Optimal configuration method for optical storage system in power distribution network

Abstract

The invention discloses an optimal configuration method for an optical storage system in a power distribution network. The method comprises the following steps: loading line parameters, load parameters, reference voltage of the system and an initial reference power value into a selected power grid model; establishing a locating and sizing optimization model of the photovoltaic system; solving the objective function, and performing optimization by taking the maximum photovoltaic capacity of the power distribution network as the objective function to find the maximum photovoltaic capacity of the power distribution network; performing clustering analysis on the daily load demand and the photovoltaic output of the location of the power distribution network by adopting a K-means clustering algorithm to obtain a typical daily load curve and a photovoltaic power generation curve; taking the daily operation cost of the micro-grid as the minimum target, considering the energy storage scheduling cost, the distributed power supply maintenance cost and the interaction cost, and meeting the charging and discharging power constraint and power balance of the energy storage system at the same time; and obtaining an optimal point scheduling scheme of energy storage by adopting a dynamic programming algorithm, and calculating the capacity of the energy storage. The method not only improves the quality of the power distribution network, but also considers the economy of energy storage investment, and reduces the investment cost.

Description

Technical Field

[0001] The present invention relates to the technical field of optimized configuration of photovoltaic energy storage systems in distribution networks, and in particular to an optimized configuration method of photovoltaic energy storage systems in distribution networks. The photovoltaic energy storage systems are comprehensively planned and the advantages of energy storage systems are utilized to effectively improve the quality of electric energy. Background Art

[0002] In recent years, with the rapid development of renewable energy, distributed generation and electric vehicles, the power system has also changed in operation, but some problems have also been exposed. For example, the increasing load demand has led to the system's installed capacity not being able to meet the distribution network's maintenance needs during peak periods; users have higher and higher requirements for power quality; photovoltaics have intermittent and unstable characteristics, which may cause safety hazards to the power grid, and may cause grid frequency and voltage fluctuations, affecting the stability of the power system and power quality. Summary of the invention

[0003] The technical problem to be solved by the present invention is to realize an optimal configuration method of a photovoltaic storage system in a distribution network, so as to obtain the optimal access position of the photovoltaic system, the optimal power capacity and the energy capacity of the energy storage system. The present invention not only improves the quality of the distribution network, but also takes into account the economy of energy storage investment and reduces the investment cost.

[0004] The technical solution adopted by the present invention is: a method for optimizing the configuration of a photovoltaic storage system in a distribution network, comprising the following steps:

[0005] S1, loading the line parameters, load parameters, system reference voltage and reference power initial values ​​into the selected power grid model;

[0006] S2. Establish a site selection and sizing optimization model for photovoltaic systems. First, define the objective function, which includes the minimum investment and operating cost of the photovoltaic system, system network loss, voltage stability margin, power balance constraint, distribution network security constraint, and total installed power constraint. Use the penalty function to establish a site selection and sizing optimization model for photovoltaic systems.

[0007] S3. The objective function is solved by using quantum particle swarm algorithm, and the maximum photovoltaic capacity of the distribution network is optimized by using Monte Carlo algorithm.

[0008] S4. Use K-means clustering algorithm to perform cluster analysis on the daily load demand and photovoltaic output of the distribution network location to obtain a typical daily load curve and photovoltaic power generation curve for determining the distributed energy storage capacity;

[0009] S5. In the lower-level planning, the daily operation cost of the microgrid is taken as the minimum target, the energy storage dispatching cost, the distributed power source maintenance cost and the interaction cost are considered, and the charging and discharging power constraints and power balance of the energy storage system are met at the same time;

[0010] S6. Use the dynamic programming algorithm to obtain the optimal scheduling plan for energy storage, and use the maximum interval algorithm to calculate the capacity of energy storage.

