Distributed photovoltaic and energy storage area autonomous capability optimization method based on robust chance constraint model

Abstract

The invention provides a distributed photovoltaic and energy storage area autonomous capability optimization method based on a robust chance constraint model, and the method comprises the steps: considering the power fluctuation caused by distributed photovoltaic and load in a distributed photovoltaic and energy storage area autonomous scheduling plan, and defining the condition for achieving the autonomous capability based on the power fluctuation quantity in the area: in any time period, the power fluctuation quantity of the distributed photovoltaic and energy storage area is smaller than the power fluctuation quantity of the distributed photovoltaic and energy storage area; the regional maximum adjustable output is greater than or equal to the regional power total fluctuation quantity; opportunity constraint and a distributed robust optimization method based on Kullback-Leibler divergence are introduced into autonomous capability optimization, and a robust opportunity constraint model is obtained; the distributed robust opportunity constraint is converted into a traditional opportunity constraint through a fuzzy set equivalent conversion formula, and the traditional opportunity constraint is solved by adopting a risk value method; on the premise of considering the regional autonomous capability, the regional scheduling operation is optimized by taking the operation cost minimization as the target.

Description

Technical Field

[0001] The present invention relates to the technical fields of distributed photovoltaic, energy storage system, regional autonomy, etc., and in particular to a method for optimizing the regional autonomy capacity of distributed photovoltaic and energy storage based on a robust chance constraint model. Background Art

[0002] The emergence of large-scale energy storage systems has promoted the development of photovoltaic and wind power generation. Large-scale energy storage systems can cooperate with photovoltaic and wind turbines to achieve smooth output, peak load shifting, and planned output tracking, which increases the controllability of power generation, reduces the randomness and volatility of the power generation system, and improves the ability of wind and solar power generation to be connected to the grid.

[0003] In actual dispatching and operation, due to the random fluctuations in the power of renewable energy and loads, the global dispatching plan formulated recently often needs to be revised according to the actual operating conditions during the day. At this time, the amount of adjustable resources in the region determines the strength of the regional autonomy. If a region can use its internal adjustable resources to calibrate the dispatching plan in real time, so that the power of the regional interconnection line reaches the planned value given by the upper layer and achieves zero-difference control, it is said that the region can achieve autonomy.

[0004] At present, the research on regional autonomy and the optimization of park scheduling based on it is still relatively weak and cannot provide effective guidance for the operation of the existing network. Summary of the invention

[0005] In view of the gaps and deficiencies in the prior art, the present invention provides a method for optimizing the regional autonomy of distributed photovoltaic and energy storage based on a robust opportunity constraint model. First, a robust opportunity constraint for regional autonomy of distributed photovoltaic and energy storage is constructed through a regional autonomy measurement method based on distributedly robust optimization (DRO). On this basis, an active distribution network scheduling model is developed with the goal of minimizing the total cost and considering operating constraints such as unit ramping and main grid power interaction to achieve the optimization of regional autonomy.

[0006] The present invention specifically adopts the following technical solutions:

[0007] A method for optimizing regional autonomous capacity of distributed photovoltaic and energy storage based on a robust chance constraint model:

[0008] In the autonomous dispatching plan of distributed photovoltaic and energy storage regions, the power fluctuation caused by distributed photovoltaic and load is taken into account. The conditions for achieving autonomy are defined based on the power fluctuation in the region: at any time period, the maximum adjustable output of the region is greater than or equal to the total power fluctuation in the region;

[0009] In the optimization of autonomous capability, chance constraints and distributed robust optimization method based on Kullback-Leibler divergence are introduced to obtain a robust chance constraint model; and the distributed robust chance constraints are transformed into traditional chance constraints through the fuzzy set equivalent transformation formula, and the traditional chance constraints are solved by the risk value method;

[0010] Under the premise of taking into account the regional autonomy capabilities, regional dispatching operations are optimized with the goal of minimizing operating costs.

[0011] Furthermore, considering the power fluctuation caused by distributed photovoltaics and loads, assuming that a certain area contains m distributed photovoltaics and n loads, the power fluctuation amount in a certain period of time is in, is the distributed photovoltaic power fluctuation amount; is the load power fluctuation; is the total power fluctuation in the region during period t;

[0012] The maximum adjustable output in a certain period of time in the region is expressed as follows:

[0013]

[0014] Where: is the maximum adjustable output of the region at time t, are the maximum adjustable output of the i-th energy storage device and the j-th distributed generation in the region at time t, n e and n p Respectively represent the total number of energy storage devices and micro gas units in the area; are the maximum charging and discharging power of the energy storage device respectively; are the charging and discharging power of the energy storage device at time t respectively; are the maximum and minimum storage capacity of the i-th energy storage device respectively; is the storage capacity of the i-th energy storage device at time t; and are the upper and lower limits of the energy storage device capacity, δ s,i is the energy self-loss rate of the energy storage device;

[0015] η ch,i and η dis,i They are respectively the charging and discharging energy efficiency; and They are respectively the downward and upward climbing rate limits of the output of the micro gas unit; and are the upper and lower limits of the output of micro gas unit i, is the output value of distributed generation at time t; and They are respectively the charging and discharging power of the energy storage device; and are 0-1 variables, representing the charging and discharging states of the energy storage equipment; Δt is the unit scheduling duration;

[0016] To ensure that the region can achieve 100% autonomy, constraint (2) must be satisfied;

[0017]

[0018] Furthermore, the basic model of the opportunity constraint is:

[0019]

[0020] Where: x is the decision variable; ξ is the random variable; g(x,ξ) is the constraint function, Pr{A} is the probability of event A occurring, and δ is the probability of the inequality constraint failing;

[0021] The Kullback-Leibler divergence is used to describe the distance d between the actual probability distribution density function f and the empirical probability distribution density function f0 of the uncertain factor. KL (f||f0), described by equation (4);

[0022]

[0023] Where: d KL represents the Kullback-Leibler divergence between f and f0;

[0024] Further describe the actual exact probability distribution D of the uncertain factors as shown in formula (5);

[0025]

[0026] Where: C is the cumulative distribution function of f; KL Indicates the tolerance value of Kullback-Leibler divergence;

[0027] From formula (5), the robust chance constraint can be derived as:

[0028]

[0029] Where inf* represents the infimum of a function*.

