User side load collaborative optimization method based on demand response

Abstract

The invention discloses a user side load collaborative optimization method based on demand response, and provides a load polymer alliance and power distribution network collaborative method (c) based on a non-cooperative game by constructing a large monomer user side load model (a) and a small monomer user side load polymer model (b) based on the non-cooperative game. And (d), a user side load robust optimization scheduling strategy is established, and efficient cooperative scheduling of large-scale user side loads in an uncertain environment is realized. According to the invention, regulation and control models of the user side load and the user side load aggregate are respectively constructed, an interaction method considering the difference of the user side load is established, and a load aggregator alliance and power distribution network cooperation method based on a non-cooperative game is provided. On the basis, a user side load robust optimization scheduling strategy is established, efficient collaborative scheduling of large-scale user side loads in an uncertain environment is achieved, and the safety, economical efficiency and voltage compliance of operation of a power distribution network are improved.

Description

Technical Field

[0001] This invention relates to the field of user-side load collaborative scheduling, and particularly to a user-side load collaborative optimization method based on demand response. This technology establishes refined models of large-scale individual user-side loads and small-scale individual user-side load aggregations, develops an interactive method considering the differences in user-side loads, and constructs a load aggregator alliance and distribution network collaborative framework based on non-cooperative game theory. Furthermore, it addresses load uncertainty through robust optimization and control strategies, achieving efficient collaborative scheduling of large-scale user-side loads and safe and economical operation of the distribution network. Background Technology

[0002] With the deepening of the construction of new power systems, the penetration rate of distributed generation and user-side loads in distribution networks has significantly increased. The volatility of their output and the randomness of their loads pose severe challenges to the safe and stable operation of distribution networks. Against this backdrop, demand-side response, as an important means of regulating supply and demand balance and improving system flexibility, is developing from traditional unidirectional load control to a direction of multi-entity collaborative interaction and refined regulation.

[0003] As a crucial link connecting the main grid and users, the distribution network's operational status is directly affected by the aggregation and dispatching effectiveness of demand-side resources. However, current distribution networks often encounter problems such as voltage exceeding limits, line overload, and increased network losses when facing a high proportion of renewable energy and small individual user-side charging loads. The traditional "source follows load" dispatching model is no longer suitable for the development requirements of the new power system. Therefore, it is urgent to effectively aggregate massive, dispersed user-side loads and other demand-side resources through technical means and incorporate them into the provincial and regional dispatching system to achieve coordinated optimization of "source, grid, load, and storage."

[0004] Therefore, this study focuses on user-side load coordination optimization in demand response. It aims to improve the adaptability of the distribution network to uncertainties, ensure the safe, economical and high-quality operation of the system, and promote the transformation of the distribution network from passive absorption to active control by establishing differentiated modeling methods, designing load aggregator interaction methods based on non-cooperative game theory, and constructing a "robust optimization" coordination framework. Summary of the Invention

[0005] The technical problem this invention aims to solve is to provide a demand-response-based user-side load collaborative optimization method to address the operational uncertainties, voltage exceedances, and line overloads brought about by the large-scale, multi-type user-side load access to the distribution network. First, differentiated behavioral models and collaborative mechanisms are established for different types of loads, such as large individual user-side loads and small individual user-side load aggregates. Based on this, a load aggregator alliance model based on non-cooperative game theory is constructed, and its distributed collaborative scheduling framework with the distribution network is designed. Further, a user-side load collaborative optimization strategy is proposed: a robust optimization method is used to construct a scheduling model considering the uncertainties of load and charging demand, improving the feasibility and robustness of the plan; and an active-reactive power collaborative control mechanism is proposed to achieve rapid response and accurate elimination of voltage exceedances and line overloads. Through duality theory, the model is transformed and linearized, converting it into an LP problem, which is then solved. Simulation results show that the proposed method can effectively improve the collaborative management and control capability of the distribution network for multi-type user-side loads, significantly improve voltage quality, reduce network losses, and enhance the system's operational resilience in the face of uncertainties.

