Power distribution network planning method and system considering flexible resources

Abstract

The invention discloses a power distribution network planning method and system considering flexible resources, and the method comprises the steps: building a model of the flexible resources, and building a power distribution network distributed energy storage planning model in a deterministic environment based on the model of the flexible resources. An IGDT model based on flexible resources and a power distribution network distributed energy storage planning model in a deterministic environment are used to construct an upper-layer power distribution network distributed energy storage planning model in an uncertain environment, and a lower-layer demand defense optimization model based on model prediction control is established. Iterative optimization solution is carried out on an upper-layer power distribution network distributed energy storage planning model in an uncertain environment and a lower-layer demand defense optimization model based on model prediction control to obtain an optimal planning result of the power distribution network, so that the economical efficiency of the power distribution network is improved, the robustness is considered, demand defense can be realized, and the power distribution network planning method is suitable for large-scale popularization and application. And the increase of the operation cost of the power distribution network in the real-time planning process is reduced, so that the economical efficiency and robustness of the power distribution network are improved, and demand defense is realized.

Description

Technical Field

[0001] The present invention relates to the technical field of distribution networks, and particularly to a distribution network planning method and system considering flexibility resources. Background Art

[0002] The grid connection of new energy sources such as photovoltaic (PV) has become the trend of energy development. However, the output of new energy sources is difficult to predict, with strong uncertainty and volatility. This seriously affects the supply-demand balance, safety, and economic operation of the distribution network. Although energy storage can smooth the fluctuations of new energy sources, reduce the demand charge by peak shaving and valley filling, and reduce the electricity charge by peak-valley arbitrage. However, the high upfront investment cost and long recovery period of energy storage seriously affect the application of energy storage, and the regulation ability of energy storage is limited by the capacity size. How to ensure the dynamic supply-demand balance of the distribution network has become an urgent problem to be solved.

[0003] It is worth noting that flexible loads, as an important part of demand-side response, can effectively improve the load power curve of the distribution network and reduce the peak value and operating cost of the load. Therefore, using flexible loads as a method to solve the dynamic supply-demand balance of the distribution network has milestone significance. As mentioned above, the flexibility regulation ability of energy storage is limited by the capacity size, and the regulation ability of flexible loads is also affected by factors such as user willingness. A single flexibility resource cannot fully cover the impact of PV uncertainty on the economy of the distribution network.

[0004] In the process of distribution network planning, the methods for dealing with uncertainty can be divided into fuzzy optimization, robust optimization, and stochastic optimization. Stochastic optimization relies on the probability density function of data to solve, generating a large number of scenarios to simulate uncertainty, which increases the computational complexity. Although fuzzy optimization captures the impact of uncertainty by setting membership parameters and user satisfaction, without relying on the distribution of the data itself, the solution obtained by the model has high subjectivity. Similarly, robust optimization also uses an uncertainty set to represent uncertainty, reducing the computational complexity, but the solution obtained is overly conservative and loses economic efficiency. Summary of the Invention

[0005] The technical problem to be solved by the present invention is: to provide a distribution network planning method and system considering flexibility resources, which can improve the economy and robustness of the distribution network and achieve demand defense.

[0006] To solve the above technical problem, a technical solution adopted by the present invention is:

[0007] A distribution network planning method considering flexibility resources, comprising the steps of:

[0008] Establish a model of flexibility resources, where the flexibility resources include distributed energy storage and flexible loads;

[0009] Build a distribution network distributed energy storage planning model in a deterministic environment based on the model of the flexibility resources;

[0010] Use the information gap decision theory to build an upper-layer distribution network distributed energy storage planning model in an uncertain environment based on the model of the flexibility resources and the distribution network distributed energy storage planning model in the deterministic environment, and establish a lower-layer demand defense optimization model based on model predictive control;

[0011] Perform iterative optimization on the upper-layer distribution network distributed energy storage planning model in the uncertain environment and the lower-layer demand defense optimization model based on model predictive control to obtain the optimal distribution network planning result.

[0012] To solve the above technical problems, another technical solution adopted by the present invention is:

[0013] A distribution network planning system considering flexibility resources, including a memory, a processor, and a computer program stored on the memory and executable on the processor. When the processor executes the computer program, the following steps are implemented:

[0014] Build a model of flexibility resources, where the flexibility resources include distributed energy storage and flexible loads;

[0015] Build a distribution network distributed energy storage planning model in a deterministic environment based on the model of the flexibility resources;

[0016] Use the information gap decision theory to build an upper-layer distribution network distributed energy storage planning model in an uncertain environment based on the model of the flexibility resources and the distribution network distributed energy storage planning model in the deterministic environment, and establish a lower-layer demand defense optimization model based on model predictive control;

[0017] Perform iterative optimization on the upper-layer distribution network distributed energy storage planning model in the uncertain environment and the lower-layer demand defense optimization model based on model predictive control to obtain the optimal distribution network planning result.

[0018] The beneficial effects of the present invention are as follows: establishing a model of flexible resources, constructing a distribution network distributed energy storage planning model in a deterministic environment based on the model of flexible resources, using the Information Gap Decision Theory (IGDT) to construct an upper-layer distribution network distributed energy storage planning model in an uncertain environment based on the model of flexible resources and the distribution network distributed energy storage planning model in a deterministic environment, and establishing a lower-layer demand defense optimization model based on model predictive control. Iteratively optimizing and solving the upper-layer distribution network distributed energy storage planning model in an uncertain environment and the lower-layer demand defense optimization model based on model predictive control to obtain the optimal distribution network planning result. Considering flexible resources including distributed energy storage and flexible loads, it can fully exploit the characteristics of various flexible resources, increase the economy of the distribution network, and take into account robustness. Moreover, the constructed upper-layer distribution network distributed energy storage planning model in an uncertain environment and the lower-layer demand defense optimization model based on model predictive control can track and feedback-correct the planning result in real time to ensure the realization of demand defense, reduce the increase in the operating cost of the distribution network during the real-time planning process, thereby improving the economy and robustness of the distribution network and achieving demand defense. BRIEF DESCRIPTION OF THE DRAWINGS

[0019] Figure 1 It is a flowchart of the steps of a distribution network planning method considering flexible resources according to an embodiment of the present invention;

[0020] Figure 2 It is a schematic structural diagram of a distribution network planning system considering flexible resources according to an embodiment of the present invention;