[0030] The robust chance constraint (6) can ensure that in all possible probability density distribution functions, the probability of an event satisfying the condition g(x,ξ)≤0 is not less than 1-δ.

[0031] Based on the Kullback-Leibler divergence, we further obtain the fuzzy set equivalent transformation formula shown in formula (7), thereby converting the distributed robust chance constraint formula (6) into the traditional chance constraint;

[0032]

[0033] Where: Pr 0 {A} represents the probability of event A occurring under the empirical probability distribution function; δ 1+ is the reliability correction value, and e is a natural constant.

[0034] Furthermore, the use of the risk value method to solve the traditional opportunity constraint is specifically as follows:

[0035] The risk value VaR of the constraint function g(x,ξ) containing decision variables x and random variables ξ is defined as:

[0036] VaR β [g(x,ξ)]=min{Pr{g(x,ξ)≤0}≥β} (8)

[0037] Where: β is the degree of risk aversion;

[0038] The value of VaR represents the maximum loss suffered by the constraint function g(x,ξ) under a given confidence level β. Referring to formula (8), the chance constraint inequality in formula (7) is equivalent to formula (9), that is, the proportion of all random variable scenarios that may lead to g(x,ξ)≥0 should not exceed δ 1+ ; By randomly selecting s scenes: ξ 1 ,ξ 2 ,…,ξ s , and define a 0-1 auxiliary variable U c (1),U c (2),…,U c (s), to characterize the effectiveness of all chance constraint inequality scenarios, when U c (k) = 1 indicates the scene ξ k The opportunity constraint inequality is invalid; therefore, formula (9) is equivalent to formula (10);

[0039]

[0040] g(x,ξ k )≤MU c (k) (10)

[0041] Where: M is a sufficiently large number. c When (k) = 0, g(x,ξ k )<0, when U cWhen (k) = 1, g(x,ξ k ) is almost unconstrained;

[0042] Then the sample average approximation equivalent form of formula (7) is:

[0043]

[0044] Formula (11) belongs to mixed integer linear constraints and can be solved by calling the commercial solver gurobi through Matlab; when x in formula (11) is ξ is When , formula (11) is expressed as a regional autonomy constraint, where the size of δ represents the strength of regional autonomy.

[0045] Furthermore, δ is introduced into the dispatching plan of distributed photovoltaic and energy storage regional autonomy to realize the evaluation of regional autonomy capability.

[0046] Furthermore, under the premise of taking into account the regional autonomy capability, the regional dispatching operation is optimized with the goal of minimizing the operating cost. Specifically, with the goal of minimizing the operating cost, an active distribution network dispatching model including energy storage equipment operation constraints, main grid power interaction constraints, micro gas unit operation constraints, distribution network park regional autonomy capability constraints, and regional maximum adjustable output constraints is considered to achieve the optimization of regional autonomy capability.

[0047] Furthermore, the objective function of the active distribution network scheduling model is shown in formula (12), which includes minimizing the production capacity cost, equipment operation and maintenance cost, and main network power exchange cost:

[0048]

[0049] in:

[0050]

[0051] Where: F fuel is the park capacity cost, Output of micro gas unit, n p is the total number of micro gas units; F om The equipment operation and maintenance cost is Output for equipment, n e is the total number of devices; F trade is the main network power exchange cost, and are the electricity purchased and sold by the park to the main grid respectively; T is the total number of dispatching periods; c is the cost coefficient corresponding to P.

[0052] Furthermore, the energy storage device operation constraints are defined as follows:

[0053]

[0054] Where: are the charging and discharging variables of the energy storage equipment, respectively, where subscripts 1, 2, and 3 represent the planned output value, upward adjustment margin, and downward adjustment margin, respectively; is the capacity status of the energy storage device, They are the upward and downward adjustment margins of the capacity status of the energy storage equipment;

[0055] Formula (14) defines the current available capacity change relationship of the energy storage device; Formula (15) defines the capacity constraint of the energy storage device, which constrains the charge and discharge depth of the energy storage device while ensuring the periodic availability of the energy storage device; Formula (16) defines the charge and discharge constraints of the energy storage device; Formula (17) defines the margin expression for the increase and decrease of the capacity of the energy storage device; Since the state change of the energy storage device is related to the timing, it is necessary to add the energy storage device capacity constraint shown in Formula (18) to ensure the adjustability of the capacity state of the energy storage device in adjacent time periods, so as to ensure that the capacity fluctuation limit value of the energy storage device in the current time period can be adjusted to the fluctuation limit value of the energy storage device in the next time period;

[0056] The main grid power interaction constraint is shown in formula (19), where the 0-1 variable constrains the park from purchasing and selling electricity at the same time;

[0057]

[0058] Where: and is the upper limit of electricity purchase and sale. and It is a 0-1 variable, indicating the state of buying or selling electricity;

[0059] The micro gas unit operation constraints are defined as follows:

[0060]

[0061] Where: is the output value of the micro gas unit, They are the margin for upward and downward adjustment of the output of the micro gas unit; Scheduling ramping for micro gas units;

[0062] Constraints on the regional autonomy of distribution network parks:

[0063]

[0064] Where: It represents the total power fluctuation generated by the distribution network park under scenario k in period t. Respectively represent the power fluctuation of load and distributed photovoltaic output. When it is greater than 0, it means that the distribution network park needs to increase power, and when it is less than 0, it means that the distribution network park needs to reduce power.