[0006] To solve the above-mentioned technical problems, the technical solution of the present invention is as follows:

[0007] A demand-response-based user-side load collaborative optimization method is proposed. A large-unit user-side load model (a) and a small-unit user-side load aggregation model based on non-cooperative game theory (b) are constructed. A collaborative method between the load aggregation alliance and the distribution network based on non-cooperative game theory is proposed (c). A robust optimization scheduling strategy for user-side load is established (d), realizing efficient collaborative scheduling of large-scale user-side loads under uncertain environments.

[0008] The above technical solution has the following beneficial effects:

[0009] 1. This invention addresses the differences in user-side load by innovatively implementing a method of reducing load on large units and aggregating small units first, followed by indirect control through adjusting tie-line power. It proposes interactive mechanisms and scheduling strategies that consider differences in information exchange and control authority, solving the problem of poor control effect caused by treating user-side load as a homogeneous object in traditional methods. This improves the load aggregator's ability to finely coordinate and schedule multiple types of user-side load.

[0010] 2. This invention proposes a robust optimization-based large single-unit user-side load control model and an efficient solution algorithm. By constructing a set that includes load fluctuations and user behavior uncertainties, the system makes decisions with the goal of optimizing system performance in the worst case. The model is transformed into a single-layer mixed integer second-order cone programming problem that can be solved efficiently by using duality theory and the second-order cone relaxation method. This significantly enhances the feasibility and robustness of the scheduling scheme in the face of multiple uncertainties. Attached Figure Description

[0011] To more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the accompanying drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are merely exemplary, and those skilled in the art can derive other embodiments based on the provided drawings without creative effort.

[0012] The structures, proportions, sizes, etc. illustrated in this specification are only for the purpose of assisting those skilled in the art in understanding and reading the content disclosed herein, and are not intended to limit the conditions under which the present invention can be implemented. Therefore, they have no substantial technical significance. Any modifications to the structure, changes in the proportions, or adjustments to the size, without affecting the effects and objectives that the present invention can produce, should still fall within the scope of the technical content disclosed in the present invention.

[0013] Figure 1 This is a diagram of the improved IEEE-33 node distribution network system according to Embodiment 1 of the present invention;

[0014] Figure 2 This is the per-unit load curve of Embodiment 1 of the present invention;

[0015] Figure 3 shows the LAs scheduling results of Embodiment 1 of the present invention;

[0016] Figure 4 This is a comparison diagram of node voltages at 19 o'clock in various embodiments of the present invention;

[0017] Figure 5 The above are 18-day voltage fluctuation diagrams for nodes in various embodiments of the present invention.

[0018] Figure 6 This is a comparison chart of network loss at different time periods in various embodiments of the present invention;

[0019] Figure 7 This is a diagram showing the maximum over-limit node voltage before and after the over-limit adjustment of this invention;

[0020] Figure 8 This is a voltage diagram of each node at 19:00 and 21:00 according to the present invention;

[0021] Figure 9This invention provides a curve showing the feeder load rate before and after adjustment.

[0022] Figure 10 This invention relates to the reactive power regulation of large-unit user-side loads.

[0023] Figure 11 This refers to the active power regulation of the large-unit user-side load in this invention. Detailed Implementation

[0024] The specific embodiments of the present invention will be further described below with reference to the accompanying drawings. It should be noted that these descriptions are for the purpose of aiding understanding the present invention, but do not constitute a limitation thereof. Furthermore, the technical features involved in the various embodiments of the present invention described below can be combined with each other as long as they do not conflict with each other.

[0025] See Figures 1-11 As shown, this invention discloses a user-side load collaborative optimization method based on demand response, comprising the following steps:

[0026] S1, Demand-side load modeling

[0027] S2. Distribution network optimization scheduling considering source-load uncertainty

[0028] S3. Model Transformation and Solution

[0029] Furthermore, step S1 specifically includes:

[0030] S1.1 Large-scale user-side load modeling

[0031] Based on load forecast results Develop a load reduction plan, as shown in equation (5.3-12).

