Continuous time method for optimizing operation of active power distribution network under high distributed photovoltaic permeability

Abstract

The invention relates to a continuous time method for optimizing the operation of an active power distribution network under high distributed photovoltaic permeability, and the method comprises the following steps: S1, obtaining all decision variables Psi D in the active power distribution network and a constraint E Psi D < = F, and combining power type energy storage and energy type energy storage into a hybrid energy storage system; s2, delimiting a scheduling time period, modeling the output of each time period of each component in the active power distribution network as a curve C (t) (t belongs to [0, 1]) based on four times of Bernstein interpolation, segmenting according to the scheduling time period, and modeling the energy storage E of an energy storage unit based on five times of Bernstein interpolation; s3, CB = [CB0, CB1, CB2, CB3] is a vector formed by interpolation of the curve C (t) in each time period, and CB is calculated according to the power values C (0) and C (1) and the slopes C '(0) and C' (1) of the head end and the tail end of the curve C (t); and S4, obtaining a constraint EBCB < = FB for interpolation, obtaining an active power distribution network model of continuous time, and solving the active power distribution network model. The performance of the hybrid energy storage system can be fully utilized, the operation cost of the active power distribution network is reduced, and the method has the characteristics of high accuracy and suitability for economic dispatching.

Description

Technical Field

[0001] The present invention relates to the field of electrical engineering technology, and more specifically to a continuous time method for optimizing the operation of an active distribution network under high distributed photovoltaic penetration. Background Art

[0002] Active distribution network (ADN), as an important part of the new power system, is an important part of achieving the "30-60" goal. The active distribution network has changed the situation that the traditional passive distribution network only connects to various types of loads, including a large number of distributed energy resources (DER) with increasing penetration in the distribution network. At the same time, the connection of a large number of DERs changes the original one-way flow at the point of common coupling (PCC) between the ADN and the upper power grid, and then a two-way flow appears. The specific structure is as follows Figure 2 shown.

[0003] Among all types of DER, photovoltaics are favored for their clean and pollution-free characteristics. With the advancement of photovoltaic technology, distributed photovoltaics directly connected to the grid have developed rapidly. According to statistics, the new installed capacity of distributed photovoltaics in my country in 2021 has exceeded that of centralized photovoltaics. However, the high proportion of distributed photovoltaic penetration has also brought many problems, mainly because distributed photovoltaics are significantly affected by the environment and have large power fluctuations, which introduces dynamic characteristics to ADN.

[0004] In order to solve the problem of distributed photovoltaic fluctuations, some studies advocate the application of energy storage systems (ESS) to help improve the efficiency and reliability of distributed energy systems. However, it should be pointed out that the impact of distributed photovoltaics is different at different times of the day. During sunny days, it is mainly reflected in power fluctuations during the scheduling period. After nightfall, there is no photovoltaic power generation, which puts pressure on the power supply of ADN. Therefore, it is necessary to combine power storage and energy storage into a hybrid energy storage system (HESS) to participate in the operation of ADN.

[0005] Other studies have shown that when a large number of distributed photovoltaics are integrated into ADN, their dynamic characteristics will increase significantly, causing their impact on the power system to become increasingly complex. In the traditional discrete time scheduling method, it is assumed that the power in each period corresponds to a certain value, and the power curve is modeled as Figure 3 It can be seen that the step curve has a low degree of fit to the actual power curve, and there is only a power jump at the junction of time periods. The dynamic characteristics brought by distributed photovoltaics are mainly reflected in the power changes within the time period. Summary of the invention

[0006] The technical problem to be solved by the present invention is to provide a continuous-time method for optimizing the operation of active distribution networks under high distributed photovoltaic penetration, which can fully utilize the performance of hybrid energy storage systems and reduce the operating costs of active distribution networks, and has the characteristics of high accuracy and suitability for economic dispatch.