[0065] In formula (24), d KL Calculated by formula (25), where M s Represents the total amount of available historical data. The larger its value, the closer the reference distribution is to the true distribution.

[0066]

[0067] Where: represents the chi-square distribution with s-1 degrees of freedom about a * The upper quantile of the distribution is used to ensure that the true distribution can be obtained with a confidence level of 1-a * be included in the reference set;

[0068] The maximum adjustable output constraint of the region is the maximum adjustable output constraint that takes into account the timing relationship, and is expressed as follows:

[0069]

[0070] The energy conservation constraint of the distribution network park is as follows;

[0071]

[0072] Where: is the predicted output value of the i-th distributed photovoltaic unit at time t, and m is the total amount of distributed photovoltaic units in the park; is the predicted value of the i-th load demand in the park at time t.

[0073] And, an electronic device, comprising a memory, a processor and a computer program stored in the memory and executable on the processor, characterized in that when the processor executes the program, the steps of the method for optimizing the regional autonomous capacity of distributed photovoltaic and energy storage based on a robust opportunity constraint model as described above are implemented.

[0074] A non-transitory computer-readable storage medium having a computer program stored thereon, characterized in that when the computer program is executed by a processor, the steps of the distributed photovoltaic and energy storage regional autonomous capacity optimization method based on a robust opportunity constraint model as described above are implemented.

[0075] Compared with the prior art, the present invention and its preferred solution construct a robust opportunity constraint for regional autonomy of distributed photovoltaic and energy storage by measuring the regional autonomy capability based on distributionally robust optimization (DRO). On this basis, an active distribution network scheduling model is adopted with the goal of minimizing the total cost and considering the operation constraints such as unit ramping and main grid power interaction to achieve the optimization of regional autonomy capability. Through the implementation of this solution, the compensation capability of energy storage equipment can be more fully utilized, and the adjustable resources of energy storage equipment can be more flexibly called. BRIEF DESCRIPTION OF THE DRAWINGS

[0076] The present invention is further described in detail below with reference to the accompanying drawings and specific embodiments:

[0077] Figure 1 This is a basic structural diagram of a distribution network park according to an embodiment of the present invention;

[0078] Figure 2 The structural diagram of the 33-node distribution network park is modified for the embodiment of the present invention;

[0079] Figure 3 A schematic diagram of predicted values ​​of distributed photovoltaic output and load according to an embodiment of the present invention;

[0080] Figure 4 Schematic diagram of the park scheduling scheme according to an embodiment of the present invention, wherein (a) the method proposed by the present invention (b) the robust optimization method (c) does not take into account the regional autonomy capability;

[0081] Figure 5 Schematic diagram of the charge state of ESS1 in an embodiment of the present invention, wherein (a) the method proposed by the present invention (b) the robust optimization method (c) does not take into account the regional autonomy capability;

[0082] Figure 6 Schematic diagram of the charge state of ESS2 in an embodiment of the present invention, wherein (a) the method proposed by the present invention (b) the robust optimization method (c) does not take into account the regional autonomy capability;

[0083] Figure 7 Schematic diagram of the charge state of ESS3 in an embodiment of the present invention, wherein (a) the method proposed by the present invention, (b) the robust optimization method, and (c) the regional autonomy capability is not taken into account. DETAILED DESCRIPTION

[0084] Hereinafter, the specific embodiments of the present application will be described in detail with reference to the accompanying drawings. According to these detailed descriptions, those skilled in the art can clearly understand the present application and can implement the present application. Without violating the principles of the present application, the features in the various embodiments can be combined to obtain new implementations, or certain features in certain embodiments can be replaced to obtain other preferred implementations.

[0085] In order to make the features and advantages of the present invention more clearly understood, the following embodiments are specifically described in detail with reference to the accompanying drawings:

[0086] Methods for measuring regional autonomy

[0087] 1. Analysis of regional autonomy capabilities

[0088] In actual dispatching and operation, due to the random fluctuations in the power of renewable energy and loads, the global dispatching plan formulated recently often needs to be revised according to the actual operating conditions during the day. At this time, the amount of adjustable resources in the region determines the strength of the regional autonomy. If a region can use its internal adjustable resources to calibrate the dispatching plan in real time, so that the power of the regional interconnection line reaches the planned value given by the upper layer and achieves zero-difference control, it is said that the region can achieve autonomy.

[0089] The present invention considers the power fluctuation caused by distributed photovoltaic and load. Assuming that a certain area contains m distributed photovoltaic and n load, the power fluctuation amount in the area in a certain period is in, is the distributed photovoltaic power fluctuation amount; is the load power fluctuation; Then it is the total power fluctuation in the region during time period t. From the definition of regional autonomy, it can be seen that the condition for the region to achieve autonomy is: in any time period, the maximum adjustable output of the region is greater than or equal to the total power fluctuation of the region. Under the conditions of considering the compensation capacity of energy storage equipment and the adjustable output of distributed power sources, without considering the timing relationship, the maximum adjustable output of the region in a certain period of time can be expressed as follows:

[0090]

[0091] Where: is the maximum adjustable output at time t in this area, are the maximum adjustable outputs of the i-th energy storage device and the j-th distributed generation in the region at time t, n e and n p Respectively represent the total number of energy storage devices and micro gas units in the area; are the maximum charging and discharging power of the energy storage device respectively;

[0092] and are the upper and lower limits of the energy storage device capacity, δ s,i is the energy self-loss rate of the energy storage device; η ch,i and η dis,i They are respectively the charging and discharging energy efficiency; and They are respectively the downward and upward ramp rate limits of the gas unit output; and are the upper and lower limits of the output of micro gas unit i, is the output value of distributed generation at time t; and They are respectively the charging and discharging power of the energy storage device; and are 0-1 variables, representing the charging and discharging status of the energy storage equipment respectively; Δt is the unit scheduling time.