[0032] (1)

[0033] In the formula: Let be the total charging power at time t for the large single-unit load at node n. For a large single-unit load, the set of nodes. This is the set of scheduling time periods.

[0034] S1.2 Modeling of small individual user-side load aggregates based on non-cooperative game theory.

[0035] S1.2.1LA Alliance Modeling

[0036] The distribution network also contains some small individual user-side loads, which need to be aggregated for easier control. Therefore, a model aggregating loads using a load aggregator (LA) is proposed, and the model is established as follows:

[0037] (2)

[0038] In the above formula, Let n be the objective function of the nth LA; and These are the on-grid electricity price and the time-of-use electricity price at time t, respectively. and These represent the electricity fed into the grid and the electricity purchased from the grid for the nth LA, respectively, where m is the individual load unit; It is the set of small individuals under the nth node; For the LA set within the distribution network, Let m be the charging and discharging power of the m-th load cell at the n-th node at time t. The number of load cells aggregated by each load aggregator (LA) is 90, and the maximum charging and discharging power of each load cell is 7kW.

[0039] The interaction between the LA and the distribution network needs to meet the following conditions:

[0040] (3)

[0041] (4)

[0042] (5)

[0043] (6)

[0044] (7)

[0045] (8)

[0046] In the above formula, The charging and discharging power of the load cell m. , Let T be the column vector of photovoltaic cells within a small individual user-side load aggregate, and T be the scheduling cycle length. , , For a set of nodes with small individual loads, The minimum power required for charging and discharging a single load cell. The maximum charging and discharging power of a single load cell at time t. Let m be the on-grid electricity consumption of load cell m under the nth node at time t. The amount of electricity purchased by the power grid for load cell m under the nth node at time t.

[0047] Based on this, a profit maximization model for the LA alliance is constructed:

[0048] (9)

[0049] Constraints:

[0050] Equations (3)-(8)(10)

[0051] In the above formula, The aggregation result of the power aggregation quotient of the tie line at time t for the nth node. The decision result of the tie line at time t of the nth node of the distribution network. For coefficients, decision variables , , and This determines the magnitude of the coupling power between the objective function of the nth LA and the distribution network tie line, which will be needed in subsequent interactions with the distribution network.

[0052] S1.2.2LA Alliance and Distribution Network Coupling Analysis

[0053] The power coupling between the distribution network and the LA consortium regarding tie-line transmission power is described by the following constraints:

[0054] (11)

[0055] In the above formula, The number of nodes in the distribution network is obtained by the LA consortium and satisfies the following constraints:

[0056] (12)

[0057] In the above formula, Represents time t Net purchase of electricity from the grid.

[0058] Furthermore, step S2 specifically includes:

[0059] S2.1 Multi-objective function modeling

[0060] Ensuring power quality by minimizing voltage offset can be mathematically represented as:

[0061] (13)

[0062] (14)

[0063] In the above formula, For nodes Voltage amplitude at time t For nodes Reference voltage value, The objective function is the voltage offset. The parameter value is set.

[0064] By minimizing the total losses in the distribution network, it can be mathematically represented as:

[0065] (15)

[0066] In the above formula, branch road The resistance value, For the side road The current value at time t. This is the sum of losses in the distribution network.

[0067] Considering the minimization of user-side load charging costs under the time-of-use pricing mechanism, as shown in the following formula.

[0068] (16)

[0069] In the formula, Let be the total charging power of the large single-unit load at node i at time t. The reference value for the overall charging power of the large single-unit load at node i at time t. This is the charging cost function for user-side loads under the time-of-use pricing mechanism.

[0070] The objective function described above can be expressed as:

[0071] (17)

[0072] In the formula: These are the weighting coefficients; Let be the LA load at node i that satisfies the distribution safety constraints at time t. Let be the tie-line power of node i at time t. is a coefficient.