[0011] In S2, the functional expression of the minimum investment and operation cost of the photovoltaic system is:

[0012]

[0013] Where: C is the sum of the investment and operation costs of distributed photovoltaics; N PV is the number of distributed photovoltaic installations; is the installed capacity of the nth distributed photovoltaic; r is the discount rate of distributed photovoltaic; y is the service life; C i is the unit investment cost of photovoltaics; C 2 is the unit operating cost of PV;

[0014] Function expression for minimizing network loss:

[0015]

[0016] Where: P loss is the active network loss of the system; I i is the current in branch i; R i is the resistance in the branch; Nbranch is the total number of branches in the system;

[0017] The function expression of voltage stability margin is:

[0018]

[0019] Where: ΔU is the voltage stability margin, U j is the voltage of each node, U e is the expected voltage, U p is the maximum allowable voltage deviation; Nbus is the total number of nodes in the system.

[0020] In S2, the node power balance constraint is:

[0021]

[0022] Where: P ij , Q ij They represent the net active and reactive power injection values ​​of node j at time t respectively; I ij is the current on branch ij at the current moment t;

[0023] The security constraints of the distribution network are:

[0024]

[0025] Where: V i max are the lower and upper limits of the voltage at node j, respectively;

[0026] The total installed power constraints are:

[0027]

[0028] Where: P PVj is the photovoltaic power installed at the jth node; P PVmax The maximum allowed installed power of the photovoltaic power source;

[0029] The power balance constraint and distribution network security constraint adopt a penalty function, and its expression is:

[0030]

[0031] Where: k 1 , k 2 and k 3 is the penalty coefficient; the penalty coefficient is determined according to the upper and lower limits of the node voltage and the value range of the maximum allowable installed power of the PV power source. Start with a smaller coefficient and gradually increase it to observe the convergence of the algorithm and the quality of the solution.

[0032] In S3, the quantum particle swarm algorithm is an improved algorithm based on the particle swarm algorithm, which is described by the Schrödinger equation φ(x, t), and then the specific position of the example is simulated. The particle position update equation is as follows:

[0033]

[0034] Where: u is a random number, u∈[0,1], which obeys normal distribution, β is the expansion coefficient, and m and n are the expansion coefficient parameters.

[0035] In the process of obtaining the typical daily load curve and the photovoltaic power generation curve through the clustering algorithm, the cluster center is determined by iteratively selecting data points far from the selected center point as new center points to increase the distance between cluster centers and reduce the deviation of clustering results.

[0036] The specific process of obtaining the typical daily load curve and photovoltaic power generation curve by the clustering algorithm in S4 is:

[0037] K-means clustering algorithm: It is expressed as follows:

[0038]

[0039] Where: a is the number of cluster centers; N is the number of samples taken; C nm Is the nth sample whether it belongs to the m class; x n is the relevant factor to be clustered; ω m For class R m The cluster center of .

[0040] In S5, the expression of energy storage dispatch cost is:

[0041]

[0042] Where: K BESS is the converted unit charge and discharge cost; P t BESS-dis , P t BESS-ch η is the discharge and charging power of the energy storage system in the microgrid at the tth moment; η is the charging and discharging efficiency of the energy storage system, Δt is the duration of time period t, which is 1h;

[0043] The maintenance cost expression of distributed power generation is:

[0044]

[0045] Where: P jt is the active power output of power generation unit j at time t; K OMj is the operation management coefficient of power generation unit j;

[0046] Interaction cost expression:

[0047]

[0048] Where: P t buy , P t sell are the power purchased and sold by the microgrid to the distribution network at time t; m(t) is the real-time electricity price.

[0049] In S5, the daily operating cost constraints of the microgrid include the energy storage system charging and discharging power constraints and the power balance constraints; the energy storage system charging and discharging power constraints:

[0050] P BESS,c,max ≤P BESS,t ≤P BESS,dis,max (13)

[0051] Where: P BESS,c,max , P BESS,dis,max Respectively represent the maximum charging power and maximum discharging power of the energy storage system;

[0052] The power balance constraint:

[0053] P pv,t +P grid,t =P load,t +P BESS ,t (14)

[0054] Where: P pv,t is the output power of the photovoltaic system at time t; P grid,t is the interaction power between the microgrid and the distribution network; P Load,t is the required power of the load at time t.