[0093] To ensure that the region can achieve 100% autonomy, constraint (2) must be satisfied.

[0094]

[0095] 2. Robust Opportunity Constraints on Regional Autonomy

[0096] From formula (2), it can be seen that if the region is to achieve 100% autonomy, the scheduling plan needs to ensure that at any time, the regional adjustable resources can cope with extreme situations such as the minimum fluctuation of distributed photovoltaic output and the maximum fluctuation of load output, which obviously does not meet the economic requirements of the actual operation of the distribution network. Therefore, in order to take into account both economy and regional autonomy, the present invention introduces opportunity constraints in the optimization of autonomy.

[0097] As an important branch of stochastic optimization, chance constraints can solve optimization problems with uncertain factors at a given confidence level. Its basic model is:

[0098]

[0099] Where: x is the decision variable; ξ is the random variable; g(x,ξ) is the constraint function, Pr{A} is the probability of event A occurring, and δ is the probability of the inequality constraint failing.

[0100] By establishing a chance constraint model as shown in formula (3), statistical characteristics such as the probability density function of the random variable ξ are indispensable. However, in actual operation, due to the quantity and quality of data, accurate statistical characteristics are often not available, which makes it difficult to construct a chance constraint model. To solve this problem, traditional stochastic optimization methods generally use empirical probability distribution models to describe the uncertainty of distributed photovoltaic output and load demand. For example, assuming that distributed photovoltaic output obeys Beta distribution and load demand obeys normal distribution, although the model construction problem is solved, the robustness and accuracy of the model are greatly reduced, and the practical feasibility of the obtained solution cannot be guaranteed. To this end, the present invention introduces a distributed robust optimization method based on Kullback-Leibler divergence, which reduces the modeling difficulty of the stochastic optimization method by finding the worst probability distribution of uncertain factors, thereby combining the advantages of stochastic optimization and robust optimization.

[0101] The Kullback-Leibler divergence is used to describe the distance d between the actual probability distribution density function f and the empirical probability distribution density function f0 of the uncertain factor. KL (f||f0), whose value can be described by formula (4).

[0102]

[0103] Where: d KL represents the Kullback-Leibler divergence between f and f0.

[0104] d defined by the Kullback-Leibler divergence KL , can describe the actual exact probability distribution D of the uncertain factors as shown in formula (5).

[0105]

[0106] Where: C is the cumulative distribution function of f; KL Indicates the tolerance value of Kullback-Leibler divergence.

[0107] By adjusting d KL The parameters can change the range of the actual probability distribution density function f. Therefore, the failure probability of the inequality constraint will become the failure probability evaluation of a cluster of probability density functions. From the perspective of robust optimization, it is necessary to ensure that among all possible probability density distribution functions, the worst case also satisfies equation (6).

[0108]

[0109] Based on the Kullback-Leibler divergence, we can further obtain the fuzzy set equivalent transformation formula shown in formula (7), thereby converting the distributed robust chance constraint formula (6) into a traditional chance constraint.

[0110]

[0111] Where: Pr 0 {A} represents the probability of event A occurring under the empirical probability distribution function; δ 1+ is the reliability correction value.

[0112] Formula (7) is used to achieve the purpose of fitting the actual precise probability distribution based on the empirical probability distribution, which effectively reduces the difficulty of building the stochastic optimization model. The transformed robust chance constraint formula (7) has good robustness and accuracy.

[0113] 3. Chance Constrained Programming Method Based on Risk Value

[0114] The distributed robust chance constraint can be transformed into a traditional chance constraint by using the Kullback-Leibler divergence. The present invention uses the risk value method to solve the obtained traditional chance constraint, thereby realizing the solution of the robust chance constraint. The value at risk (VaR) is a mainstream risk measurement framework. For the function g(x,ξ) containing the decision variable x and the random variable ξ, the VaR can be defined as:

[0115] VaR β [g(x,ξ)]=min{Pr{g(x,ξ)≤0}≥β} (8)

[0116] Where: β is the degree of risk aversion.

[0117] The value of VaR represents the maximum loss suffered by the function g(x,ξ) under a given confidence level β. Referring to formula (8), the chance constraint inequality in formula (7) can be equivalent to formula (9), that is, the proportion of all random variable scenarios that may lead to g(x,ξ)≥0 should not exceed δ 1+ By randomly selecting s scenes: ξ 1 ,ξ 2 ,…,ξ s , and define a 0-1 auxiliary variable U c (1),U c (2),…,U c (s), to characterize the effectiveness of all chance constraint inequality scenarios, when U c (k) = 1 indicates the scene ξ k The chance constraint inequality is invalid. Therefore, equation (9) can be equivalent to (10).

[0118]

[0119] g(x,ξ k )≤MU c (k) (10)

[0120] Where: M is a sufficiently large number.

[0121] Then the sample average approximation equivalent form of formula (7) is:

[0122]

[0123] Formula (11) belongs to mixed integer linear constraints, which can be quickly solved by calling the commercial solver gurobi through Matlab, thus solving the problem of solving distributed robust opportunity constraints. When x in formula (11) is ξ is When , formula (11) is expressed as a regional autonomy constraint, where the size of δ is the strength of the regional autonomy. By introducing it into the scheduling plan, the regional autonomy can be examined.

