described in detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present invention and are not intended to limit the present invention.

[0060] It is understood that the terms "first," "second," etc., used herein may be used to describe various elements, but unless otherwise specified, these elements are not limited by these terms. These terms are only used to distinguish a first element from another element. For example, a first xx script may be referred to as a second xx script, and similarly, a second xx script may be referred to as a first xx script without departing from the scope of this application.

[0061] like Figure 1 As shown in FIG, the distributed energy storage optimization method for distribution network based on the two-layer model includes:

[0062] Step S1: multiple nodes are set based on the transmission mode of the distribution network, the load output error of the node is simulated and obtained based on the random scenario method, and a two-layer model is established for the light intensity of the node based on the load output error and Beta distribution. The two-layer model includes an inner model and an outer model.

[0063] Specifically, in this embodiment, the nodes of the distribution network refer to the points where different power components in the distribution network are interconnected, including but not limited to physical connection points and virtual nodes, such as the points where loads, power generation equipment, and energy storage equipment are connected to the power grid. The random scenario method is used to process and evaluate the uncertainty and randomness in the system. The load output error refers to the difference between the actual load power and the expected value of the load power in the uncertainty scenario. The load output error can be described by the Beta distribution to describe the corresponding random variable, combined with the light intensity of the node and a two-layer model is formed, where the light intensity is the most important factor affecting the photovoltaic output of the node. The two-layer model refers to a two-layer optimization model that correlates the optimization objectives and constraints of the inner and outer layers, which can achieve safe operation and efficiency optimization of distributed energy storage in the distribution network.

[0064] Step S2: Calculate the probability flow of the distribution network based on the semi-invariant method. Generate the node voltage probability density function based on the first criterion. Calculate the risk coefficient of the node voltage based on the node voltage probability density function. Set the minimum risk coefficient as the inner optimization target of the inner model.

[0065] Specifically, in this embodiment, the semi-invariant method can be used to process random variables and probability distributions in the distribution network and generate a probabilistic power flow for the distribution network. This probabilistic power flow can quantify the randomness and uncertainty of the distribution network. The node voltage probability density function refers to the safety probability density function of the node voltage, which can be used to calculate the node voltage risk factor. The inner optimization objective refers to the optimization objective of the inner model in the two-layer model.

[0066] Step S3: The distributed energy storage of the distribution network constructs an energy efficiency function based on the acquired first variable value and second variable value, and calculates an outer optimization target of the outer model based on the energy efficiency function.

[0067] Specifically, in this embodiment, distributed energy storage refers to energy storage devices or systems dispersed across the distribution network. The first variable value refers to the operational efficiency of the distributed energy storage, and the second variable value refers to the system loss of the distributed energy storage. The energy efficiency function is a comprehensive function used to measure and analyze distributed energy storage, and the outer optimization objective refers to the optimization objective of the outer layer of the two-layer model.

[0068] Step S4: setting multiple constraints, and obtaining a constraint data set of the double-layer model based on the constraints, wherein the constraints include system flow constraints, photovoltaic output constraints, energy storage charge state and capacity constraints.

[0069] Specifically, in this embodiment, constraints are used to limit the feasible solution space of the two-layer model. In the distribution network, constraints ensure that the corresponding constraint dataset is not violated when optimizing the energy storage device operation plan, thereby achieving safe and efficient operation of the distribution network. System flow constraints refer to the constraints corresponding to the probabilistic flow of the distribution network, photovoltaic output constraints refer to the constraints on photovoltaic output fluctuations, energy storage state of charge refers to the constraints on the charging and discharging of the energy storage device, and capacity constraints refer to the constraints on the capacity of the energy storage device.

[0070] Step S5: The two-layer model obtains the node voltage safety margin and operating efficiency based on the node voltage probability density function and the energy efficiency function, and solves the two-layer model based on the superior-inferior solution distance method to obtain the optimal value of the two-layer model. The energy storage operation plan is generated based on the optimal values ​​of the node voltage safety margin and the energy efficiency function.