[0055] In S6, the specific steps of the maximum interval algorithm are:

[0056] S6.1 According to the lower-level planning, the optimal power curve of the energy storage system within a time period is obtained;

[0057] S6.2 divides the charging and discharging process of the energy storage system operation curve into different intervals, accumulates the time in each interval, and obtains the capacity of each interval;

[0058]

[0059] Where P BESS,t is the charging and discharging power of the energy storage system at time t;

[0060] S6.3 selects the maximum interval capacity as the configuration capacity of the energy storage system:

[0061] S=max(S(t)),t=1,2,...,T (16).

[0062] Beneficial effects of the present invention: The optimization configuration method of the photovoltaic storage system in the distribution network of the present invention can help balance the supply and demand relationship of the power grid by using the energy storage system. The energy storage system adopted in this article can store excess electricity to ensure the stability of the power grid. Because of its fast energy response speed, peak shaving and valley filling, and improved voltage, it provides strong support for improving the power quality of the distribution network and improving the reliability of system operation. By establishing a site selection and capacity optimization model for distributed photovoltaic systems, the optimal access location and optimal power capacity of the photovoltaic system in the distribution network are calculated, and this is used as input data in the lower-level planning. Combined with the distributed energy storage system operation strategy planning model, the optimal output curve of the energy storage is finally solved, and the maximum interval algorithm is used to obtain the energy capacity of the energy storage device. Not only the power quality of the distribution network is improved, but also the economy of energy storage investment is considered to reduce investment risks. BRIEF DESCRIPTION OF THE DRAWINGS

[0063] Figure 1 It is a flow chart of the optimization configuration method of the photovoltaic storage system in the distribution network of the present invention;

[0064] Figure 2 This is a schematic diagram of the voltage occupied by the distributed power supply and energy storage without access;

[0065] Figure 3 This is a schematic diagram of node voltage after distributed power supply and energy storage are connected. DETAILED DESCRIPTION

[0066] A method for optimizing the configuration of a photovoltaic storage system in a distribution network, such as Figure 1 As shown, the following steps are included:

[0067] S1, loading the line parameters, load parameters, system reference voltage and reference power initial values ​​into the selected power grid model;

[0068] S2. Establish a site selection and sizing optimization model for photovoltaic systems. First, define the objective function, which includes the minimum investment and operating cost of the photovoltaic system, system network loss, voltage stability margin, power balance constraint, distribution network security constraint, and total installed power constraint. Use the penalty function to establish a site selection and sizing optimization model for photovoltaic systems.

[0069] S3. The objective function is solved by using quantum particle swarm algorithm, and the maximum photovoltaic capacity of the distribution network is optimized by using Monte Carlo algorithm.

[0070] S4. Use K-means clustering algorithm to perform cluster analysis on the daily load demand and photovoltaic output of the distribution network location to obtain a typical daily load curve and photovoltaic power generation curve for determining the distributed energy storage capacity;

[0071] S5. In the lower-level planning, the daily operation cost of the microgrid is taken as the minimum target, the energy storage dispatching cost, the distributed power source maintenance cost and the interaction cost are considered, and the charging and discharging power constraints and power balance of the energy storage system are met at the same time;

[0072] S6. Use the dynamic programming algorithm to obtain the optimal scheduling plan for energy storage, and use the maximum interval algorithm to calculate the capacity of energy storage.