[0021] Figure 3 It is a schematic diagram of the IEEE 33-node structure in the distribution network planning method considering flexible resources according to an embodiment of the present invention;

[0022] Figure 4 It is a schematic diagram of the photovoltaic load output curve in the distribution network planning method considering flexible resources according to an embodiment of the present invention;

[0023] Figure 5 In (a), it is a schematic diagram of the change in grid trading power in the distribution network planning method considering flexible resources according to an embodiment of the present invention;

[0024] Figure 5 In (b), it is a schematic diagram of the change in DES power in the distribution network planning method considering flexible resources according to an embodiment of the present invention;

[0025] Figure 6 It is a schematic diagram of the change in the uncertainty radius and cost in the distribution network planning method considering flexible resources according to an embodiment of the present invention. DETAILED DESCRIPTION OF THE INVENTION

[0026] To describe the technical content, achieved objectives and effects of the present invention in detail, the following will be described in conjunction with the embodiments and with reference to the accompanying drawings.

[0027] Please refer to Figure 1 , a distribution network planning method considering flexibility resources, comprising the steps of:

[0028] Establish a model of flexibility resources, where the flexibility resources include distributed energy storage and flexible loads;

[0029] Build a distributed energy storage planning model for the distribution network under a deterministic environment based on the model of flexibility resources;

[0030] Use the information gap decision theory to build an upper-layer distributed energy storage planning model for the distribution network under an uncertain environment based on the model of flexibility resources and the distributed energy storage planning model for the distribution network under a deterministic environment, and establish a lower-layer demand defense optimization model based on model predictive control;

[0031] Iteratively optimize and solve the upper-layer distributed energy storage planning model for the distribution network under an uncertain environment and the lower-layer demand defense optimization model based on model predictive control to obtain the optimal planning result of the distribution network.

[0032] As can be seen from the above description, the beneficial effects of the present invention are as follows: establishing a model of flexibility resources, building a distributed energy storage planning model for the distribution network under a deterministic environment based on the model of flexibility resources, using IGDT to build an upper-layer distributed energy storage planning model for the distribution network under an uncertain environment based on the model of flexibility resources and the distributed energy storage planning model for the distribution network under a deterministic environment, and establishing a lower-layer demand defense optimization model based on model predictive control, iteratively optimizing and solving the upper-layer distributed energy storage planning model for the distribution network under an uncertain environment and the lower-layer demand defense optimization model based on model predictive control to obtain the optimal planning result of the distribution network, considering flexibility resources including distributed energy storage and flexible loads, being able to fully exploit the characteristics of various flexibility resources, increasing the economy of the distribution network, taking into account robustness, and the established upper-layer distributed energy storage planning model for the distribution network under an uncertain environment and the lower-layer demand defense optimization model based on model predictive control can track and feedback correct the planning result in real time to ensure that demand defense can be achieved, reduce the increase in the operating cost of the distribution network during the real-time planning process, thereby improving the economy and robustness of the distribution network and achieving demand defense.

[0033] Furthermore, the model of flexibility resources includes a distributed energy storage model, specifically:

[0034]

[0035] Wherein, represents the charging power of the k-th distributed energy storage at the t-th moment, represents the power of the k-th distributed energy storage, represents the discharging power of the k-th distributed energy storage at the t-th moment, represents the capacity of the k-th distributed energy storage, v represents the capacity coefficient of the k-th distributed energy storage, SOE k (T) represents the energy state of the k-th distributed energy storage at the T-th moment, T represents the total optimization time, SOE k (0) represents the energy state of the k-th distributed energy storage at the 0-th moment, SOE k (t) represents the energy state of the k-th distributed energy storage at the t-th moment, SOE k (t - 1) represents the energy state of the k-th distributed energy storage at the (t - 1)-th moment, η c represents the charging efficiency of the distributed energy storage, Δt represents the optimization time interval, η d represents the discharging efficiency of the distributed energy storage.

[0036] As can be seen from the above description, by establishing a distributed energy storage model, the distributed energy storage, as a flexible and fast regulation resource, can not only suppress the random fluctuations of distributed photovoltaics, but also improve the local consumption of distributed photovoltaics and reduce the burden on the power grid.

[0037] Furthermore, the model of the flexibility resource further includes a shiftable load model, a transferable load model, and a curtailable load model;

[0038] The shiftable load model is specifically as follows:

[0039]

[0040] Wherein, P i,Shift (t) represents the load power before shifting at the t-th moment of the i-th node, represents the load power after shifting at the t-th moment of the i-th node, α i,shift (t) represents the 0-1 variable indicating whether the shiftable load at the t-th moment of the i-th node is shifted, represents the end moment of the shiftable period, represents the start moment of the shiftable period, represents the maximum shiftable moment of the shiftable load of the i-th node, represents the total continuous operation time of the shiftable load of the i-th node;

[0041] The transferable load model is specifically as follows:

[0042]

[0043] Wherein, P i,Tran (t) represents the load power of the i-th node before transfer at the t-th moment, represents the load power of the i-th node after transfer at the t-th moment, and β i,Tran (t) represents a 0-1 variable indicating whether the shiftable load of the i-th node is transferred at the t-th moment, represents the minimum power value of the transferred load of the i-th node, represents the maximum power value of the transferred load of the i-th node, and β i,Tran (τ) represents a 0-1 variable indicating whether the shiftable load of the i-th node is transferred at the τ-th moment, represents the total continuous operation time of the shiftable load of the i-th node;

[0044] The shiftable load model is specifically as follows:

[0045]

[0046]

[0047] Wherein, P i,Cut (t) represents the load before curtailment of the i-th node at the t-th moment, represents the load curtailment coefficient of the i-th node at the t-th moment, and γ i,Cut (t) represents a 0-1 variable indicating whether the load of the i-th node is curtailed at the t-th moment, represents the load after curtailment of the i-th node at the t-th moment, represents the minimum continuous curtailment time, and γ i,Cut (t - 1) represents a 0-1 variable indicating whether the load of the i-th node is curtailed at the (t - 1)-th moment, represents the maximum continuous curtailment time, and N i,max represents the maximum curtailment times of the shiftable load of the i-th node.

[0048] As can be seen from the above description, flexible loads are divided into shiftable loads, transferable loads, and curtailable loads. Flexible loads refer to electrical loads in a power system that can actively adjust their power consumption time, power, or operation mode according to external signals. The participation of flexible loads in the distribution network planning is conducive to reducing the planned capacity of distributed energy storage, optimizing the power consumption behavior of loads, reducing the peak-valley difference of loads, and realizing the economic operation of the distribution network.