[0071] Specifically, in this embodiment, the node voltage safety margin refers to the difference between the voltage level of each node in the distribution network and the voltage range required for its safe operation. The size of the node voltage safety margin can be used to quantify the risk factor of the node voltage and reflect the stability and reliability of the distribution network. Therefore, the energy storage operation plan of the distribution network is formulated based on the node voltage safety margin and the economic benefits of operation. Among them, the energy storage operation plan determines the operation strategy of the distributed energy storage device or system in the distribution network through the optimal value in the two-layer model to achieve a series of predetermined target values, which can improve the stability and reliability of the distribution network and improve energy utilization efficiency.

[0072] As a preferred technical solution of the present invention, establishing a two-layer model includes the following steps:

[0073] The random scenario method combines the hyperbolic distribution function and the probability density function and describes the load output error based on the first formula, which is: ,in, is the probability density function, π 、 ξ 、 μ and δ are the parameters of the hyperbolic distribution function, K λ ( ξ )for λ The modified Bessel function of the third kind of order, is the load output error of the node in the random scenario method.

[0074] Based on the second formula, the load output error and Beta distribution are combined to generate a two-layer model to represent the photovoltaic output. , the second formula is: ,in, α and β are the parameters of the Beta distribution, P max is the maximum value of photovoltaic active power, P PV is the photovoltaic active power.

[0075] Calculate the actual load demand of the i-th node in the distribution network at time t based on the third formula , the second formula is: ,in, 、 and are the user load, photovoltaic output and energy storage charging and discharging power of the i-th node at time t, is the load output error of the i-th node at time t in the random scenario method.

[0076] If the i-th node is not connected to a photovoltaic device or an energy storage device at time t, then or The value of is zero.

[0077] Specifically, in this embodiment, the random scenario method can clearly describe the node load output error by combining multiple functional formulas. The third type of modified Bessel function is a Hankel function used to deal with fluctuation problems. Photovoltaic output refers to the photovoltaic active power generated by the photovoltaic power generation system, that is, the power of the photovoltaic system converting solar energy into electrical energy. Substituting the photovoltaic active power into the second formula can represent the corresponding photovoltaic output.

[0078] As a preferred technical solution of the present invention, the semi-invariant method includes:

[0079] Obtain the power reference point of the distribution network, generate the Taylor first-order form of the AC power flow equation based on the power reference point, and obtain the fluctuation of node power based on the Taylor first-order form. The mathematical form of the AC power flow equation is: ,in, is a first-order semi-invariant, X 0 represents the expected value at the power reference point, is the power reference point, J 0 is the Jacobian matrix of the power reference point, W 0 is J The inverse matrix of 0, S Inject power fluctuation amount into the node.

[0080] Calculate the r-order semi-invariant based on the fourth formula , the fourth formula is: , where r is a positive integer, is a matrix W The e-th power of each element in 0, is the r-order semi-invariant of photovoltaic active power, is the r-order semi-invariant of the load power.

[0081] Replacement of r-order semi-invariants based on conversion formula , the conversion formula is: , ,in, is the probability combination operator symbol.

[0082] Specifically, in this embodiment, a power reference point refers to a reference node in the distribution network used for power flow analysis and calculations, representing the distribution network's operating state under standard conditions. The semi-invariant method has the mathematical property of linear additivity. Therefore, in the fourth formula, the PV active power and the load power can be added to generate the semi-invariants of various orders.

[0083] As a preferred technical solution of the present invention, generating a node voltage probability density function includes:

[0084] The first criterion refers to the maximum entropy criterion of random variables. The first criterion is set based on the mathematical model formula, which is: ,in, is the entropy value, To satisfy the constraint symbols, is the node voltage probability density function, is the constraint function.

[0085] Based on the mathematical model, the node voltage probability density function is output and the Lagrangian function is constructed. The Lagrangian function is: ,in, λ 0 and λ r is the Lagrange multiplier, and m is any positive integer.

[0086] The extreme value condition of the Lagrangian function is set based on the third formula, which is: ,exist When , the node voltage probability density function is: , where exp is the exponential function symbol.

[0087] Specifically, the node voltage probability density function is generated based on the probability density function and the semi-invariant method combined with the Lagrangian function.

[0088] As a preferred technical solution of the present invention, obtaining the inner layer optimization target includes:

[0089] Obtain node voltage deviation and establish the i-th node voltage comprehensive function based on the node voltage deviation , the comprehensive function of the voltage of the i-th node is: ,in, is the voltage deviation of the i-th node, and These are voltage deviation parameters. is the voltage amplitude of the i-th node.