[0007] The technical solution adopted by the present invention to solve the technical problem is: constructing a continuous time method for optimizing the operation of active distribution network under high distributed photovoltaic penetration, comprising the following steps:

[0008] S1. Obtain all decision variables Ψ in the active distribution network D and constraint EΨ D ≤F, and combine power-type energy storage and energy-type energy storage into a hybrid energy storage system;

[0009] S2. Define the dispatch period, and model the output of each component in the active distribution network in each period as a curve C(t)(t∈[0,1]) based on the fourth Bernstein interpolation and divide it into segments according to the dispatch period. The storage energy E of the energy storage unit is modeled based on the fifth Bernstein interpolation;

[0010] S3, C B =[C B0 ,C B1 ,C B2 ,C B3 ] is the vector composed of the interpolation of the curve C(t) in each time period. C(t) is calculated based on the power values ​​C(0), C(1) and the slopes C′(0), C′(1) at the beginning and end of the curve C(t). B ;

[0011] S4. Obtain the constraint E on the interpolation B C B ≤F B , the continuous-time active distribution network model is obtained and solved.

[0012] According to the above scheme, in step S1, the active distribution network is connected to the upper power grid through the common coupling point, and the decision variable Ψ D and constraint EΨ D ≤F is based on discrete-time ADN modeling, and the combined energy storage system is composed of flywheel energy storage and all-vanadium liquid flow battery;

[0013] Constraint EΨ of active distribution network D ≤F includes the unit's output constraints, climbing constraints and start-stop constraints, energy storage unit charging and discharging constraints and storage capacity constraints, distributed photovoltaic abandonment constraints; transmission line flow constraints, upper-level power grid power purchase constraints and power balance constraints.

[0014] According to the above scheme, all decision variables and prediction curves are modeled as continuous time curves C(t), which are Bezier curves.

[0015] According to the above scheme, in step S2, each curve C(t) is a piecewise curve, which is segmented into each scheduling period t∈[0,1] based on the scheduling period.

[0016] The interpolation vector of the curve C(t) is obtained based on the quartic Bernstein interpolation, C B =[C B0 ,C B1 ,C B2 ,C B3 ], where C Bi (i=0,1,2,3) is the interpolation value.

[0017] The curve C(t) is uniquely determined by its set of control points in each time period t∈[0,1]. The control points of the curve C(t) and its interpolation vector C B =[C B0 ,C B1 ,C B2 ,C B3 ] one-to-one correspondence, the control points are:

[0018]

[0019] The analytical expression of the curve in the time period t∈[0,1] is given by its interpolation vector C B The only way to determine is as follows:

[0020]

[0021] Each interpolation value is calculated from the values ​​C(0), C(1) and the slopes C'(0), C'(1) at the beginning and end of the curve C(t) in the time period t∈[0,1]:

[0022]

[0023] After integration over t∈[0,1], C(t) has nothing to do with time t and only with the interpolation vector C B related:

[0024]

[0025] The differential curve C'(t) of the curve C(t) corresponds to the Bernstein interpolation after reducing the order by one:

[0026]

[0027]

[0028] If the curve C(t) is always equal to 0, then its interpolated values ​​must also be 0:

[0029]

[0030] If the curve C(t) is always less than a certain positive constant α, then the interpolation value is also always less than α:

[0031]

[0032] All decision variables except energy storage Ψ D , and the corresponding curves all correspond to the quartic Bernstein interpolation coefficient vector (·) B The energy storage capacity of the energy storage unit is based on the fifth-order Bernstein interpolation model E B ;

[0033] The decision variables related to the hybrid energy storage system are the energy storage charging / discharging power P CHB / P DISB , energy storage charge / discharge status Energy storage E B ; Other parameters are the upper and lower limits of energy storage charge / discharge power P ch / P dis , energy storage charge / discharge efficiency η ch / η dis , energy storage limit E max ;

[0034] The interpolation coefficient vector corresponding to all other decision variables is the unit output value P GB , unit start and stop status Photovoltaic output value P pvB , the amount of abandoned light P pvdB , the transmission line power P LB , power purchase quantity P exB ; Other parameters are the unit climbing / climbing rate R up / R low , unit operation cost coefficient α, unit start-up and shutdown cost coefficient β, electricity price γ, transmission line capacity P Lmax , the upper limit of power purchase and sale capacity P max ;

[0035] all(·) B All the above-mentioned interpolation vectors C are observed B the nature of;

[0036] The constraints of all continuous-time models are obtained according to the properties described in equations (3)-(8) for each interpolation vector (·) B The analytical expressions of the curves can be obtained by using equations (1) and (2).