[0124] Park scheduling optimization method considering regional autonomy capabilities

[0125] 1. Basic structure of the park

[0126] This embodiment is aimed at Figure 1 The basic structure of the distribution network park shown in the figure optimizes its dispatching operation with the goal of minimizing operating costs, taking into account the regional autonomy. Among them, the energy storage equipment can store or release electric energy according to the operating conditions to cope with the random fluctuations of distributed photovoltaic output and load; distributed power sources and distributed energy storage can supply power to the superior power grid if there is surplus capacity under the premise of ensuring regional autonomy to obtain better economic benefits; when the park cannot guarantee its autonomy, it can purchase electricity from the superior power grid to maintain the stable power supply of the park.

[0127] 2. Stochastic Optimization Model of Distribution Network Park Considering Regional Autonomy Capability

[0128] 1) Objective function

[0129] The distribution network park scheduling operation objective function constructed in this embodiment is shown in formula (12), which includes minimizing the production capacity cost, equipment operation and maintenance cost, and main network power exchange cost.

[0130]

[0131] in:

[0132]

[0133] Where: F fuel is the park capacity cost, Output of micro gas unit, n p is the total number of micro gas units; F om The equipment operation and maintenance cost is Output for equipment, n e is the total number of devices; F trade is the main network power exchange cost, and are the electricity purchased and sold by the park to the main grid respectively; T is the total number of dispatching periods; and c is the corresponding cost coefficient.

[0134] 2) Energy storage equipment operation constraints

[0135] The operation constraints of energy storage equipment are defined as follows:

[0136]

[0137] Where: are the charging and discharging variables of the energy storage equipment, respectively, where subscripts 1, 2, and 3 represent the planned output value, upward adjustment margin, and downward adjustment margin, respectively; is the capacity status of the energy storage device, They are the upward and downward adjustment margins respectively.

[0138] Formula (14) defines the change relationship of the current available capacity of the energy storage device. Formula (15) defines the capacity constraint of the energy storage device. While ensuring the periodic availability of the energy storage device, it constrains the charging and discharging depth of the energy storage device and protects the life of the energy storage device. Formula (16) defines the charging and discharging constraints of the energy storage device to avoid the phenomenon of simultaneous charging and discharging. Formula (17) defines the margin expression for the increase and decrease of the capacity of the energy storage device. Since the state change of the energy storage device is related to the timing, it is necessary to add the energy storage device capacity constraint shown in Formula (18) to ensure the adjustability of the capacity state of the energy storage device between adjacent time periods, and to ensure that the capacity fluctuation limit value of the energy storage device in the current time period can be adjusted to the fluctuation limit value of the energy storage device in the next time period.

[0139] 3) Main network power interaction constraints

[0140] The main grid power interaction constraint is shown in formula (19), where the 0-1 variable constrains the park from purchasing and selling electricity at the same time.

[0141]

[0142] Where: and is the upper limit of electricity purchase and sale. and It is a 0-1 variable, indicating the state of purchasing or selling electricity.

[0143] 4) Operation constraints of micro gas turbines

[0144] The operating constraints of the micro gas unit are defined as follows:

[0145]

[0146] Where: is the unit output value, They are the margin for upward and downward adjustment of unit output respectively; Schedule the ramp amount for the unit.

[0147] 5) Regional autonomy constraints

[0148] Combining the above regional autonomy analysis, the distributed robust optimization transformation method based on Kullback-Leibler divergence, and the opportunity constraint planning method based on VaR, the regional autonomy constraint of the distribution network park can be obtained as shown in the following formula:

[0149]

[0150] Where: It represents the total power fluctuation generated by the distribution network park under scenario k in period t. They represent the power fluctuation of load and distributed photovoltaic output respectively. When it is greater than 0, it means that the distribution network park needs to increase power. When it is less than 0, it means that the distribution network park needs to reduce power.

[0151] In formula (24), d KL It can be calculated by formula (25), where M s It represents the total amount of available historical data. The larger its value is, the closer the obtained reference distribution is to the true distribution.

[0152]

[0153] Where: represents the chi-square distribution with s-1 degrees of freedom about a * The upper quantile of , ensures that the true distribution can be obtained with confidence level 1-a * are included in the reference set.

[0154] 6) Regional maximum adjustable output constraints

[0155] Combined with the above regional autonomy analysis, the maximum adjustable output constraint considering the timing relationship can be expressed as follows:

[0156]

[0157] 7) Energy conservation constraints

[0158] The energy conservation constraint of the distribution network park is shown in the following formula.

[0159]

[0160] Where: is the predicted output value of the i-th distributed photovoltaic unit at time t, and m is the total amount of distributed photovoltaic units in the park; is the predicted value of the i-th load demand in the park at time t.

[0161] Case Analysis

[0162] 1. Case study

[0163] This embodiment is based on the IEEE 33 node network. Figure 2 The distribution network park shown in the figure is simulated and analyzed. The distribution network park contains 3 distributed power sources, 3 sets of energy storage equipment and 3 distributed photovoltaic generators, and maintains contact with the main grid. It can buy or sell electricity from the main grid. The price follows the time-of-use electricity price shown in Table 1. The load and distributed photovoltaic output forecast values ​​in each period in the park are shown in Table 1. Figure 3 As shown, the present invention assumes that the distributed photovoltaic output and load deviation obey the normal distribution, with a mean of 0 and a standard deviation of 20% of the predicted value. This embodiment uses Monte Carlo simulation to randomly generate 10,000 scenarios as available historical data.

[0164] Table 1 Time-of-use electricity price

[0165]

[0166] The parameter configurations of distributed power supply and energy storage equipment are shown in Table 2 and Table 3 respectively.