[0090] The node voltage probability density function is based on the fourth formula to calculate the node voltage risk coefficient , the fourth formula is: , where the minimum risk factor is set as the inner optimization target , is the maximum amplitude of the voltage at the i-th node, is the minimum amplitude of the voltage at the i-th node.

[0091] Specifically, the node voltage comprehensive function is a mathematical function used to quantify the impact of node voltage deviation in the distribution network on the efficiency of the distribution network system. It combines factors such as the magnitude of the voltage deviation, voltage stability, and power efficiency. By combining the node voltage probability density function and the node voltage comprehensive function to generate the fourth formula, the risk coefficient of the inner layer model can be calculated.

[0092] As a preferred technical solution of the present invention, calculating the outer optimization target includes:

[0093] According to the change in photovoltaic output, the distributed energy storage is charged during the peak period of photovoltaic output and discharged during the valley period when photovoltaic output is less than the load demand. The first variable value is calculated based on the fifth formula. , the fifth formula is: ,in, and are the electricity prices during discharge and charging respectively, and are the energy storage discharge power and energy storage charging power respectively, and The valley period and peak period are respectively and are the discharge efficiency and charging efficiency of distributed energy storage respectively;

[0094] Calculate the second variable value based on the sixth formula , the sixth formula is: ,in, C s and They are the line loss cost of distributed energy storage and the energy storage charging and discharging loss cost, The system operation period, P s is the line loss power of the time period;

[0095] The energy efficiency function outputs the maximum operating benefit based on the seventh formula , and determine the energy storage output power, the seventh formula is: , set the energy storage output power under maximum operating benefit as the outer optimization target .

[0096] Specifically, in this embodiment, the energy storage power refers to the power of distributed energy storage in the distribution network. Setting the energy storage output power under maximum operating benefit as the outer optimization target can maximize the efficiency of the device corresponding to the distributed energy storage.

[0097] As a preferred technical solution of the present invention, obtaining the constraint data set includes:

[0098] The system flow constraint and node voltage safety margin are set as the first constraint of the inner optimization objective in the two-layer model, and the photovoltaic output constraint, energy storage charge state and capacity constraint are set as the second constraint of the outer optimization objective in the two-layer model. The first constraint and the second constraint are combined to generate the constraint data set of the two-layer model.

[0099] The first constraint is: , ,in, and are the power source active power, photovoltaic active power, energy storage active power and load active power of the i-th node respectively, and are the power supply reactive power, photovoltaic reactive power, energy storage reactive power and load reactive power of the i-th node respectively, and are the voltage amplitudes of the i-th node and the j-th node respectively, is the voltage phase between the i-th node and the j-th node, and are the system conductance and system susceptance of the distribution network respectively.

[0100] The second constraint is: , ,in, and The minimum and maximum values ​​corresponding to the photovoltaic active power in the photovoltaic output constraint, and are the minimum and maximum values ​​corresponding to the photovoltaic reactive power in the photovoltaic output constraint, is the value corresponding to the energy storage charge state, is the value corresponding to the capacity constraint.

[0101] Specifically, in this embodiment, by setting the optimization objectives of the double-layer model respectively through the first constraint and the second constraint, the accuracy of distributed energy storage can be improved, and the safe operation of the distribution network can be guaranteed.

[0102] As a preferred technical solution of the present invention, obtaining the optimal value of the two-layer model includes:

[0103] In the two-layer model, the inner model transmits the photovoltaic output and node voltage safety margin to the outer model based on the inner optimization objective. The outer model formulates the energy storage operation plan based on the outer optimization objective and risk factor of the distributed energy storage, and provides probabilistic power flow information to the inner model.

[0104] Based on the good and bad solution distance method, the maximum index of the inner optimization target and the outer optimization target of the two-layer model are calculated respectively. The good and bad solution distance method is: ,in, When is the inner model, When is the outer model, is the maximum size index of model n in the f-th solution, is the optimization target of model n in the f-th solution, is the minimum value of the optimization target of model n in all solutions, is the maximum value of the optimization target of model n among all solutions, f is a positive integer from 1 to F, and F is the number of solutions.