[0037] According to the above scheme, the operation constraints of the hybrid energy storage system are as follows:

[0038]

[0039] Among them, equations (9)-(10) are the energy storage curve expressions corresponding to the fifth-order Bernstein interpolation; equations (11)-(12) are the energy storage curves to ensure the energy storage E B The corresponding curve is a first-order continuous constraint with smooth and continuous connection from beginning to end. Equations (13)-(14) are charge and discharge constraints; equations (15)-(16) are energy storage constraints.

[0040] According to the above scheme, in step S4, the operation constraints of the active distribution network are as follows:

[0041]

[0042] Among them, formula (18) is the optimization objective of the overall model; formulas (19)-(21) are the unit operation cost, the ring microgrid power purchase cost, and the photovoltaic abandonment penalty; formula (22) is the power balance constraint; formulas (23)-(24) are the unit output constraint and climbing constraint; formula (25) is the converted upper power grid power purchase and sales constraint; formula (26) is the transmission line flow constraint; formulas (27)-(29) are distributed photovoltaic abandonment constraints; formulas (30)-(31) are first-order continuity constraints to ensure that the beginning and end of the curve are smoothly and continuously connected.

[0043] The continuous time method for optimizing the operation of active distribution networks under high distributed photovoltaic penetration of the present invention has the following beneficial effects:

[0044] 1. The present invention can give full play to the advantages of hybrid energy storage in terms of power density and energy density to cope with the fluctuation of distributed photovoltaics during the day and the increased load at night, and is suitable for multi-energy complementation;

[0045] 2. The present invention provides a continuous-time model of an active distribution network, which can fully characterize the sub-hourly power fluctuations of DERs in the active distribution network and is suitable for participating in the power market and economic dispatch. BRIEF DESCRIPTION OF THE DRAWINGS

[0046] The present invention will be further described below with reference to the accompanying drawings and embodiments, in which:

[0047] Figure 1 is a conceptual diagram of an active distribution network of a continuous time method for optimizing the operation of an active distribution network under high distributed photovoltaic penetration of the present invention;

[0048] Figure 2 It is a comparison diagram of the discrete time step power curve of the present invention and the actual curve;

[0049] Figure 3 is a flow chart of a continuous time method for optimizing the operation of an active distribution network under high distributed photovoltaic penetration of the present invention;

[0050] Figure 4 1 is a schematic diagram of a fourth-order Bernstein interpolation corresponding to the Bezier curve of the present invention;

[0051] Figure 5 It is the improved IEEE-13 node active distribution network topology diagram of the present invention;

[0052] Figure 6 This is the intraday HESS operation status diagram obtained by the continuous time method;

[0053] Figure 7 This is the intraday HESS operation status diagram obtained by the discrete time method. DETAILED DESCRIPTION

[0054] In order to have a clearer understanding of the technical features, purposes and effects of the present invention, specific embodiments of the present invention are now described in detail with reference to the accompanying drawings.

[0055] like Figure 1-7 As shown, the continuous time method for optimizing the operation of the active distribution network under high distributed photovoltaic penetration of the present invention obtains all decision variables Ψ in the active distribution network D and constraint EΨ D ≤F, and combine power-type energy storage and energy-type energy storage into a hybrid energy storage system; define the dispatch period, and model the output of each component in the active distribution network in each period as a curve C(t)(t∈[0,1]) based on the fourth Bernstein interpolation and divide it into segments according to the dispatch period, and the energy storage E of the energy storage unit is modeled based on the fifth Bernstein interpolation;

[0056] C B =[C B0 ,C B1 ,C B2 ,C B3 ] is the vector composed of the interpolation of the curve C(t) in each time period. C(t) is calculated based on the power values ​​C(0), C(1) and the slopes C′(0), C′(1) at the beginning and end of the curve C(t). B ; Convert all constraints to constraints on interpolation E B C B ≤F B , the continuous-time active distribution network model is obtained and solved.