[0073] S2.2 Constraint Modeling

[0074] Distribution network nodes must satisfy active / reactive power balance constraints (18)-(19):

[0075] (18)

[0076] (19)

[0077] In the above formula, , They are respectively The net active and reactive loads of the node at time t. , They are respectively The normal active and reactive loads of the node at time t. branch road The active power output of DG at time t, For nodes The reactive power compensation provided by CB at time t.

[0078] The Distflow must satisfy the constraints shown in formulas (20)-(23):

[0079] (20)

[0080] (twenty one)

[0081] (twenty two)

[0082] (twenty three)

[0083] In the above formula, , Branch roads The active and reactive power transmitted at time t, For nodes The voltage amplitude at time t, For nodes The voltage amplitude at time t, , Branch roads Resistance and reactance values , For time t The transmission power.

[0084] The voltage and current must satisfy the constraints shown in formulas (24)-(25):

[0085] (twenty four)

[0086] , (25)

[0087] In the above formula, and For nodes The minimum and maximum voltage constraints at the point are set to a value between 0.95 pu and 1.05 pu. branch road Maximum current constraint.

[0088] S2.3 Simplified Second-Order Cone Model

[0089] First, perform variable substitution on the squares of voltage and current:

[0090] (26)

[0091] (27)

[0092] Then formulas (20)-(25) can be reformulated as:

[0093] (28)

[0094] (29)

[0095] (30)

[0096] (31)

[0097] (32)

[0098] (33)

[0099] In the above formula, , .

[0100] Subsequently, by performing second-order cone relaxation on formula (31), we can obtain:

[0101] (34)

[0102] After variable substitution, the voltage deviation and network loss in the original objective function are corrected as follows:

[0103] (35)

[0104] (36)

[0105] After the above processing, the original nonlinear programming model is transformed into a mixed-integer second-order cone programming problem, which can be solved efficiently using commercial solvers such as Cplex.

[0106] S2.4 Robust Optimization Model

[0107] Considering the impact of various uncertainties on the system, a robust optimization model is constructed.

[0108] (37)

[0109] (38)

[0110] In the above formula, each parameter represents: ; ; and , and It is a constant matrix. i .

[0111] Constructing an uncertain set:

[0112] Considering the uncertainty of load forecasting results for large individual users, uncertainty set modeling is performed:

[0113] (39)

[0114] In the above formula, This is a predicted value; It is a positive deviation value; It is a negative deviation value.

[0115] For equation (39), normalized auxiliary variables are used. , and Further rewritten as:

[0116] (40)

[0117] In the above formula, This is the lower bound of the prediction interval; The difference between the upper and lower bounds; belong Continuous values.

[0118] Furthermore, step S3 specifically includes:

[0119] S3.1 Dual Process Modeling

[0120] (41)

[0121] (42)

[0122] Among them, let , , , Then we can obtain , .

[0123] Construct the Lagrange function as follows:

[0124] (43)

[0126] As shown above, it is easy to see You can get

[0127] (44)

[0128] (45)

[0129] (46)

[0130] Therefore, its dual problem can be expressed as:

[0131] (47)

[0132] (48)

[0133] S3.2 Final Robust Optimization Model

[0134] (49)

[0135]

[0136] I. On constructing the model and simulation system:

[0137] The proposed coordinated control model was verified through simulation using an improved IEEE 33-node distribution network system. The system's base capacity was set to 10 MVA, the base voltage to 10 kV, the initial voltage of the slack node to 1.0 pu, the safe operating range of the node voltage to 0.95 pu to 1.05 pu, and the transmission power limit of critical feeder 1 to 4.6 MW.

[0138] At nodes 6, 15, 19, and 25, configure public large-unit user-side loads (large-unit user-side load 1), dedicated large-unit user-side loads (large-unit user-side load 2), large-unit user-side load 3, and large-unit user-side load 4, respectively. Each load aggregator (LA) aggregates 90 load units, and each load unit has a maximum charge / discharge power of 7kW. The minimum and maximum state of charge (SOC) are set to 0 and 1, respectively.