[0105] Based on the eighth formula, the extremely large index is standardized and a standard index is generated. The eighth formula is: ,in, is the standard index of model n in the f-th solution.

[0106] Based on the ninth formula, the first distance between the standard index of model n in the f-th scheme and the corresponding maximum value is calculated respectively. , the second distance to the corresponding minimum , the ninth formula is: , calculate the fit of all standard indicators based on the tenth formula , the tenth formula is: , extract the standard index corresponding to the maximum fit and set it as the optimal value of the two-layer model.

[0107] Specifically, in this embodiment, the optimal value of the two-layer model can be used to formulate an energy storage operation plan, where the number of schemes refers to the total number of all schemes that can solve the two-layer model, and the energy storage operation plan refers to the operation plan of the device or system corresponding to the distributed energy storage, that is, the charging and discharging plan. Formulated by the optimal value of the two-layer model, it can ensure the safe operation of the distribution network and improve the operating efficiency.

[0108] like Figure 2 As shown, the present invention also provides a distribution network distributed energy storage optimization system based on a double-layer model, which is used to implement the above-mentioned distribution network distributed energy storage optimization method based on a double-layer model. The system mainly includes:

[0109] The node analysis module sets multiple nodes based on the transmission mode of the distribution network, simulates and obtains the load output error of the node based on the random scenario method, and establishes a two-layer model for the light intensity of the node based on the load output error and Beta distribution. The two-layer model includes an inner model and an outer model.

[0110] The optimization setting module calculates the probability flow of the distribution network based on the semi-invariant method. The probability flow generates the node voltage probability density function based on the first criterion, calculates the risk coefficient of the node voltage based on the node voltage probability density function, and sets the minimum risk coefficient as the inner optimization target of the inner model. The distributed energy storage of the distribution network constructs an energy efficiency function based on the obtained first variable value and second variable value, and calculates the outer optimization target of the outer model based on the energy efficiency function.

[0111] The model constraint module sets multiple constraint conditions and obtains the constraint data set of the two-layer model based on the constraint conditions. The constraint conditions include system flow constraints, photovoltaic output constraints, energy storage charge state and capacity constraints.

[0112] Model output module, the two-layer model obtains the node voltage safety margin based on the node voltage probability density function and energy efficiency function, and solves the two-layer model based on the superior-inferior solution distance method to obtain the optimal value of the two-layer model. The energy storage operation plan is generated based on the node voltage safety margin and the optimal value of the two-layer model.

[0113] Specifically, in this embodiment, the IEEE 33-node standard model is improved through a simulation model. The IEEE 33-node model is a benchmark system model widely used in power system dynamic research and teaching. It contains 33 nodes and 31 transmission lines with different generator, load and switchgear configurations, such as Figure 3 As shown in the figure, a distribution network simulation model with a distributed energy storage system was constructed. In the simulation model, the system base voltage was set to 12.66 kV. Photovoltaic power plants (PV) with a rated capacity of 400 kW were installed at nodes 12, 20, 25, and 28, respectively. Distributed energy storage (ESS) was installed at nodes 3 and 9, with a maximum charge and discharge power of 300 kW and a charge and discharge efficiency of 0.9. The transformer rated capacity was 2.5 kVA.

[0114] like Figure 4 The following are the system load and output curves of a single photovoltaic unit: Figure 5 As shown in the figure, the photovoltaic access nodes (12, 20, 25, 28) and the feeder end nodes (18, 33) are selected as key nodes. When the distributed energy storage output is not optimized, the voltage of the photovoltaic access nodes 12, 20, and 28 rises rapidly during the photovoltaic output peak period (10-16 o'clock), causing the voltage to exceed the upper limit, and at some moments it exceeds 1.06pu. At the same time, the voltage of the feeder end nodes 18 and 33 drops due to insufficient power in the evening hours (21-24 o'clock, 0-04 o'clock), and at some moments it is lower than 0.93pu, which is not conducive to the safe operation of the system. Figure 6 The figure shows the active power output plan of distributed energy storage after double-layer optimization operation. A positive output indicates energy storage charging, while a negative output indicates energy storage discharging. Figure 7The figure shows the voltage of key nodes after the distributed energy storage charging and discharging output is optimized. Figure 7 It can be seen that after optimization, the voltage of key nodes at each moment fluctuates within the safe range of 0.95~1.05pu. Figure 8 The following is a graph showing the change in system losses before and after optimization. Compared to the system without energy storage, the average system loss decreased by 23.6% after energy storage optimization. These simulation results demonstrate that the proposed distributed energy storage optimization method, system, and medium for distribution networks based on a two-layer model can achieve safe operation and improve operational efficiency.