[0049] Furthermore, the distribution network distributed energy storage planning model constructed based on the flexibility resource model under a deterministic environment includes:

[0050] Establish an objective function to minimize the planning cost;

[0051] Establish the demand power constraint, power flow constraint, and power limit constraint for trading with the power grid, and obtain the constraint conditions according to the demand power constraint, the power flow constraint, the power limit constraint for trading with the power grid, and the model of the flexibility resources;

[0052] Generate a distribution network distributed energy storage planning model under a deterministic environment based on the objective function and the constraint conditions.

[0053] As can be seen from the above description, an objective function is established to minimize the planning cost, the constraint conditions are obtained according to the demand power constraint, the power flow constraint, the power limit constraint for trading with the power grid, and the model of the flexibility resources, and a distribution network distributed energy storage planning model under a deterministic environment is generated based on the objective function and the constraint conditions, which can make full use of the diversified regulation capabilities of flexibility resources such as distributed energy storage and flexible loads, and effectively cope with the challenges of the uncertainty of distributed photovoltaics to the operation reliability and economy of the distribution network.

[0054] Furthermore, the establishment of the objective function to minimize the planning cost is specifically as follows:

[0055]

[0056] In the formula, F represents the planning cost, F Grid represents the cost of trading with the power grid, F DES represents the planning cost of distributed energy storage, F Shift represents the compensation cost of shiftable loads, F Tran represents the compensation cost of transferable loads, F Cut represents the compensation cost of curtailable loads, F P2P represents the transmission cost, c b (t) represents the time-of-use electricity price, P b (t) represents the power purchased by the user from the power grid at the t-th moment, c s represents the feed-in tariff, P s (t) represents the power sold by the user to the power grid at the t-th moment, c dec represents the demand tariff, P bmax represents the demand power, N represents the number of distributed energy storage, represents the power of distributed energy storage, represents the capacity investment and construction cost of distributed energy storage, represents the operation and maintenance cost of distributed energy storage, I represents the number of nodes, c Shift represents the compensation cost coefficient of shiftable loads, c Tran represents the compensation cost coefficient of transferable loads, c Cut represents the compensation cost coefficient of curtailable loads, c P2P represents the transmission cost coefficient, Denote the power exchanged between the $i$-th node and the $j$-th node at the $t$-th moment.

[0057] As can be seen from the above description, establishing an objective function to minimize the planning cost, where the planning cost includes the transaction cost with the power grid, the planning cost of distributed energy storage, and the compensation cost of flexible loads, can ensure the economy of the system.

[0058] Furthermore, the demand power constraint is specifically:

[0059] Pb max =max[P b (t)];

[0060] The power flow constraint is specifically:

[0061]

[0062] In the formula, P i,Load (t) represents the load power of the $i$-th node at the $t$-th moment, represents the charging power of the $i$-th distributed energy storage at the $t$-th moment, represents the photovoltaic power of the $i$-th node at the $t$-th moment, represents the discharging power of the $i$-th distributed energy storage at the $t$-th moment, represents the base load power of the $i$-th node at the $t$-th moment;

[0063] The power limit constraint for trading with the power grid is specifically:

[0064] 0≤P b (t)≤Pb max ;

[0065] 0≤P s (t)≤Ps max ;

[0066] In the formula, P bmax represents the upper limit of the purchased power, and P smax represents the upper limit of the sold power.

[0067] As can be seen from the above description, establishing the demand power constraint, the power flow constraint, and the power limit constraint for trading with the power grid can ensure the safe operation of the distributed energy storage planning process.

[0068] Furthermore, the use of the information gap decision theory to construct the upper-layer distributed energy storage planning model for the distribution network under the uncertain environment based on the model of the flexibility resources and the distributed energy storage planning model for the distribution network under the deterministic environment includes:

[0069] Use the information gap decision theory to construct a distribution network distributed energy storage planning model under the initial uncertain environment, and use the model of the flexibility resources and the distribution network distributed energy storage planning model under the deterministic environment as the constraint conditions of the distribution network distributed energy storage planning model under the initial uncertain environment;

[0070] Based on robust optimization, transform the distribution network distributed energy storage planning model under the initial uncertain environment to obtain the upper-layer distribution network distributed energy storage planning model under the uncertain environment.

[0071] As can be seen from the above description, to address the impact of photovoltaic uncertainty on distributed energy storage planning, use IGDT to construct a distribution network distributed energy storage planning model under the initial uncertain environment, and then transform the distribution network distributed energy storage planning model under the initial uncertain environment based on robust optimization to obtain the upper-layer distribution network distributed energy storage planning model under the uncertain environment, which is convenient for direct solution.

[0072] Furthermore, the upper-layer distribution network distributed energy storage planning model under the uncertain environment is specifically:

[0073] maxmin[σ1,..., σ i ;

[0074] F ≤ (1 + ω)F0;

[0075]

[0076] In the formula, σ i represents the uncertainty radius of the i-th photovoltaic, F represents the optimal planning cost under the deterministic environment, F0 represents the optimal planning cost under the uncertain environment, ω represents the risk aversion factor, represents the power of the i-th photovoltaic, represents the expected power of the i-th photovoltaic.

[0077] As can be seen from the above description, the upper-layer distribution network distributed energy storage planning model under the uncertain environment effectively addresses photovoltaic uncertainty to improve system robustness.

[0078] Furthermore, the establishment of the lower-layer demand defense optimization model based on model predictive control is specifically:

[0079]

[0080] In the formula, F Lower represents the lower-layer demand defense optimization model based on model predictive control, H represents the model predictive control prediction time domain, represents the charging power of the k-th distributed energy storage in the lower layer at the h-th moment, denotes the charging power of the k-th distributed energy storage obtained from the upper-layer planning at the h-th moment, denotes the discharging power of the k-th distributed energy storage in the lower layer at the h-th moment, denotes the discharging power of the k-th distributed energy storage obtained from the upper-layer planning at the h-th moment, where Ω represents the set of nodes, denotes the power exchanged between the i-th node and the j-th node in the lower layer at the h-th moment, denotes the power exchanged between the i-th node and the j-th node obtained from the upper-layer planning at the h-th moment, P b,low (h) denotes the power purchased by users from the power grid at the h-th moment in the lower layer, P b (h) denotes the power purchased by users from the power grid at the h-th moment obtained from the upper-layer planning, P s,low (h) denotes the power sold by users to the power grid at the h-th moment, P s (h) denotes the power sold by users to the power grid at the h-th moment obtained from the upper-layer planning.