[0139] For distributed power generation, wind turbines (WT) are connected at nodes 8 and 16, and photovoltaic (PV) power is connected at nodes 10, 17, 21, and 28. The installed capacity of both PV and WT is 1.0MW. Regarding reactive power and voltage control equipment, an on-load tap changer (OLTC) is configured at the first-end node 1, with 5 tap positions, an adjustment step size of 0.025, and an adjustment range of 0.95pu~1.05pu. Capacitor banks (CB) are connected at nodes 12 and 29, with each CB having 5 capacitor banks, each with a capacity of 60kvar. The maximum number of adjustments per day for both OLTC and CB is set to 5. The voltage-reactive power threshold for voltage over-limit regulation is 0.2, and the voltage-active power threshold is 0.35.

[0140] To verify the effectiveness of the model, a multi-scenario comparative experiment was designed. Three cases were set up:

[0141] Case 1 (Benchmark Case): Disorderly charging of user-side loads, used to quantitatively assess the impact of disorderly charging behavior on the distribution network.

[0142] Case 2: User-side load cluster collaborative participation in regulation, used to analyze the effect of single-dimensional regulation of user-side load.

[0143] Case 3: Coordinated regulation of voltage regulating equipment (OLTC, CB) with user-side load to verify the further improvement effect of coordinated regulation.

[0144] II. Analysis of Optimized Scheduling Results:

[0145] The optimization results of the power distribution network are analyzed. Figure 4 The node voltage distribution of three cases during peak load periods was compared. In baseline case 1, disordered charging by user-side loads caused the voltages at nodes 6-18 and 26-33 of the distribution network to generally fall below the safe threshold, with the maximum voltage exceedance reaching 0.89 pu. In case 2, through user-side load regulation, the number of nodes exceeding the voltage limit decreased, and the maximum voltage exceedance increased to 0.93 pu, indicating that user-side load regulation can effectively alleviate local voltage exceedances. In case 3, through coordinated regulation by voltage regulation equipment and user-side loads, the overall node voltage of the distribution network recovered to the safe operating range, successfully eliminating voltage exceedances.

[0146] Terminal node 18 was selected as the observation node, and its daily voltage fluctuation is as follows: Figure 5 As shown in Case 1 and Case 2, the lowest voltage values ​​of this node during peak load periods were 0.88 pu and 0.89 pu, respectively. However, in Case 3, through coordinated regulation, the voltage was stabilized within the range of [0.97 pu, 1.02 pu], with a significant reduction in fluctuation amplitude and effective guarantee of power supply quality.

[0147] Analysis of distribution network losses, such as Figure 6 As shown in Table 1, the total daily network loss in Case 1 is 2681.0 kW. Compared to Case 1, the total daily network loss in Case 3 is reduced by 940.4 kW, a reduction of 35%. The results show that the collaborative control model proposed in this invention can effectively reduce network loss and improve system economy.

[0148] Table 1. Comparison of Total Network Loss for Case 1-3

[0149] Case Case 1 Case 2 Case 3 Daily total network loss (kW) 2681.0 2536.7 1740.6

[0150] S3. Analysis of the effect of flexible regulation

[0151] The control effect was analyzed. When the system experienced voltage over-limit and line overload during the peak load period from 17:00 to 22:00, the proposed control method was triggered. Figure 7 The voltage change of the maximum voltage over-limit node (node ​​18) before and after regulation is shown. Through the reactive-active coordinated compensation strategy, its voltage amplitude was successfully corrected to the safe range, and the maximum deviation value was increased from 0.91pu to 0.96pu.

[0152] Figure 9 The load rate of feeder 1 before and after the control was compared. Before the control, the feeder was in a state of power flow overload from 19:00 to 22:00, with a peak load rate of 141.6%. After the control, the average load of the feeder during the peak period decreased from 5.64MW to 4.6MW (limit), effectively alleviating the heavy overload phenomenon of the line.

[0153] Figure 10 and Figure 11 The reactive and active power regulation of each individual user-side load was further quantified. During the regulation process, the reactive power compensation of the large individual user-side load (such as large individual user-side load 2) was prioritized to alleviate voltage overload. When alleviating feeder overload, the active power resources of multiple large individual user-side loads (such as large individual user-side loads 1, 2, and 4) were coordinated, proving the effectiveness and rationality of the proposed method in resource allocation.