[0115] The present invention also provides a computer medium storing program instructions, wherein when the program instructions are executed, the device where the computer medium is located is controlled to execute the above-mentioned distribution network distributed energy storage optimization method based on the double-layer model.

[0116] It should be understood that, although the various steps in the flow chart of each embodiment of the present invention are shown in sequence according to the indication of the arrows, these steps are not necessarily performed in sequence according to the order indicated by the arrows. Unless otherwise specified herein, the execution of these steps is not strictly limited in order, and these steps can be performed in other orders. Moreover, at least a portion of the steps in each embodiment may include a plurality of sub-steps or a plurality of stages, and these sub-steps or stages are not necessarily performed at the same time, but can be performed at different times, and the execution order of these sub-steps or stages is not necessarily performed in sequence, but can be performed in turn or alternately with at least a portion of other steps or sub-steps or stages of other steps.

[0117] Those skilled in the art will appreciate that all or part of the processes in the above-described method embodiments can be implemented by instructing the relevant hardware through a computer program. The above-described program can be stored in a non-volatile computer-readable storage medium. When executed, the program can include the processes of the above-described method embodiments. Any reference to memory, storage, database, or other media used in the various embodiments provided herein may include non-volatile and / or volatile memory. Non-volatile memory may include read-only memory (ROM), programmable ROM (PROM), electrically programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM), or flash memory. Volatile memory may include random access memory (RAM) or external cache memory. By way of illustration and not limitation, RAM is available in various forms, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), double data rate SDRAM (DDRSDRAM), enhanced SDRAM (ESDRAM), synchronous link DRAM (SLDRAM), RAMbus direct RAM (RDRAM), direct RAMbus dynamic RAM (DRDRAM), and RAMbus dynamic RAM (RDRAM).

[0118] The technical features of the above-mentioned embodiments can be combined arbitrarily. In order to make the description concise, not all possible combinations of the technical features in the above-mentioned embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0119] The above embodiments merely illustrate several implementations of the present invention, and while their descriptions are relatively specific and detailed, they should not be construed as limiting the scope of the present invention. It should be noted that a person skilled in the art would be able to make numerous variations and improvements without departing from the spirit of the present invention, all of which fall within the scope of protection of the present invention. Therefore, the scope of protection of the present invention shall be determined by the appended claims.

[0120] The above are only preferred embodiments of the present invention and are not intended to limit the present invention. Any modifications, equivalent substitutions and improvements made within the spirit and principles of the present invention should be included in the scope of protection of the present invention.

Claims

1. A distributed energy storage optimization method for distribution network based on a two-layer model, characterized in that: The method comprises the following steps: Step S1: multiple nodes are set based on the transmission mode of the distribution network, the load output error of the node is simulated and obtained based on the random scenario method, and a double-layer model is established for the light intensity of the node based on the load output error and Beta distribution, wherein the double-layer model includes an inner model and an outer model; Step S2: Calculate the probability flow of the distribution network based on the semi-invariant method, generate a node voltage probability density function based on the probability flow according to the first criterion, calculate the risk coefficient of the node voltage based on the node voltage probability density function, and set the minimum risk coefficient as the inner optimization target of the inner model; Step S3: The distributed energy storage of the distribution network constructs an energy efficiency function based on the acquired first variable value and second variable value, and calculates an outer optimization target of the outer model based on the energy efficiency function; Step S4: setting a plurality of constraint conditions, and obtaining a constraint data set of the double-layer model based on the constraint conditions, wherein the constraint conditions include system flow constraint, photovoltaic output constraint, energy storage charge state and capacity constraint; Step S5: The two-layer model obtains the node voltage safety margin and the operating benefit based on the node voltage probability density function and the energy efficiency function, and solves the two-layer model based on the superior-inferior solution distance method to obtain the optimal value of the two-layer model, and generates an energy storage operation plan based on the node voltage safety margin and the optimal value of the two-layer model.