[0073] In S2, the functional expression of the minimum investment and operation cost of the photovoltaic system is:

[0074]

[0075] Where: C is the sum of the investment and operation costs of distributed photovoltaics; N PV The number of distributed photovoltaic installations; is the installed capacity of the nth distributed photovoltaic; r is the discount rate of distributed photovoltaic; y is the service life; C i is the unit investment cost of photovoltaics; C 2is the unit operating cost of PV;

[0076] Function expression for minimizing network loss:

[0077]

[0078] Where: P loss is the active network loss of the system; I i is the current in branch i; R i is the resistance in the branch; Nbranch is the total number of branches in the system;

[0079] The function expression of voltage stability margin is:

[0080]

[0081] Where: ΔU is the voltage stability margin, U j is the voltage of each node, U e is the expected voltage, U p is the maximum allowable voltage deviation; Nbus is the total number of nodes in the system.

[0082] In S2, the node power balance constraint is:

[0083]

[0084] Where: P ij , Q ij They represent the net active and reactive power injection values ​​of node j at time t respectively; I ij is the current on branch ij at the current moment t;

[0085] The security constraints of the distribution network are:

[0086]

[0087] Where: V i max are the lower and upper limits of the voltage at node j, respectively;

[0088] The total installed power constraints are:

[0089]

[0090] Where: P PVj is the photovoltaic power installed at the jth node; P PVmax The maximum allowed installed power of the photovoltaic power source;

[0091] The power balance constraint and distribution network security constraint adopt a penalty function, and its expression is:

[0092]

[0093] Where: k 1 , k 2 and k 3 is the penalty coefficient; the penalty coefficient is determined according to the upper and lower limits of the node voltage and the value range of the maximum allowable installed power of the PV power source. Start with a smaller coefficient and gradually increase it to observe the convergence of the algorithm and the quality of the solution.

[0094] In S3, the quantum particle swarm algorithm is an improved algorithm based on the particle swarm algorithm, which is described by the Schrödinger equation φ(x, t), and then the specific position of the example is simulated. The particle position update equation is as follows:

[0095]

[0096] Where: u is a random number, u∈[0,1], which obeys normal distribution, β is the expansion coefficient, and m and n are the expansion coefficient parameters.

[0097] In S4, the clustering center is determined in the process of obtaining the typical daily load curve and the photovoltaic power generation curve through the clustering algorithm: the data point farther away from the selected center point is iteratively selected as the new center point to increase the distance between the cluster centers and reduce the deviation of the clustering results.

[0098] The specific process of obtaining the typical daily load curve and the photovoltaic power generation curve by the clustering algorithm in S4 is:

[0099] K-means clustering algorithm: It is expressed as follows:

[0100]

[0101] Where: a is the number of cluster centers; N is the number of samples taken; C nm Is the nth sample whether it belongs to the m class; x n is the relevant factor to be clustered; ω m For class R m The cluster center of .

[0102] In S5, the expression of energy storage dispatch cost is:

[0103]

[0104] Where: K BESS is the converted unit charge and discharge cost; P t BESS-dis , P t BESS-chη is the discharge and charging power of the energy storage system in the microgrid at the tth moment; η is the charging and discharging efficiency of the energy storage system, Δt is the duration of time period t, which is 1h;

[0105] The maintenance cost expression of distributed power generation is:

[0106]

[0107] Where: P jt is the active power output of power generation unit j at time t; K OMj is the operation management coefficient of power generation unit j;

[0108] Interaction cost expression:

[0109]

[0110] Where: P t buy , P t sell are the power purchased and sold by the microgrid to the distribution network at time t; m(t) is the real-time electricity price.

[0111] In S5, the daily operating cost constraints of the microgrid include the energy storage system charging and discharging power constraints and the power balance constraints; the energy storage system charging and discharging power constraints:

[0112] P BESS,c,max ≤P BESS,t ≤P BESS,dis,max (13)

[0113] Where: P BESS,c,max , P BESS,dis,max Respectively represent the maximum charging power and maximum discharging power of the energy storage system;

[0114] The power balance constraint:

[0115] P pv,t +P grid,t =P load,t +P BESS ,t (14)

[0116] Where: P pv,t is the output power of the photovoltaic system at time t; P grid,t is the interaction power between the microgrid and the distribution network; P Load,t is the required power of the load at time t.