[0081] As can be seen from the above description, during real-time operation, the predicted value of photovoltaic power will have errors, resulting in deviations in the charging and discharging plans of distributed energy storage during the lower-layer planning process, leading to the failure of demand defense and thus increasing the operating cost of the distribution network. Therefore, to reduce the increase in the operating cost of the distribution network during real-time planning, it is necessary to perform demand defense on the distribution network. By establishing a demand defense optimization model based on model predictive control for the lower layer, model predictive control can utilize its characteristics of rolling optimization and feedback correction to track the upper-layer planning results in real time according to the actual output value of photovoltaic power and correct the output plans of current distributed energy storage, etc., reducing the probability of demand defense failure and thus effectively realizing demand defense.

[0082] Please refer to Figure 2 , another embodiment of the present invention provides a distribution network planning system considering flexibility resources, including a memory, a processor, and a computer program stored on the memory and executable on the processor. When the processor executes the computer program, it realizes each step in the above-mentioned distribution network planning method considering flexibility resources.

[0083] The above-mentioned distribution network planning method and system considering flexibility resources of the present invention can be applied to the distribution network planning scenario, which will be described through specific embodiments as follows:

[0084] Please refer to Figure 1 、 Figures 3 - 6 , the first embodiment of the present invention is as follows:

[0085] A distribution network planning method considering flexibility resources, including the steps of:

[0086] S1. Establish a model of flexibility resources.

[0087] Among them, the flexibility resources include distributed energy storage and flexible loads. The models of the flexibility resources include distributed energy storage models, shiftable load models, transferable load models, and curtailable load models.

[0088] The distributed energy storage model is specifically as follows:

[0089]

[0090] In the formula, represents the charging power of the k-th distributed energy storage at the t-th moment, represents the power of the k-th distributed energy storage, represents the discharging power of the k-th distributed energy storage at the t-th moment, represents the capacity of the k-th distributed energy storage, ε represents the capacity coefficient of the k-th distributed energy storage, SOE k (T) represents the energy state of the k-th distributed energy storage at the T-th moment, T represents the total optimization time, SOE k (0) represents the energy state of the k-th distributed energy storage at the 0-th moment, SOE k (t) represents the energy state of the k-th distributed energy storage at the t-th moment, SOE k (t - 1) represents the energy state of the k-th distributed energy storage at the (t - 1)-th moment, η c represents the charging efficiency of the distributed energy storage, Δt represents the optimized time interval, η d represents the discharging efficiency of the distributed energy storage.

[0091] The first and second formulas in formula (1) represent the charging and discharging restrictions on the k-th distributed energy storage, the third formula indicates whether the k-th distributed energy storage can charge and discharge simultaneously, the fourth and fifth formulas represent the power and capacity calculation formulas of the k-th distributed energy storage, the sixth formula represents that to ensure the recycling of the k-th distributed energy storage, the initial and final energy states of the distributed energy storage should be equal, the seventh formula indicates that the energy state of the k-th distributed energy storage should be maintained within and to extend the service life, and the eighth formula represents the calculation formula of the energy state SOE k (t) of the k-th distributed energy storage at the t-th moment.

[0092] When participating in the operation of the distribution network, the shiftable load needs to be shifted as a whole, and its power consumption time cannot be interrupted and the total power consumption time remains unchanged. The shiftable load model is specifically as follows:

[0093]

[0094] In the formula, P i,Shift(t) represents the load power of the i-th node before translation at the t-th moment, represents the load power of the i-th node after translation at the t-th moment, α i,shift (t) represents the 0-1 variable indicating whether the shiftable load of the i-th node is shifted at the t-th moment, represents the end time of the shiftable period, represents the start time of the shiftable period, represents the maximum shiftable time of the shiftable load of the i-th node, represents the total continuous operation time of the shiftable load of the i-th node.

[0095] In equation (2), the first formula represents that the total amount of load before and after translation remains unchanged, and the second formula ensures that the power consumption time of the shiftable load remains unchanged.

[0096] On the premise of ensuring the equality of the total load throughout the optimization period, the shiftable load can be flexibly adjusted at each moment. The specific shiftable load model is as follows:

[0097]

[0098] In the formula, P i,Tran (t) represents the load power of the i-th node before transfer at the t-th moment, represents the load power of the i-th node after transfer at the t-th moment, β i,Tran (t) represents the 0-1 variable indicating whether the transferable load of the i-th node is transferred at the t-th moment, represents the minimum power value of the transferred load of the i-th node, represents the maximum power value of the transferred load of the i-th node, β i,Tran (τ) represents the 0-1 variable indicating whether the transferable load of the i-th node is transferred at the τ-th moment, represents the total continuous operation time of the transferable load of the i-th node.

[0099] In equation (3), the first formula ensures that the total load demand of the transferable load remains unchanged before and after transfer, and the second and third formulas ensure that the equipment will not start and stop frequently, damaging the equipment state.

[0100] By reducing the node load, the curtailable load reduces the power consumption of users. By reasonably controlling the curtailable load, the peak load of the power system can be effectively reduced, improving the system economy. And to ensure the comfort of users, the curtailable load needs to meet the constraints of the minimum and maximum continuous curtailment time and the number of curtailments. The specific curtailable load model is as follows:

[0101]

[0102]

[0103] In the formula, P i,Cut (t) represents the load before curtailment of the i-th node at the t-th moment, represents the load curtailment coefficient of the i-th node at the t-th moment, γ i,Cut (t) represents a 0-1 variable indicating whether the load of the i-th node is curtailed at the t-th moment, represents the load after curtailment of the i-th node at the t-th moment, represents the minimum continuous curtailment time, γ i,Cut (t - 1) represents a 0-1 variable indicating whether the load of the i-th node is curtailed at the (t - 1)-th moment, represents the maximum continuous curtailment time, N i,max represents the maximum number of curtailments of the load that can be curtailed by the i-th node.