[0167] Table 2 Distributed power supply operation parameters

[0168]

[0169] Table 3 Energy storage equipment operating parameters

[0170]

[0171] 3.3.1.4.2 Method performance analysis

[0172] For the distance d defined based on the Kullback-Leibler divergence KL and the modified robust chance constraint model probability requirement δ 1+ (the original probability requires δ to be 0.2), the present invention conducted a study as shown in Table 4, and analyzed the quantiles a of the two on different chi-square distributions. * and degrees of freedom N f The value below.

[0173] Table 4 d under different conditions KL and δ 1+ Value

[0174]

[0175] As shown in Table 4, the upper quantile a of the chi-square distribution * Within a certain range, the corrected allowable failure probability δ 1+ The influence of can be almost ignored. The present invention selects a * =0.1 Conduct follow-up research; d KL With N f increases with the increase of 1+ With N f decreases with the increase of f The smaller the d KL The smaller the value, the smaller the range of correction required, and the closer the true probability distribution is to the probability distribution of empirical assumptions. 1+ The larger the value, the less comprehensive the probability information of each randomly selected scene will be. f =10, 100 and 200, the fluctuation of each random scene is very likely to be too small, which is difficult to reflect the actual operation fluctuation of the park. Therefore, in order to correctly reflect the probability information in the historical data and improve the reliability of the robust chance constraint model, the present invention selects N f =1000 for follow-up research.

[0176] The present invention compares the proposed distribution network park day-ahead optimization scheduling results based on the robust chance constraint model with the results obtained by the traditional random programming and robust optimization based on the chance constraint model, and the results are shown in Table 5.

[0177] Table 5 Comparison of optimization results before

[0178] Method Type Total dispatch cost / $ Probability of achieving regional autonomy Proposed method 10750.53 90.8％ Stochastic Programming 10697.74 80％ Robust Optimization 14831.24

[0179] Depend on Figure 4It can be seen that based on the economic benefits of time-of-use electricity prices, the park will purchase electricity from the main grid during low electricity price periods and sell electricity to the main grid during high electricity price periods. Among them, in the case of ignoring the regional autonomy capability and in the robust optimization scheme that only considers the worst case, the park can satisfy the energy conservation relationship in a single scenario by purchasing more electricity from the main grid during low electricity price periods. However, the method proposed in the present invention cannot purchase more electricity during low electricity price periods because it needs to ensure the autonomy of the park in multiple scenarios so that each unit has sufficient margin for up and down adjustment. In high electricity price periods, the park scheduling scheme that does not consider the regional autonomy capability can sell electricity to the main grid to the greatest extent to obtain the maximum economic benefits; the method proposed in the present invention needs to take into account the requirements of regional autonomy capability in multiple scenarios, so the remaining capacity is small and the electricity sales volume is low; the robust optimization method needs to ensure the balance of energy supply and demand in the worst scenario, with the least remaining capacity, the lowest electricity sales volume, and the worst economic benefits. During the peak load period (t=8), when the regional autonomy is not taken into account, the park can meet the load demand through the output of each unit and the discharge of energy storage equipment; when the regional autonomy is taken into account, the method proposed in the present invention needs to purchase electricity from the main grid and retain the adjustment margin of the energy storage equipment to absorb the random fluctuations in various scenarios. The robust optimization rule needs to purchase more electricity and fully discharge the energy storage equipment to meet the load demand in the worst scenario. The specific working status of the energy storage equipment under the three scheduling schemes is as follows: Figure 5-Figure 7 shown.

[0180] Depend on Figure 5-Figure 7 It can be seen that, without considering the regional autonomy capability, the energy storage device only releases energy during the peak load period to meet the load requirements of the park, and charges at the end of the scheduling cycle to meet the periodic adjustability of the energy storage device. The robust optimization scheduling scheme needs to meet the load requirements of the park in the worst scenario, so the discharge intensity of the energy storage device is greater than that of the energy storage device without considering the regional autonomy capability, and continues to discharge after the peak load period to meet the scheduling and operation needs of the park. However, overall, the charge state of the energy storage device in the above two scheduling schemes changes relatively smoothly and does not fluctuate much, indicating that the compensation capacity of the energy storage device is not fully utilized when the regional autonomy capability is not taken into account and in the robust optimization scheduling scheme, and its flexible and adjustable resources are not fully called. In the method proposed in the present invention, since it is necessary to ensure the autonomous operation capability of the park in multiple scenarios and leave sufficient margin for up and down adjustments, the energy storage device works more frequently, not only releasing energy during the peak load period, but also exerting its flexible adjustment capability through charging and discharging energy in the remaining time periods, so that the compensation capacity of the energy storage device is fully utilized. Therefore, compared with the method that ignores regional autonomy and robust optimization, the method proposed in the present invention can give full play to the compensation capacity of the energy storage device and more flexibly call the adjustable resources of the energy storage device.

[0181] Based on the same inventive concept, the present invention also provides a computer device, which includes: one or more processors, and a memory for storing one or more computer programs; the program includes program instructions, and the processor is used to execute the program instructions stored in the memory. The processor may be a central processing unit (CPU), or other general-purpose processors, digital signal processors (DSP), application-specific integrated circuits (ASIC), field-programmable gate arrays (FPGA) or other programmable logic devices, discrete gates or transistor logic devices, discrete hardware components, etc. It is the computing core and control core of the terminal, which is used to implement one or more instructions, specifically for loading and executing one or more instructions in a computer storage medium to implement the above method.

[0182] It needs to be further explained that, based on the same inventive concept, the present invention also provides a computer storage medium, on which a computer program is stored, and the computer program is executed by a processor to execute the above method. The storage medium can adopt any combination of one or more computer-readable media. The computer-readable medium can be a computer-readable signal medium or a computer-readable storage medium. The computer-readable storage medium can be, for example, but not limited to, an electrical, magnetic, optical, electrical, magnetic, infrared, or semiconductor system, device or device, or any combination of the above. More specific examples (non-exhaustive list) of computer-readable storage media include: an electrical connection with one or more wires, a portable computer disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the above. In the present invention, a computer-readable storage medium can be any tangible medium containing or storing a program, which can be used by an instruction execution system, device or device or used in combination with it.