2. The method according to claim 1, characterized in that: Establishing the two-layer model includes the following steps: The random scenario method combines a hyperbolic distribution function and a probability density function and describes the load output error based on a first formula, which is: ,in, is the probability density function, π , ξ , μ and δ are the parameters of the hyperbolic distribution function, K λ ( ξ )for λ The modified Bessel function of the third kind is is the load output error of the node in the random scenario method; Based on the second formula, the load output error and the Beta distribution are combined to generate the double-layer model to represent the photovoltaic output. , the second formula is: ,in, α and β are the parameters of the Beta distribution, P max is the maximum value of photovoltaic active power, is the photovoltaic active power; Based on the third formula, the actual load demand of the i-th node in the distribution network at time t is calculated , the second formula is: ,in, , and are the user load, photovoltaic output and energy storage charging and discharging power of the i-th node at time t, is the load output error of the i-th node at time t in the random scenario method; If the i-th node is not connected to a photovoltaic device or an energy storage device at time t, then or The value of is zero.

3. The method according to claim 1, characterized in that The semi-invariant method includes: The power reference point of the distribution network is obtained, a Taylor first-order form of the AC power flow equation is generated based on the power reference point, and the fluctuation change of the node power is obtained based on the Taylor first-order form. The mathematical form of the AC power flow equation is: ,in, is a first-order semi-invariant, X 0 represents the expected value at the power reference point, is the power reference point, J 0 is the Jacobian matrix of the power reference point, W 0 is J The inverse matrix of 0, S Inject power fluctuation amount into the node; Calculate the r-order semi-invariant based on the fourth formula , the fourth formula is: , where r is a positive integer, For the matrix W The e-th power of each element in 0, is the r-order semi-invariant of the photovoltaic active power, is the r-order semi-invariant of load power; Replacement of r-order semi-invariants based on conversion formula , the conversion formula is: , , where C is the probability combination operator symbol.

4. The method according to claim 3, characterized in that Generating the node voltage probability density function includes: The first criterion refers to the maximum entropy criterion of the random variable, and the first criterion is set based on a mathematical model, and the mathematical model is: ,in, is the entropy value, To satisfy the constraint symbol, is the node voltage probability density function, is the constraint function; The optimal solution of the node voltage probability density function is output based on the mathematical model, and a Lagrangian function is constructed based on the maximum entropy value. The Lagrangian function is: ,in, λ 0 and λ r is the Lagrange multiplier, m is any positive integer; The extreme value condition of the Lagrangian function is set based on the third formula, and the third formula is: ,exist When , the node voltage probability density function is: , where exp is the exponential function symbol.

5. The method according to claim 4, characterized in that Obtaining the inner layer optimization target includes: Obtain node voltage deviation, and establish the i-th node voltage comprehensive function based on the node voltage deviation , the i-th node voltage comprehensive function is: ,in, is the voltage deviation of the ith node, and These are voltage deviation parameters. is the voltage amplitude of the ith node; The node voltage probability density function calculates the risk coefficient of the node voltage based on the fourth formula , the fourth formula is: , where the minimum risk factor is set as the inner optimization target , is the maximum amplitude of the voltage at the i-th node, is the minimum amplitude of the voltage at the i-th node.

6. The method according to claim 1, characterized in that Calculating the outer optimization target includes: Distributed energy storage determines the change value of the photovoltaic output, charges during the peak period of the photovoltaic output, and discharges during the valley period when the photovoltaic output is less than the load demand, and calculates the first variable value based on the fifth formula , the fifth formula is: ,in, and are the electricity price during discharge and the electricity price during charge, and are the energy storage discharge power and energy storage charging power respectively, and are the valley time period and the peak time period respectively, and are respectively the discharge efficiency and the charging efficiency of the distributed energy storage; Calculate the second variable value based on the sixth formula , the sixth formula is: ,in, C s and are the line loss cost and energy storage charging and discharging loss cost of the distributed energy storage, The system operation period, P s is the line loss power in the period; the energy efficiency function outputs the maximum operating benefit based on the seventh formula , and determine the energy storage output power, the seventh formula is: , the energy storage output power under the maximum operating benefit is set as the outer optimization target .