[0104] In formula (5), the first formula is the minimum continuous curtailment time constraint of the load that can be curtailed by the i-th node at the t-th moment, the second formula is the maximum continuous curtailment time constraint of the load that can be curtailed by the i-th node at the t-th moment, and the third formula represents the curtailment number constraint of the load that can be curtailed by the i-th node.

[0105] S2. Based on the model of the flexibility resources, construct a distribution network distributed energy storage planning model under a deterministic environment, specifically including S21 - S23:

[0106] S21. Establish an objective function to minimize the planning cost, specifically as follows:

[0107]

[0108] In the formula, F represents the planning cost, F Grid represents the cost of trading with the power grid, F DES represents the planning cost of distributed energy storage, F Shift represents the compensation cost of shiftable load, F Tran represents the compensation cost of transferable load, F Cut represents the compensation cost of curtailable load, F P2P represents the transmission cost, c b (t) represents the time-of-use electricity price, P b (t) represents the power purchased by the user from the power grid at the t-th moment, c s represents the feed-in tariff, P s (t) represents the power sold by the user to the power grid at the t-th moment, c dec represents the demand tariff, P bmax represents the demand power, N represents the number of distributed energy storage, represents the power of distributed energy storage, Denotes the capacity investment and construction cost of distributed energy storage. Denotes the operation and maintenance cost of distributed energy storage. I represents the number of nodes, c Shift Denotes the compensation cost coefficient of shiftable load, c Tran Denotes the compensation cost coefficient of transferable load, c Cut Denotes the compensation cost coefficient of curtailable load, c P2P Denotes the transmission cost coefficient. Denotes the power exchanged between the i-th node and the j-th node at the t-th moment.

[0109] S22. Establish the demand power constraint, the power flow constraint, and the power limit constraint for trading with the power grid, and obtain the constraint conditions according to the demand power constraint, the power flow constraint, the power limit constraint for trading with the power grid, and the model of the flexibility resources.

[0110] Among them, the demand power constraint is specifically:

[0111] Pb max = max[P b (t)];

[0112] The power flow constraint is specifically:

[0113]

[0114] In the formula, P i,Load (t) denotes the load power of the i-th node at the t-th moment. Denotes the charging power of the i-th distributed energy storage at the t-th moment. Denotes the photovoltaic power of the i-th node at the t-th moment. Denotes the discharging power of the i-th distributed energy storage at the t-th moment. Denotes the base load power of the i-th node at the t-th moment.

[0115] The power limit constraint for trading with the power grid is specifically:

[0116] 0 ≤ P b (t) ≤ Pb max ;

[0117] 0 ≤ P s (t) ≤ Ps max ;

[0118] In the formula, P smax Denotes the upper limit of the selling power.

[0119] The constraint conditions can be expressed as:

[0120]

[0121] S23. Generate a distribution network distributed energy storage planning model in a deterministic environment based on the objective function and the constraint conditions.

[0122] S3. Use the information gap decision theory to construct an upper-layer distribution network distributed energy storage planning model in an uncertain environment based on the flexibility resource model and the distribution network distributed energy storage planning model in a deterministic environment, and establish a lower-layer demand defense optimization model based on model predictive control, specifically including S31 - S33:

[0123] S31. Use the information gap decision theory to construct a distribution network distributed energy storage planning model in an initial uncertain environment, and use the flexibility resource model and the distribution network distributed energy storage planning model in a deterministic environment as the constraint conditions of the distribution network distributed energy storage planning model in the initial uncertain environment.

[0124] Among them, the distribution network distributed energy storage planning model in the initial uncertain environment is specifically:

[0125] max[σ1,..., σ i ;

[0126] maxF ≤ (1 + ω)F0;

[0127]

[0128] s.t. (1) - (7).

[0129] S32. Based on robust optimization, transform the distribution network distributed energy storage planning model in the initial uncertain environment to obtain an upper-layer distribution network distributed energy storage planning model in an uncertain environment.

[0130] As can be seen from the above formula, the distribution network distributed energy storage planning model in the initial uncertain environment is a two-layer multi-objective optimization model, which is difficult to solve directly. Therefore, first, use the idea of robust optimization to transform max[σ1,..., σ i into maxmin[σ1,..., σ i , and secondly, when the photovoltaic output is the smallest, the operating cost of the system is the highest. At this time, transform maxF ≤ (1 + ω)F0 into F ≤ (1 + ω)F0 to obtain the upper-layer distribution network distributed energy storage planning model in an uncertain environment, specifically:

[0131] maxmin[σ1,..., σ i ;

[0132] F ≤ (1 + ω)F0;

[0133]

[0134] In the formula, σ i represents the uncertainty radius of the i-th photovoltaic, F represents the optimal planning cost in a deterministic environment, that is, the planning cost in Equation (6), F0 represents the optimal planning cost in an uncertain environment, and ω represents the risk aversion factor. represents the power of the i-th photovoltaic. represents the expected power of the i-th photovoltaic.

[0135] The constraint conditions of the distribution network distributed energy storage planning model in the upper-layer uncertain environment are s.t. (1) to (7).

[0136] S33. Establish a demand defense optimization model based on model predictive control for the lower layer, specifically:

[0137]

[0138] In the formula, F Lower represents the demand defense optimization model based on model predictive control for the lower layer, H represents the prediction horizon of model predictive control, represents the charging power of the k-th distributed energy storage in the lower layer at the h-th moment, represents the charging power of the k-th distributed energy storage obtained from the upper-layer planning at the h-th moment, represents the discharging power of the k-th distributed energy storage in the lower layer at the h-th moment, represents the discharging power of the k-th distributed energy storage obtained from the upper-layer planning at the h-th moment, Ω represents the set of nodes, represents the power exchanged between the i-th node and the j-th node at the h-th moment in the lower layer, represents the power exchanged between the i-th node and the j-th node at the h-th moment obtained from the upper-layer planning, P b,low (h) represents the power purchased by users from the power grid at the h-th moment in the lower layer, P b (h) represents the power purchased by users from the power grid at the h-th moment obtained from the upper-layer planning, P s,low (h) represents the power sold by users to the power grid at the h-th moment, P s (h) represents the power sold by users to the power grid at the h-th moment obtained from the upper-layer planning.

[0139] S4. Iteratively optimize and solve the distribution network distributed energy storage planning model in the upper-layer uncertain environment and the demand defense optimization model based on model predictive control for the lower layer to obtain the optimal planning result of the distribution network.