[0117] In S6, the specific steps of the maximum interval algorithm are:

[0118] 1) Based on the lower-level planning, the optimal power curve of the energy storage system within a time period is obtained;

[0119] 2) Divide the charging and discharging process of the energy storage system operation curve into different intervals, accumulate the time in each interval, and obtain the capacity of each interval;

[0120]

[0121] Where P BESS,t is the charging and discharging power of the energy storage system at time t;

[0122] 3) Select the maximum interval capacity as the configuration capacity of the energy storage system:

[0123] S=max(S(t)),t=1,2,...,T (16).

[0124] The optical storage system planning method of this embodiment is applied to the IEEE-33 node system. Figure 2 , Figure 3 The voltage conditions of each node in the IEEE-33 node distribution network system before and after the distributed energy storage system is connected are divided into two scenarios. Figure 2 Scenario 1: No distributed power generation and energy storage; Figure 3 Scenario 2: Connecting to distributed power sources and energy storage. It can be seen that the system model after connecting to distributed power sources and energy storage in Scenario 2 is smoother, improving the quality and reliability of electricity.

[0125] The above are only specific embodiments of the present invention and are not intended to limit the present invention. For those skilled in the art, the present invention may have various modifications and variations. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present invention shall be included in the protection scope of the present invention.

Claims

1. A method for optimizing the configuration of a photovoltaic storage system in a distribution network, characterized in that: The following steps are involved: S1, loading the line parameters, load parameters, system reference voltage and reference power initial values ​​into the selected power grid model; S2. Establish a site selection and sizing optimization model for photovoltaic systems. First, define the objective function, which includes the minimum investment and operating cost of the photovoltaic system, system network loss, voltage stability margin, power balance constraint, distribution network security constraint, and total installed power constraint. Use the penalty function to establish a site selection and sizing optimization model for photovoltaic systems. S3. The objective function is solved by using quantum particle swarm algorithm, and the maximum photovoltaic capacity of the distribution network is optimized by using Monte Carlo algorithm. S4. Use K-means clustering algorithm to perform cluster analysis on the daily load demand and photovoltaic output of the distribution network location to obtain a typical daily load curve and photovoltaic power generation curve for determining the distributed energy storage capacity; S5. In the lower-level planning, the daily operation cost of the microgrid is taken as the minimum target, the energy storage dispatching cost, the distributed power source maintenance cost and the interaction cost are considered, and the charging and discharging power constraints and power balance of the energy storage system are met at the same time; S6. Use the dynamic programming algorithm to obtain the optimal scheduling plan for energy storage, and use the maximum interval algorithm to calculate the capacity of energy storage.

2. The method for optimizing configuration of a photovoltaic storage system in a distribution network according to claim 1, characterized in that: In S2, the functional expression of the minimum investment and operation cost of the photovoltaic system is: Where: C is the sum of the investment and operation costs of distributed photovoltaics; N PV The number of distributed photovoltaic installations; is the installed capacity of the nth distributed photovoltaic; r is the discount rate of distributed photovoltaic; y is the service life; C i is the unit investment cost of photovoltaics; C2 is the unit operating cost of photovoltaics; Function expression for minimizing network loss: Where: P loss is the active network loss of the system; I i is the current in branch i; R i is the resistance in the branch; Nbranch is the total number of branches in the system; The function expression of voltage stability margin is: Where: ΔU is the voltage stability margin, U j is the voltage of each node, U e is the expected voltage, U p is the maximum allowable voltage deviation; Nbus is the total number of nodes in the system.