7. The method according to claim 1, characterized in that Obtaining constraint data sets includes: The system power flow constraint and the node voltage safety margin are set as the first constraint of the inner optimization target in the double-layer model, the photovoltaic output constraint, the energy storage charge state and the capacity constraint are set as the second constraint of the outer optimization target in the double-layer model, and the first constraint and the second constraint are combined to generate the constraint data set of the double-layer model; The first constraint is: , ,in, and are the power source active power, photovoltaic active power, energy storage active power and load active power of the i-th node respectively. and are the power source reactive power, photovoltaic reactive power, energy storage reactive power and load reactive power of the i-th node respectively, and are the voltage amplitudes of the i-th node and the j-th node respectively, is the voltage phase between the i-th node and the j-th node, and are respectively the system conductance and the system susceptance of the distribution network; The second constraint is: , ,in, and The minimum value and the maximum value corresponding to the photovoltaic active power in the photovoltaic output constraint are respectively, and are respectively the minimum and maximum values ​​corresponding to the photovoltaic reactive power in the photovoltaic output constraint, is the value corresponding to the energy storage charge state, is the value corresponding to the capacity constraint.

8. The method according to claim 1, characterized in that: Obtaining the optimal value of the two-layer model includes: The inner model in the double-layer model transmits the photovoltaic output and the node voltage safety margin to the outer model based on the inner optimization target, and the outer model formulates the energy storage operation plan based on the outer optimization target and the risk factor of the distributed energy storage, and provides probabilistic power flow information to the inner model; The maximum index of the inner optimization target and the outer optimization target of the two-layer model are calculated based on the good and bad solution distance method, and the good and bad solution distance method is: ,in, When is the inner model, When is the outer model, is the maximum size index of model n in the f-th solution, is the optimization target of model n in the f-th solution, is the minimum optimization target value of model n in all solutions, is the maximum value of the optimization target of model n among all the schemes, f is a positive integer from 1 to F, and F is the number of schemes; The extremely large index is standardized based on the eighth formula to generate a standard index. The eighth formula is: ,in, is the standard index of model n in the f-th scheme; Based on the ninth formula, the first distance between the standard index of the model n in the f-th scheme and the corresponding maximum value is calculated respectively. , the second distance to the corresponding minimum , the ninth formula is: , based on the tenth formula, calculate the fit of all standard indicators , the tenth formula is: , extract the standard index corresponding to the maximum fit and set it as the optimal value of the two-layer model.

9. A distributed energy storage optimization system for a distribution network based on a two-layer model, used to implement the method according to any one of claims 1 to 8, characterized in that: The system includes the following modules: A node analysis module, which sets a plurality of nodes based on the transmission mode of the distribution network, simulates and obtains the load output error of the node based on the random scenario method, and establishes a double-layer model for the light intensity of the node based on the load output error and Beta distribution, wherein the double-layer model includes an inner model and an outer model; An optimization setting module, which calculates the probability flow of the distribution network based on a semi-invariant method, generates a node voltage probability density function based on a first criterion, calculates the risk coefficient of the node voltage based on the node voltage probability density function, sets the minimum risk coefficient as the inner optimization target of the inner model, constructs an energy efficiency function of the distributed energy storage of the distribution network based on the acquired first variable value and second variable value, and calculates the outer optimization target of the outer model based on the energy efficiency function; A model constraint module, which sets a plurality of constraint conditions and obtains a constraint data set of the double-layer model based on the constraint conditions, wherein the constraint conditions include system flow constraint, photovoltaic output constraint, energy storage charge state and capacity constraint; A model output module, wherein the double-layer model obtains the node voltage safety margin based on the node voltage probability density function and the energy efficiency function, solves the double-layer model based on the superior-inferior solution distance method, obtains the optimal value of the double-layer model, and generates an energy storage operation plan based on the node voltage safety margin and the optimal value of the double-layer model.

10. A computer medium, characterized in that The computer medium stores program instructions, wherein when the program instructions are executed, the device where the computer medium is located is controlled to execute the method according to any one of claims 1 to 8.

Images