[0183] In the description of this specification, the description with reference to the terms "one embodiment", "example", "specific example", etc. means that the specific features, structures, materials or characteristics described in conjunction with the embodiment or example are included in at least one embodiment or example of the present disclosure. In this specification, the schematic representation of the above terms does not necessarily refer to the same embodiment or example. Moreover, the specific features, structures, materials or characteristics described can be combined in any one or more embodiments or examples in a suitable manner.

[0184] The above shows and describes the basic principles, main features and advantages of the present disclosure. Those skilled in the art should understand that the present disclosure is not limited by the above embodiments, and the above embodiments and descriptions are only for explaining the principles of the present disclosure. Without departing from the spirit and scope of the present disclosure, the present disclosure may have various changes and improvements, and these changes and improvements fall within the scope of the present disclosure to be protected.

[0185] The present invention is not limited to the above-mentioned optimal implementation mode. Anyone can derive other forms of distributed photovoltaic and energy storage regional autonomous capacity optimization methods based on robust opportunity constraint models under the inspiration of the present invention. All equal changes and modifications made according to the scope of the patent application of the present invention should fall within the scope of the present invention.

Claims

1. A method for optimizing regional autonomous capacity of distributed photovoltaic and energy storage based on a robust chance constraint model, characterized by: In the autonomous dispatching plan of distributed photovoltaic and energy storage regions, the power fluctuation caused by distributed photovoltaic and load is taken into account. The conditions for achieving autonomy are defined based on the power fluctuation in the region: at any time period, the maximum adjustable output of the region is greater than or equal to the total power fluctuation in the region; In the optimization of autonomous capability, chance constraints and distributed robust optimization method based on Kullback-Leibler divergence are introduced to obtain a robust chance constraint model; and the distributed robust chance constraints are transformed into traditional chance constraints through the fuzzy set equivalent transformation formula, and the traditional chance constraints are solved by the risk value method; Under the premise of taking into account the regional autonomy capabilities, regional dispatching operations are optimized with the goal of minimizing operating costs.

2. The method for optimizing the regional autonomous capacity of distributed photovoltaic and energy storage based on a robust opportunity constraint model according to claim 1, characterized in that: Considering the power fluctuation caused by distributed photovoltaic and load, assuming that a certain area contains m distributed photovoltaic and n load, the power fluctuation in a certain period of time is in, is the distributed photovoltaic power fluctuation amount; is the load power fluctuation; is the total power fluctuation in the region during period t; The maximum adjustable output in a certain period of time in the region is expressed as follows: Where: is the maximum adjustable output of the region at time t, are the maximum adjustable output of the i-th energy storage device and the j-th distributed generation in the region at time t, n e and n p Respectively represent the total number of energy storage devices and micro gas units in the area; are the maximum charging and discharging power of the energy storage device respectively; are the charging and discharging power of the energy storage device at time t respectively; are the maximum and minimum storage capacity of the i-th energy storage device respectively; is the storage capacity of the i-th energy storage device at time t; and are the upper and lower limits of the energy storage device capacity, δ s,i is the energy self-loss rate of the energy storage device; η ch,i and η dis,i They are respectively the charging and discharging energy efficiency; and They are respectively the downward and upward climbing rate limits of the output of the micro gas unit; and are the upper and lower limits of the output of micro gas unit i, is the output value of distributed generation at time t; and They are respectively the charging and discharging power of the energy storage device; and are 0-1 variables, representing the charging and discharging states of the energy storage equipment; Δt is the unit scheduling duration; To ensure that the region can achieve 100% autonomy, constraint (2) must be satisfied; 3. The method for optimizing the regional autonomous capacity of distributed photovoltaic and energy storage based on a robust opportunity constraint model according to claim 2 is characterized in that: The basic model of the opportunity constraint is: Where: x is the decision variable; ξ is the random variable; g(x,ξ) is the constraint function, Pr{A} is the probability of event A occurring, and δ is the probability of the inequality constraint failing; The Kullback-Leibler divergence is used to describe the distance d between the actual probability distribution density function f and the empirical probability distribution density function f0 of the uncertain factor. KL (f||f0), described by equation (4); Where: d KL represents the Kullback-Leibler divergence between f and f0; Further describe the actual exact probability distribution D of the uncertain factors as shown in formula (5); Where: C is the cumulative distribution function of f; KL Indicates the tolerance value of Kullback-Leibler divergence; The robust chance constraint is derived from formula (5): Where: inf* represents the infimum of function*; The robust chance constraint (6) is used to ensure that in all possible probability density distribution functions, the probability of an event satisfying the condition g(x,ξ)≤0 is not less than 1-δ; Based on the Kullback-Leibler divergence, the fuzzy set equivalent transformation formula shown in formula (7) is further obtained, thereby converting the distributed robust chance constraint formula (6) into a traditional chance constraint; Where: Pr 0 {A} represents the probability of event A occurring under the empirical probability distribution function; δ 1+ is the reliability correction value, and e is a natural constant.