[0140] Among them, the output of the distribution network distributed energy storage planning model under the upper-layer uncertainty environment is used as the input of the demand defense optimization model based on model predictive control in the lower layer, and the feedback result of the lower-layer model affects the decision-making of the upper-layer model. Through an iterative method, the upper and lower layer models cooperate with each other and are gradually optimized until the optimal solution is found. The upper-layer planning results include the power and capacity of distributed energy storage, the charging and discharging power, the flexible load results, the power purchased from and sold to the power grid, and the line exchange power. The lower-layer planning results include the real-time charging and discharging power of distributed energy storage, the real-time line exchange power, and the real-time power purchased from and sold to the power grid.

[0141] The above method of the present invention will be specifically described below through an example. Figure 3 It is a schematic diagram of the IEEE33 structure. The IEEE33 node example includes 6 distributed photovoltaics, and energy storage is configured at six locations of nodes 7, 9, 18, 20, 24, and 31, as Figure 3 shown. The time-of-use electricity price is specifically expressed as follows: the valley electricity price is 0.3139 CNY / kWh (00:00 - 08:00), the peak electricity price is 1.0697 CNY / kWh (08:00 - 12:00, 17:00 - 21:00), and the normal electricity price is 0.6418 CNY / kWh (12:00 - 17:00, 21:00 - 24:00). The demand electricity price is 38 CNY / kW per month, the feed-in tariff is 0.25 CNY / kWh, the optimization time T = 96, and Δt = 0.5. The construction cost of the power and capacity of distributed energy storage (DES) is 600 CNY / kW and 1000 CNY / kWh, the discount rate is 10%, and the service life is 10 years. The prediction horizon of model predictive control is 4, and the control horizon is 1. The load and distributed photovoltaic output curves are as Figure 4 shown.

[0142] To verify the impact of the combination of multiple types of flexibility resources on the distribution network planning, the following 4 scenarios (Cases) are set for comparative analysis:

[0143] Scenario 1: Consider shiftable loads, transferable loads, curtailable loads, and distributed energy storage participating in the distribution network planning process;

[0144] Scenario 2: Only consider shiftable loads, transferable loads, and curtailable loads participating in the distribution network planning process;

[0145] Scenario 3: Only consider distributed energy storage participating in the distribution network planning process;

[0146] Scenario 4: Do not consider flexible loads and distributed energy storage participating in the distribution network planning.

[0147] Table 1 Upper-layer model costs under four scenarios

[0148]

[0149] Table 2 Lower-layer model costs under four scenarios

[0150]

[0151] Table 3 Peak-valley differences of load under four scenarios

[0152] Case1 Case2 Case3 Case4 Peak - valley difference of load 5975.1707 5977.9942 7476 7476

[0153] First, to compare and analyze the economic benefits under 4 scenarios, v = 0.05 and a lower-layer PV error of 10% are selected to solve the distribution network planning models for the upper and lower layers under 4 scenarios in Tables 1 and 2. The results are shown in Tables 1 and 2. In the upper-layer model, due to the peak-valley arbitrage of distributed energy storage and the role of flexible load curtailment in reducing peak load, the distribution network planning cost is reduced by 8.9381%, 4.1495%, and 13.8487% compared with Scenarios 2, 3, and 4. It is proved that flexible resources can improve the economy of the distribution network. At the same time, the combination of the two flexible resources can better reduce the cost of the distribution network. When operating in the lower layer, the cost of the lower-layer model is below the cost of the upper-layer model, ensuring the success of demand defense and reducing the cost of real-time operation. Compared with Scenarios 2, 3, and 4, the economy of Scenario 1 is improved by 4.3070%, 4.0791%, and 6.9591% respectively.

[0154] Secondly, to prove the superiority of flexible load and distributed energy storage in reducing the peak-valley difference of net load. Table 3 shows the specific values of the peak-valley difference of net load under four scenarios. The peak-valley differences of net load in Scenarios 1 and 2 are reduced by 1500.8293 kW and 1498.0058 kW compared with Scenarios 3 and 4. This is because Scenarios 1 and 2 use transferable and shiftable loads to transfer and shift the load at the peak to other time periods. At the same time, the curtailable load curtails the load at the peak to reduce the power traded with the grid and the demand power.

[0155] Finally, Figure 5 shows the graph of the power traded with the grid and the change of distributed energy storage power. From Figure 5(a) It can be seen that during the first 31 time instants, the time-of-use electricity price is at the valley price. To improve the economy of the distribution network, the DES purchases a large amount of electric energy from the grid for charging, and the SOE (remaining energy of the battery) continuously rises, preparing for the peak electricity price in the following period. From the 32nd to the 47th time instants and from the 68th to the 83rd time instants, the time-of-use electricity price is at the peak price. The DES conducts discharging activities to ensure the supply-demand balance of the distribution network and reduce the cost of the distribution network. At the same time, the flexible load plays a role, and the shiftable and translatable loads are shifted and translated to the valley electricity price to reduce costs. From the 48th to the 68th time instants and from the 84th to the 96th time instants, the time-of-use electricity price is at the normal price. The DES conducts charging activities to prepare for the next peak electricity price period and ensure the recycling of the DES. At the same time, from Figure 5 (b) It can be seen from this that the SOE at the lower layer tracks the planning results at the upper layer in real time to conduct demand defense.

[0156] To prove the superiority of the above method of the present invention in dealing with the uncertainty of photovoltaic power. Figure 6 Show the graphs of the uncertainty radius and cost variation under different risk aversion factors. From Figure 6 it can be seen that when the risk aversion factor increases, the planning cost and the uncertainty radius also increase accordingly. This is because the decision maker believes that the uncertainty of photovoltaic power will have a negative impact on the planning cost and is pessimistic about the uncertainty. In addition, the planning cost always does not exceed the expected cost of the decision maker. From the perspective of system operation, as the uncertainty radius increases, the power value of the photovoltaic power is less than the predicted value, and the amount of electricity purchased from the external grid needs to be increased, resulting in a further increase in the actual operation cost. The data change trend is consistent with the trend shown in the graph, further verifying the effectiveness of the robust IGDT model.