[0057] The present invention provides a continuous time method for optimizing the operation of an active distribution network under high distributed photovoltaic penetration, comprising the following steps:

[0058] S1. Obtain all decision variables Ψ in the active distribution network D and constraint EΨ D ≤F, and combine power-type energy storage and energy-type energy storage into a hybrid energy storage system.

[0059] S2. Define the dispatch period, and model the output of each component in the active distribution network in each period as a curve C(t)(t∈[0,1]) based on the fourth-order Bernstein interpolation and divide it into segments according to the dispatch period. The storage energy E of the energy storage unit is modeled based on the fifth-order Bernstein interpolation.

[0060] S3, C B =[C B0 ,C B1 ,C B2 ,C B3 ] is the vector composed of the interpolation of the curve C(t) in each time period. C(t) is calculated based on the power values ​​C(0), C(1) and the slopes C′(0), C′(1) at the beginning and end of the curve C(t). B .

[0061] S4. Obtain the constraint E on the interpolation B C B ≤F B , the continuous-time active distribution network model is obtained and solved.

[0062] like Figure 5 As shown in the conceptual diagram, the active distribution network is connected to the upper power grid through the point of common coupling (PCC). D and constraint EΨ D ≤F is based on discrete-time ADN modeling. All decision variables and prediction curves in the active distribution network are modeled as continuous-time curves C(t), which are Bezier curves. Each curve C(t) is a piecewise curve, which is segmented into each scheduling period t∈[0,1] based on the scheduling period. The interpolation vector of curve C(t) is obtained based on the quartic Bernstein interpolation, C B =[C B0 ,C B1 ,C B2 ,C B3 ], where C Bi (i=0,1,2,3) is interpolation. The curve C(t) is uniquely determined by its control point set in each time period t∈[0,1], such as Figure 4 As shown. The control points of the curve C(t) and its interpolation vector C B =[C B0 ,C B1 ,C B2 ,C B3 ] one-to-one correspondence, the control points are:

[0063]

[0064] The analytical expression of the curve in the time period t∈[0,1] is given by its interpolation vector CB The only way to determine is as follows:

[0065]

[0066] Each interpolation value can be calculated from the values ​​C(0), C(1) and the slopes C'(0), C'(1) at the beginning and end of the curve C(t) in the time period t∈[0,1]:

[0067]

[0068] After integration over t∈[0,1], C(t) has nothing to do with time t and only with the interpolation vector C B related:

[0069]

[0070] The differential curve C'(t) of the curve C(t) corresponds to the Bernstein interpolation after reducing the order by one:

[0071]

[0072] If the curve C(t) is always equal to 0, then its interpolated values ​​must also be 0:

[0073]

[0074] If the curve C(t) is always less than a certain positive constant α, then the interpolation value is also always less than α:

[0075]

[0076] All decision variables except energy storage Ψ D , and the corresponding curves all correspond to the quartic Bernstein interpolation coefficient vector (·) B The energy storage capacity of the energy storage unit is based on the fifth-order Bernstein interpolation model E B , because the energy storage charging / discharging power is its differential relative to the energy storage, and the power corresponds to the fourth-order Bernstein interpolation, so the energy storage corresponds to the fifth-order Bernstein interpolation. The decision variable related to the hybrid energy storage system is the energy storage charging / discharging power P CHB / P DISB , energy storage charge / discharge status Energy storage E B ; Other parameters are the upper and lower limits of energy storage charge / discharge power P ch / P dis , energy storage charge / discharge efficiency η ch / η dis , energy storage limit E max The interpolation coefficient vector corresponding to all other decision variables is the unit output value P GB , unit start and stop status Photovoltaic output value P pvB , the amount of abandoned light P pvdB , the transmission line power P LB , power purchase quantity P exB ; Other parameters are the unit climbing / climbing rate R up / R low , unit operation cost coefficient α, unit start-up and shutdown cost coefficient β, electricity price γ, transmission line capacity P Lmax , the upper limit of power purchase and sale capacity P max .all(·) B All the above-mentioned interpolation vectors C are observed B nature.

[0077] The constraints of all continuous-time models are obtained according to the properties described in equations (35)-(40) for each interpolation vector (·) B Constrained by , the analytical expressions of each curve can be calculated according to equations (33)-(34).