[0154] III. Based on the comprehensive experimental results, we can conclude that:

[0155] (1) The user-side load collaborative optimization based on demand response proposed in this invention can effectively cope with the uncertainty brought about by user-side load and renewable energy access by formulating an economical and secure scheduling plan and performing rapid and accurate flexible regulation.

[0156] (2) The proposed method can significantly improve the power quality of the distribution network, stabilize the node voltage within a safe range, and effectively reduce the network loss of the system.

[0157] (3) The iterative interaction between the LA aggregation model based on non-cooperative game theory and the distribution network scheduling model, as well as the robust optimization planning and control, together constitute an efficient and robust user-side load collaborative control framework, ensuring the safe, stable and economical operation of the distribution network.

[0158] The embodiments of the present invention have been described in detail above with reference to the accompanying drawings, but the present invention is not limited to the described embodiments. For those skilled in the art, various changes, modifications, substitutions, and variations can be made to these embodiments without departing from the principles and spirit of the present invention, and these variations still fall within the protection scope of the present invention.

Claims

1. A demand response based user side load coordination optimization method, characterized in that: The large monomer user side load model (a) and the small monomer user side load aggregation model (b) based on non-cooperative game are constructed. The collaborative method (c) of load aggregation alliance based on non-cooperative game and distribution network is proposed. The user side load robust optimization scheduling strategy (d) is established. The efficient collaborative scheduling of large-scale user side load in uncertain environment is realized. 2.The demand response based user side load coordination optimization method according to claim 1, characterized in that: The macromonomer user side load model is: (1) wherein: is the total charging power at the moment t for the large monolithic load n, is the set of nodes for the large monolithic load, is the set of dispatch time periods.

3. The user side load collaborative optimization method based on demand response according to claim 2, characterized in that In the present application: The small monomer user side load aggregation alliance (LA) model based on non-cooperative game is （2）。 4.The demand response based user side load coordination optimization method of claim 3, wherein: The interaction between LA and distribution network needs to meet the following conditions: (3) (4) (5) (6) (7) (8) The LA alliance revenue maximization model is constructed: （9） The constraint conditions of LA alliance revenue maximization model are formula (3)-(8) (10).

5. The demand response based user side load coordination optimization method according to claim 4, characterized in that: The constraint conditions of LA alliance and distribution network about transmission power coupling of tie line are: （11） In the above formula, is the number of distribution network nodes, which is obtained by solving the LA alliance, and satisfies the following constraints: (12) In the above formula, represents the time t Net electricity is purchased from the grid. 6.The demand response based user side load coordination optimization method of claim 5, wherein: The multi-objective function modeling is voltage offset target function: (13) （14） A power distribution network network loss summation function: (15) The user side load charging cost function under time-of-use pricing mechanism is: （16） The above-mentioned objectives are integrated to form the function expression: （17）。 7. The demand response based user side load coordination optimization method of claim 6, wherein: The constraint condition modeling needs to meet the following balance constraint conditions of distribution network node: (18) (19) Recursive derivation is carried out by using Distflow power flow model, which needs to meet the following constraint conditions: (20) (21) (22) (23) The current and voltage need to meet the following constraint conditions: (24) , (25)。 8.The demand response based user side load coordination optimization method according to claim 6 or 7, characterized in that: After the second-order cone model simplification and relaxation of the target function and constraint conditions: The voltage deviation and network loss correction in the original target function are: (35) (36) The original nonlinear programming model is converted into a mixed integer second-order cone programming problem, which can be quickly solved by a commercial solver to obtain the minimum network loss and other parameters. 9.The demand response based user side load coordination optimization method of claim 8, wherein: The robust optimization model is: （37） （38） The uncertainty set of large monomer user side load prediction results is modeled: (40) The dual processing model is: (47) (48) The final robust optimization model is: (49) (50)。

Images