[0157] As shown in Table 4 and Table 5, Table 4 and Table 5 show the comparison results of the above method of the present invention with robust optimization and fuzzy optimization. It can be seen from Table 4 that the costs of the above method of the present invention under the same uncertainty radius are respectively reduced by 257.927 yuan, 255.1522 yuan, 266.5693 yuan, 269.8985 yuan, and 264.3043 yuan compared with the costs of robust optimization. The above method of the present invention takes into account both the robustness and the economy of the system while ensuring the robustness of the system. Table 5 shows the comparison results of the above method of the present invention with fuzzy optimization, and the economy of the above method of the present invention is compared by setting different confidence levels of fuzzy optimization. Compared with fuzzy optimization, the costs of the above method of the present invention are respectively reduced by 1149.118 yuan, 874.1289 yuan, 601.7932 yuan, 327.4863 yuan, and 52.6811 yuan.

[0158] Table 4 Comparison of the multi-objective IGDT of the present invention and the results of robust optimization

[0159] Risk aversion factor Multi - objective IGDT Uncertainty radius Robust optimization 0.01 57092.8735 0.0386 57350.8005 0.02 57658.1495 0.0773 57913.3017 0.03 58223.4255 0.117 58489.9948 0.04 58788.7015 0.1562 59058.6000 0.05 59353.9774 0.1945 59618.2817

[0160] Table 5 Comparison of Multi-objective IGDT and Fuzzy Optimization Results of the Present Invention

[0161] Risk aversion factor Multi - objective IGDT Confidence level Fuzzy optimization 0.01 57092.8735 0.75 58241.9915 0.02 57658.1495 0.8 58532.2784 0.03 58223.4255 0.85 58825.2187 0.04 58788.7015 0.9 59116.1878 0.05 59353.9774 0.95 59406.6585

[0162] Please refer to Figure 2 , the second embodiment of the present invention is as follows:

[0163] A distribution network planning system considering flexible resources, including a memory, a processor, and a computer program stored on the memory and executable on the processor. When the processor executes the computer program, each step in the distribution network planning method considering flexible resources in the first embodiment is implemented.

[0164] In summary, the present invention provides a distribution network planning method and system considering flexible resources, establishes a model of flexible resources, constructs a distribution network distributed energy storage planning model in a deterministic environment based on the model of flexible resources, uses the Information Gap Decision Theory (IGDT) to construct an upper-layer distribution network distributed energy storage planning model in an uncertain environment based on the model of flexible resources and the distribution network distributed energy storage planning model in a deterministic environment, and establishes a lower-layer demand defense optimization model based on model predictive control. Iteratively optimize and solve the upper-layer distribution network distributed energy storage planning model in an uncertain environment and the lower-layer demand defense optimization model based on model predictive control to obtain the optimal distribution network planning result. It considers flexible resources including distributed energy storage and flexible loads, can fully explore the characteristics of various flexible resources, increases the economy of the distribution network, and takes into account robustness. Moreover, the constructed upper-layer distribution network distributed energy storage planning model in an uncertain environment and the lower-layer demand defense optimization model based on model predictive control can track and feedback-correct the planning result in real time to ensure that demand defense can be achieved, reduce the increase in the operating cost of the distribution network during the real-time planning process, thereby improving the economy and robustness of the distribution network and realizing demand defense; in addition, due to errors in the predicted value of photovoltaic in real-time operation, it will cause deviations in the charge and discharge plans of distributed energy storage during the lower-layer planning process, resulting in the failure of demand defense and thus increasing the operating cost of the distribution network. Therefore, to reduce the increase in the operating cost of the distribution network during the real-time planning process, it is necessary to perform demand defense on the distribution network. By establishing a lower-layer demand defense optimization model based on model predictive control, model predictive control can utilize its characteristics of rolling optimization and feedback correction to track the upper-layer planning result in real time according to the actual output value of photovoltaic and correct the output plan of the current distributed energy storage, etc., reducing the probability of demand defense failure and thus effectively realizing demand defense.

[0165] The above are only embodiments of the present invention, and thus do not limit the patent scope of the present invention. Any equivalent transformation made by using the content of the specification and drawings of the present invention, or directly or indirectly applied in the relevant technical fields, shall be similarly included in the patent protection scope of the present invention.

Claims

1. A distribution network planning method considering flexible resources, characterized in that, Including the steps: Establish a model of flexibility resources, where the flexibility resources include distributed energy storage and flexible loads; Construct a distributed energy storage planning model for a distribution network under a deterministic environment based on the model of the flexibility resources; Use the information gap decision theory to construct an upper-layer distributed energy storage planning model for a distribution network under an uncertain environment based on the model of the flexibility resources and the distributed energy storage planning model for a distribution network under a deterministic environment, and establish a lower-layer demand defense optimization model based on model predictive control; Iteratively optimize and solve the upper-layer distributed energy storage planning model for a distribution network under an uncertain environment and the lower-layer demand defense optimization model based on model predictive control to obtain the optimal planning result of the distribution network.

2. The distribution network planning method considering flexible resources according to claim 1, characterized in that The model of the flexibility resources includes a distributed energy storage model, specifically: In the formula, represents the charging power of the k-th distributed energy storage at the t-th moment, represents the power of the k-th distributed energy storage, represents the discharging power of the k-th distributed energy storage at the t-th moment, represents the capacity of the k-th distributed energy storage, ε represents the capacity coefficient of the k-th distributed energy storage, SOE k (T) represents the energy state of the k-th distributed energy storage at the T-th moment, T represents the total optimization time, SOE k (0) represents the energy state of the k-th distributed energy storage at the 0-th moment, SOE k (t) represents the energy state of the k-th distributed energy storage at the t-th moment, SOE k (t - 1) represents the energy state of the k-th distributed energy storage at the (t - 1)-th moment, η c represents the charging efficiency of the distributed energy storage, Δt represents the optimized time interval, η d represents the discharging efficiency of the distributed energy storage.