[0078] The operating constraints of the hybrid energy storage system are as follows:

[0079]

[0080] Among them, equations (41)-(42) are the energy storage curve expressions corresponding to the fifth-order Bernstein interpolation; equations (43)-(44) are the energy storage curves to ensure B The corresponding curve is a first-order continuous constraint with smooth and continuous connection from beginning to end. Equations (45)-(46) are charge and discharge constraints; equations (47)-(48) are energy storage constraints.

[0081] The operating constraints of the active distribution network are as follows:

[0082]

[0083]

[0084] Among them, formula (50) is the optimization objective of the overall model; formulas (51)-(53) are the unit operating costs, the ring microgrid power purchase costs, and the photovoltaic abandonment penalty; formula (22) is the power balance constraint; formulas (23)-(24) are the unit output constraints and climbing constraints; formula (25) is the converted upper power grid power purchase and sales constraints; formula (58) is the transmission line flow constraint; formulas (59)-(61) are distributed photovoltaic abandonment constraints; formulas (62)-(63) are first-order continuity constraints to ensure that the beginning and end of the curve are smoothly and continuously connected.

[0085] The following further describes the beneficial effects that can be achieved by this embodiment in conjunction with a specific application scenario.

[0086] This embodiment is verified in the improved IEEE-13 node ADN. The node system includes 1 unit, 3 distributed photovoltaics, 3 flywheel energy storages, and 1 all-vanadium liquid flow battery energy storage. The topology diagram is as follows: Figure 5 As shown, the electricity price refers to the summer industrial and commercial time-of-use electricity price in Wuhan.

[0087] The corresponding program was written on the computing software MATLAB R2024a, and the Yalmip solver equipped with Gurobi was called to solve the problem. To solve the optimization problem, the computing device used was: a HUAWEI matebook laptop with an Intel Core i7-10510U processor, 16GB of RAM, and Windows 10 Professional operating system. The solution results are shown in Table 1. Figure 6 and Figure 7 shown.

[0088] Table 1 Comparison of the running results of different methods

[0089]

[0090]

[0091] The embodiments of the present invention are described above in conjunction with the accompanying drawings, but the present invention is not limited to the above-mentioned specific implementation modes, which are merely illustrative rather than restrictive. Under the guidance of the present invention, ordinary technicians in this field can also make many forms without departing from the scope of protection of the present invention and the claims, all of which are within the protection of the present invention.

Claims

1. A continuous-time method for optimizing the operation of active distribution networks under high distributed photovoltaic penetration, characterized in that: The following steps are involved: S1. Obtain all decision variables Ψ in the active distribution network D and constraint EΨ D ≤F, and combine power-type energy storage and energy-type energy storage into a hybrid energy storage system; S2. Define the dispatch period, and model the output of each component in the active distribution network in each period as a curve C(t)(t∈[0,1]) based on the fourth Bernstein interpolation and divide it into segments according to the dispatch period. The storage energy E of the energy storage unit is modeled based on the fifth Bernstein interpolation; S3, C B =[C B0 ,C B1 ,C B2 ,C B3 ] is the vector composed of the interpolation of the curve C(t) in each time period. C(t) is calculated based on the power values ​​C(0), C(1) and the slopes C′(0), C′(1) at the beginning and end of the curve C(t). B ; S4. Obtain the constraint E on the interpolation B C B ≤F B , the continuous-time active distribution network model is obtained and solved.

2. The continuous time method for optimizing the operation of active distribution networks under high distributed photovoltaic penetration according to claim 1, characterized in that: In step S1, the active distribution network is connected to the upper power grid through the common coupling point, and the decision variable Ψ D and constraint EΨ D ≤F is based on discrete-time ADN modeling, and the combined energy storage system is composed of flywheel energy storage and all-vanadium liquid flow battery; Constraint EΨ of active distribution network D ≤F includes the unit's output constraints, climbing constraints and start-stop constraints, energy storage unit charging and discharging constraints and storage capacity constraints, distributed photovoltaic abandonment constraints; transmission line flow constraints, upper-level power grid power purchase constraints and power balance constraints.