3. The method for optimizing the configuration of a photovoltaic storage system in a distribution network according to claim 2, characterized in that : S2 The node power balance constraint is: Where: P ij , Q ij They represent the net active and reactive power injection values ​​of node j at time t respectively; I ij is the current on branch ij at the current moment t; The security constraints of the distribution network are: Where: are the lower and upper limits of the voltage at node j, respectively; The total installed power constraints are: Where: P PVj is the photovoltaic power installed at the jth node; P PVmax The maximum allowed installed power of the photovoltaic power source; The power balance constraint and distribution network security constraint adopt a penalty function, and its expression is: Where: k1, k2 and k3 are penalty coefficients; the penalty coefficients are determined according to the upper and lower limits of the node voltage and the value range of the maximum allowable installed power of the photovoltaic power source. Start with a smaller coefficient and gradually increase it to observe the convergence of the algorithm and the quality of the solution.

4. The method for optimizing configuration of a photovoltaic storage system in a distribution network according to claim 1, characterized in that: In S3, the quantum particle swarm algorithm is an improved algorithm based on the particle swarm algorithm, which is described by the Schrödinger equation φ(x, t), and then the specific position of the example is simulated. The particle position update equation is as follows: Where: u is a random number, u∈[0,1], which obeys normal distribution, β is the expansion coefficient, and m and n are the expansion coefficient parameters.

5. The method for optimizing configuration of a photovoltaic storage system in a distribution network according to claim 1, characterized in that: In the process of obtaining the typical daily load curve and the photovoltaic power generation curve through the clustering algorithm, the cluster center is determined by iteratively selecting data points far from the selected center point as new center points to increase the distance between cluster centers and reduce the deviation of clustering results.

6. The method for optimizing configuration of a photovoltaic storage system in a distribution network according to claim 1, characterized in that: The specific process of obtaining the typical daily load curve and photovoltaic power generation curve by the clustering algorithm in S4 is: K-means clustering algorithm: It is expressed as follows: Where: a is the number of cluster centers; N is the number of samples taken; C nm Is the nth sample whether it belongs to the m class; x n is the relevant factor to be clustered; ω m For class R m The cluster center of .

7. The method for optimizing configuration of a photovoltaic storage system in a distribution network according to claim 1, characterized in that: In S5, the expression of energy storage dispatch cost is: C t BESS =K BESS (P t BESS-dis / η+P t BESS-ch η)Δt (10) Where: K BESS is the converted unit charge and discharge cost; P t BESS-dis , P t BESS-ch η is the discharge and charging power of the energy storage system in the microgrid at the tth moment; η is the charging and discharging efficiency of the energy storage system, Δt is the duration of time period t, which is 1h; The maintenance cost expression of distributed power generation is: Where: P jt is the active power output of power generation unit j at time t; K OMj is the operation management coefficient of power generation unit j; Interaction cost expression: Where: P t buy , P t sell are the power purchased and sold by the microgrid to the distribution network at time t; m(t) is the real-time electricity price.

8. The method for optimizing configuration of a photovoltaic storage system in a distribution network according to claim 1, characterized in that: In S5, the daily operating cost constraints of the microgrid include the energy storage system charging and discharging power constraints and the power balance constraints; the energy storage system charging and discharging power constraints: P BESS,c,max ≤P BESS,t ≤P BESS,dis,max (13) Where: P BESS,c,max , P BESS,dis,max Respectively represent the maximum charging power and maximum discharging power of the energy storage system; The power balance constraint: P pv,t +P grid,t =P load,t +P BESS ,t (14) Where: P pv,t is the output power of the photovoltaic system at time t; P grid,t is the interaction power between the microgrid and the distribution network; P Load,t is the required power of the load at time t.

9. The method for optimizing configuration of a photovoltaic storage system in a distribution network according to claim 1, characterized in that: In S6, the specific steps of the maximum interval algorithm are: 1) Based on the lower-level planning, the optimal power curve of the energy storage system within a time period is obtained; 2) Divide the charging and discharging process of the energy storage system operation curve into different intervals, and accumulate the time in each interval. Get the capacity of each interval; Where P BESS,t is the charging and discharging power of the energy storage system at time t; 3) Select the maximum interval capacity as the configuration capacity of the energy storage system: S=max(S(t)),t=1,2,...,T(16).

Images