3. A distribution network planning method considering flexible resources according to claim 2, characterized in that, The model of the flexibility resources further includes a shiftable load model, a transferable load model, and a curtailable load model; The shiftable load model is specifically: Where, P i,Shift (t) represents the load power of the i-th node before translation at the t-th moment, represents the load power of the i-th node after translation at the t-th moment, α i,shift (t) represents a 0-1 variable indicating whether the shiftable load of the i-th node is shifted at the t-th moment, represents the end time of the shiftable period, represents the start time of the shiftable period, represents the maximum shiftable time of the shiftable load of the i-th node, represents the total continuous operation time of the shiftable load of the i-th node; The transferable load model is specifically: Wherein, P i,Tran (t) represents the load power of the i-th node before transfer at the t-th moment, represents the load power of the i-th node after transfer at the t-th moment, and β i,Tran (t) represents a 0-1 variable indicating whether the transferable load of the i-th node is transferred at the t-th moment, represents the minimum power value of the transferred load of the i-th node, represents the maximum power value of the transferred load of the i-th node, and β i,Tran (τ) represents a 0-1 variable indicating whether the transferable load of the i-th node is transferred at the τ-th moment, represents the total continuous operation time of the transferable load of the i-th node; The curtailable load model is specifically: Where, P i,Cut (t) represents the load of the i-th node before curtailment at the t-th moment, θ i,Cut (t) represents the load curtailment coefficient of the i-th node at the t-th moment, γ i,Cut (t) represents a 0-1 variable indicating whether the load of the i-th node is curtailed at the t-th moment, represents the load of the i-th node after curtailment at the t-th moment, represents the minimum continuous curtailment time, γ i,Cut (t - 1) represents a 0-1 variable indicating whether the load of the i-th node is curtailed at the (t - 1)-th moment, represents the maximum continuous curtailment time, N i,max represents the maximum number of curtailments of the load that can be curtailed by the i-th node.

4. A distribution network planning method considering flexible resources according to claim 3, characterized in that The construction of the distributed energy storage planning model for a distribution network under a deterministic environment based on the model of the flexibility resources includes: Establish an objective function to minimize the planning cost; Establish demand power constraints, power flow constraints, and power limit constraints for trading with the power grid, and obtain constraint conditions according to the demand power constraints, the power flow constraints, the power limit constraints for trading with the power grid, and the model of the flexibility resources; Generate a distributed energy storage planning model for a distribution network under a deterministic environment based on the objective function and the constraint conditions.

5. A distribution network planning method considering flexible resources according to claim 4, characterized in that, The establishment of the objective function to minimize the planning cost is specifically: In the formula, F represents the planning cost, F Grid represents the cost of trading with the power grid, F DES represents the planning cost of distributed energy storage, F Shift represents the compensation cost of shiftable load, F Tran represents the compensation cost of transferable load, F Cut represents the compensation cost of curtailable load, F P2P represents the transmission cost, c b (t) represents the time-of-use electricity price, P b (t) represents the power purchased by the user from the power grid at the t-th moment, c s represents the feed-in tariff, P s (t) represents the power sold by the user to the power grid at the t-th moment, c dec represents the demand tariff, P bmax represents the demand power, N represents the number of distributed energy storage units, represents the power of the distributed energy storage, represents the capacity investment and construction cost of the distributed energy storage, represents the operation and maintenance cost of the distributed energy storage, I represents the number of nodes, c Shift represents the compensation cost coefficient of shiftable load, c Tran represents the compensation cost coefficient of transferable load, c Cut represents the compensation cost coefficient of curtailable load, c P2P represents the transmission cost coefficient, represents the power exchanged between the i-th node and the j-th node at the t-th moment.

6. The distribution network planning method considering flexibility resources according to claim 5, characterized in that, The demand power constraint is specifically: P bmax = max[P b (t)]; The power flow constraint is specifically: Where, P i,Load (t) represents the load power of the i-th node at the t-th moment, represents the charging power of the i-th distributed energy storage at the t-th moment, represents the photovoltaic power of the i-th node at the t-th moment, represents the discharging power of the i-th distributed energy storage at the t-th moment, represents the base load power of the i-th node at the t-th moment; The power limit constraint for trading with the power grid is specifically: 0 ≤ P b (t) ≤ P bmax ; 0 ≤ P s (t) ≤ P Smax ; where P bmax represents the upper limit of the purchase power, and P smax represents the upper limit of the selling power.

7. A distribution network planning method considering flexible resources according to claim 1, characterized in that The construction of the upper-layer distributed energy storage planning model for a distribution network under an uncertain environment using the information gap decision theory based on the model of the flexibility resources and the distributed energy storage planning model for a distribution network under a deterministic environment includes: Use the information gap decision theory to construct a distributed energy storage planning model for a distribution network under an initial uncertain environment, and use the model of the flexibility resources and the distributed energy storage planning model for a distribution network under a deterministic environment as the constraint conditions of the distributed energy storage planning model for a distribution network under the initial uncertain environment; Based on robust optimization, transform the distributed energy storage planning model for a distribution network under the initial uncertain environment to obtain an upper-layer distributed energy storage planning model for a distribution network under an uncertain environment.

8. A distribution network planning method considering flexible resources according to claim 7, characterized in that, The upper-layer distributed energy storage planning model for a distribution network under an uncertain environment is specifically: max min[σ1,...,σ i ; F ≤ (1 + ω)F0; where, σ i represents the uncertainty radius of the i-th photovoltaic, F represents the optimal planning cost under a deterministic environment, F0 represents the optimal planning cost under an uncertain environment, ω represents the risk aversion factor, represents the power of the i-th photovoltaic, represents the expected power of the i-th photovoltaic.

9. A distribution network planning method considering flexible resources according to claim 5, characterized in that The establishment of the lower-layer demand defense optimization model based on model predictive control is specifically: where F Lower represents the demand defense optimization model of model predictive control at the lower layer, H represents the prediction time domain of model predictive control, represents the charging power of the k-th distributed energy storage at the h-th moment at the lower layer, represents the charging power of the k-th distributed energy storage at the h-th moment obtained by the upper layer planning, represents the discharging power of the k-th distributed energy storage at the h-th moment at the lower layer, represents the discharging power of the k-th distributed energy storage at the h-th moment obtained by the upper layer planning, Ω represents the set of nodes, represents the power exchanged between the i-th node and the j-th node at the h-th moment at the lower layer, represents the power exchanged between the i-th node and the j-th node at the h-th moment obtained by the upper layer planning, P b,low (h) represents the power purchased by users from the power grid at the h-th moment at the lower layer, P b (h) represents the power purchased by users from the power grid at the h-th moment obtained by the upper layer planning, P s,low (h) represents the power sold by users to the power grid at the h-th moment, P s (h) represents the power sold by users to the power grid at the h-th moment obtained by the upper layer planning.

10. A distribution network planning system considering flexible resources, comprising a memory, a processor, and a computer program stored on the memory and executable on the processor, characterized in that, When the processor executes the computer program, it realizes each step in a distribution network planning method considering flexibility resources according to any one of claims 1 to 9.

Images