3. The continuous time method for optimizing the operation of active distribution networks under high distributed photovoltaic penetration according to claim 2, characterized in that: All decision variables and prediction curves are modeled as continuous time curves C(t) and are Bezier curves.

4. The continuous time method for optimizing the operation of active distribution networks under high distributed photovoltaic penetration according to claim 1, characterized in that: In step S2, each curve C(t) is a piecewise curve, which is segmented into each scheduling period t∈[0,1] based on the scheduling period; The interpolation vector of the curve C(t) is obtained based on the quartic Bernstein interpolation, C B =[C B0 ,C B1 ,C B2 ,C B3 ], where C Bi (i=0,1,2,3) is interpolation; The curve C(t) is uniquely determined by its set of control points in each time period t∈[0,1]. The control points of the curve C(t) and its interpolation vector C B =[C B0 ,C B1 ,C B2 ,C B3 ] one-to-one correspondence, the control points are: The analytical expression of the curve in the time period t∈[0,1] is given by its interpolation vector C B The only way to determine is as follows: Each interpolation value is calculated from the values ​​C(0), C(1) and the slopes C'(0), C'(1) at the beginning and end of the curve C(t) in the time period t∈[0,1]: After integration over t∈[0,1], C(t) has nothing to do with time t and only with the interpolation vector C B related: The differential curve C'(t) of the curve C(t) corresponds to the Bernstein interpolation after reducing the order by one: If the curve C(t) is always equal to 0, then its interpolated values ​​must also be 0: If the curve C(t) is always less than a certain positive constant α, then the interpolation value is also always less than α: All decision variables except energy storage Ψ D , and the corresponding curves all correspond to the quartic Bernstein interpolation coefficient vector (·) B The energy storage capacity of the energy storage unit is based on the fifth-order Bernstein interpolation model E B ; The decision variables related to the hybrid energy storage system are the energy storage charging / discharging power P CHB / P DISB , energy storage charge / discharge status Energy storage E B ; Other parameters are the upper and lower limits of energy storage charge / discharge power P ch / P dis , energy storage charge / discharge efficiency η ch / η dis , energy storage limit E max ; The interpolation coefficient vector corresponding to all other decision variables is the unit output value P GB , unit start and stop status Photovoltaic output value P pvB , the amount of abandoned light P pvdB , the transmission line power P LB , power purchase quantity P exB ; Other parameters are the unit climbing / climbing rate R up / R low , unit operation cost coefficient α, unit start-up and shutdown cost coefficient β, electricity price γ, transmission line capacity P Lmax , the upper limit of power purchase and sale capacity P max ; all(·) B All the above-mentioned interpolation vectors C are observed B the nature of; The constraints of all continuous-time models are obtained according to the properties described in equations (3)-(8) for each interpolation vector (·) B The analytical expressions of the curves can be obtained by using equations (1) and (2).

5. The continuous time method for optimizing the operation of active distribution networks under high distributed photovoltaic penetration according to claim 2, characterized in that: The operating constraints of the hybrid energy storage system are as follows: Among them, equations (9)-(10) are the energy storage curve expressions corresponding to the fifth-order Bernstein interpolation; equations (11)-(12) are the energy storage curves to ensure the energy storage E B The corresponding curve is a first-order continuous constraint with smooth and continuous connection from beginning to end; Equations (13)-(14) are charge and discharge constraints; Equations (15)-(16) are energy storage constraints.

6. The continuous time method for optimizing the operation of active distribution networks under high distributed photovoltaic penetration according to claim 1, characterized in that: In step S4, the operation constraints of the active distribution network are as follows: Among them, formula (18) is the optimization objective of the overall model; formulas (19)-(21) are the unit operation costs, the ring microgrid power purchase costs, and the photovoltaic abandonment penalty; formula (22) is the power balance constraint; formulas (23)-(24) are the unit output constraints and ramp constraints; formula (25) is the converted upper power grid power purchase and sales constraints; formula (26) is the transmission line power flow constraint; formulas (27)-(29) are distributed photovoltaic abandonment constraints; Formulas (30) and (31) are first-order continuity constraints to ensure that the beginning and the end of the curve are smoothly and continuously connected.

Images