4. The method for optimizing regional autonomous capacity of distributed photovoltaic and energy storage based on a robust opportunity constraint model according to claim 3 is characterized by: The method of using the value-at-risk method to solve the traditional opportunity constraint is specifically as follows: The risk value VaR of the constraint function g(x,ξ) containing decision variables x and random variables ξ is defined as: VaR β [g(x,ξ)]=min{Pr{g(x,ξ)≤0}≥β} (8) Where: β is the degree of risk aversion; The value of VaR represents the maximum loss suffered by the constraint function g(x,ξ) under a given confidence level β. Referring to formula (8), the chance constraint inequality in formula (7) is equivalent to formula (9), that is, the proportion of all random variable scenarios that may lead to g(x,ξ)≥0 should not exceed δ 1+ ; By randomly selecting s scenes: ξ 1 ,ξ 2 ,…,ξ s , and define a 0-1 auxiliary variable U c (1),U c (2),…,U c (s), to characterize the effectiveness of all chance constraint inequality scenarios, when U c (k) = 1 indicates the scene ξ k The opportunity constraint inequality is invalid; therefore, formula (9) is equivalent to formula (10); g(x,ξ k )≤MU c (k) (10) Where: M is a sufficiently large number; Then the sample average approximation equivalent form of formula (7) is: Formula (11) belongs to mixed integer linear constraints and can be solved by calling the commercial solver gurobi through Matlab; when x in formula (11) is ξ is When , formula (11) is expressed as a regional autonomy constraint, where the size of δ represents the strength of regional autonomy.

5. The method for optimizing regional autonomous capacity of distributed photovoltaic and energy storage based on a robust chance constraint model according to claim 4 is characterized in that: δ is introduced into the dispatching plan of distributed photovoltaic and energy storage regional autonomy to realize the evaluation of regional autonomy capability.

6. The method for optimizing regional autonomous capacity of distributed photovoltaic and energy storage based on a robust chance constraint model according to claim 4 is characterized in that: The optimization of regional dispatching operation with the goal of minimizing operating costs under the premise of taking regional autonomy capabilities into account is specifically as follows: with the goal of minimizing operating costs, an active distribution network dispatching model is considered, including energy storage equipment operation constraints, main grid power interaction constraints, micro gas unit operation constraints, distribution network park regional autonomy constraints, and regional maximum adjustable output constraints, to achieve optimization of regional autonomy capabilities.

7. The method for optimizing regional autonomous capacity of distributed photovoltaic and energy storage based on a robust chance constraint model according to claim 6 is characterized by: The objective function of the active distribution network dispatching model is shown in formula (12), which includes minimizing the production capacity cost, equipment operation and maintenance cost, and main network power exchange cost: in: Where: F fuel is the park capacity cost, Output of micro gas unit, n p is the total number of micro gas units; F om The equipment operation and maintenance cost is Output for equipment, n e is the total number of devices; F trade is the main network power exchange cost, and are the electricity purchased and sold by the park to the main grid respectively; T is the total number of dispatching periods; c is the cost coefficient corresponding to P.

8. The method for optimizing regional autonomous capacity of distributed photovoltaic and energy storage based on a robust chance constraint model according to claim 7, characterized in that: The energy storage equipment operation constraints are defined as follows: Where: are the charging and discharging variables of the energy storage equipment, respectively, where subscripts 1, 2, and 3 represent the planned output value, upward adjustment margin, and downward adjustment margin, respectively; is the capacity status of the energy storage device, They are the upward and downward adjustment margins of the capacity status of the energy storage equipment; Formula (14) defines the current available capacity change relationship of the energy storage device; Formula (15) defines the capacity constraint of the energy storage device, which constrains the charge and discharge depth of the energy storage device while ensuring the periodic availability of the energy storage device; Formula (16) defines the charge and discharge constraints of the energy storage device; Formula (17) defines the margin expression for the increase and decrease of the capacity of the energy storage device; Since the state change of the energy storage device is related to the timing, it is necessary to add the energy storage device capacity constraint shown in Formula (18) to ensure the adjustability of the capacity state of the energy storage device in adjacent time periods, so as to ensure that the capacity fluctuation limit value of the energy storage device in the current time period can be adjusted to the fluctuation limit value of the energy storage device in the next time period; The main grid power interaction constraint is shown in formula (19), where the 0-1 variable constrains the park from purchasing and selling electricity at the same time; Where: and is the upper limit of electricity purchase and sale. and It is a 0-1 variable, indicating the state of buying or selling electricity; The micro gas unit operation constraints are defined as follows: Where: is the output value of the micro gas unit, They are the margin for upward and downward adjustment of the output of the micro gas unit; Scheduling ramping for micro gas units; Constraints on the regional autonomy of distribution network parks: Where: It represents the total power fluctuation generated by the distribution network park in scenario k during period t. Respectively represent the power fluctuation of load and distributed photovoltaic output. When it is greater than 0, it means that the distribution network park needs to increase power, and when it is less than 0, it means that the distribution network park needs to reduce power. In formula (24), d KL Calculated by formula (25), where M s Represents the total amount of available historical data. The larger its value, the closer the reference distribution is to the true distribution. Where: represents the chi-square distribution with s-1 degrees of freedom about a * The upper quantile of the distribution is used to ensure that the true distribution can be obtained with a confidence level of 1-a * be included in the reference set; The maximum adjustable output constraint of the region is the maximum adjustable output constraint that takes into account the timing relationship, and is expressed as follows: The energy conservation constraint of the distribution network park is as follows; Where: is the predicted output value of the i-th distributed photovoltaic unit at time t, and m is the total amount of distributed photovoltaic units in the park; is the predicted value of the i-th load demand in the park at time t.

9. An electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that: When the processor executes the program, the steps of the distributed photovoltaic and energy storage regional autonomous capacity optimization method based on the robust opportunity constraint model as described in any one of claims 1-8 are implemented.

10. A non-transitory computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the steps of the method for optimizing the regional autonomous capacity of distributed photovoltaic and energy storage based on a robust chance constraint model as described in any one of claims 1 to 8 are implemented.

Images