Voltage quality control method, device and system for distributed photovoltaic large-scale access of power distribution network and medium

Abstract

The invention provides a voltage quality control method, device and system for distributed photovoltaic large-scale access of a power distribution network and a medium. The method comprises the following steps: firstly, analyzing the problems of voltage out-of-limit and network loss increase caused by photovoltaic access; secondly, defining an electrical distance between nodes based on reactive voltage sensitivity, and realizing dynamic partitioning of the power distribution network by adopting a clustering algorithm; secondly, constructing a double-layer optimization model which takes the minimum subarea network loss and the optimal voltage quality as an upper layer target and takes the minimum total operation cost as a lower layer target; and finally, cooperatively solving the double-layer model through an improved iterative algorithm, and generating a comprehensive control strategy of reactive power regulation, energy storage charging and discharging, SVG switching and flexible load response of the parallel photovoltaic inverter. According to the invention, through an optimization mechanism of partition autonomy and double-layer cooperation, rapid, accurate and economic control of the voltage of the power distribution network under high-proportion photovoltaic access is realized, and the safe and stable operation capability of the power grid and the new energy consumption level are effectively improved.

Description

Technical Field

[0001] This invention relates to the field of power system technology, specifically to a voltage quality control method, device, system, and medium for large-scale distributed photovoltaic (PV) grid integration in power distribution networks. Background Technology

[0002] With the rapid development of the new energy industry, distributed photovoltaic (PV) power has been widely used due to its clean and renewable advantages, and large-scale integration into the distribution network has become a development trend. However, distributed PV power output is characterized by intermittency, fluctuation, and randomness. Large-scale integration will change the power flow distribution of the distribution network, leading to a series of voltage quality problems such as voltage exceeding limits, voltage fluctuations and flicker, and three-phase imbalance, which seriously affect the safe and stable operation of the distribution network and the power quality for users.

[0003] Existing voltage quality control methods for distribution networks mainly include traditional reactive power compensation device regulation and reactive power regulation by distributed generation sources themselves. However, traditional reactive power compensation devices are mostly fixed-capacity compensation or semi-controllable compensation, with slow response speeds, making it difficult to adapt to the rapid power fluctuations brought about by the large-scale integration of distributed photovoltaic (PV) power. Reactive power regulation relying solely on distributed PV inverters is limited by their rated capacity, resulting in limited regulation capabilities. During peak PV output or off-peak load periods, they cannot effectively suppress voltage rises, leading to voltage quality issues. Therefore, there is an urgent need for a voltage quality control method that can achieve hierarchical regulation and coordinated optimization to improve the distribution network's capacity to accommodate large-scale distributed PV integration and ensure the safe and stable operation of the distribution network. Summary of the Invention

[0004] The purpose of this invention is to provide a voltage quality control method, device, system, and medium for large-scale distributed photovoltaic (PV) grid integration in power distribution networks. Through hierarchical and zoned collaborative control, it achieves millisecond-level dynamic response and optimized resource allocation, thereby solving the problems mentioned in the background art.

[0005] To achieve the above objectives, the present invention provides the following technical solution:

[0006] A voltage quality control method for large-scale distributed photovoltaic (PV) grid integration in a distribution network includes the following steps:

[0007] Step 1: Analyze the impact of distributed photovoltaic (PV) grid connection on distribution network node voltage and network losses, and determine the voltage exceeding limits and increased network losses caused by the connection.

[0008] Step 2: Based on the impact analysis results determined in Step 1, calculate the reactive voltage sensitivity between each node of the distribution network, calculate the electrical distance between any two nodes in the distribution network based on the reactive voltage sensitivity, and partition the distribution network using a clustering method based on the electrical distance to obtain the optimal partitioning scheme;

[0009] Step 3: Based on the aforementioned partitioning scheme, establish an upper-level scheduling model with the optimization objectives of minimizing distribution network losses and optimizing voltage quality, and a lower-level scheduling model with the optimization objective of optimizing distribution network operation economy, thus constructing a two-layer optimization model;

[0010] Step 4: Iteratively solve the two-layer optimization model. Through alternating solution and information exchange between the upper-layer scheduling model and the lower-layer scheduling model, obtain a comprehensive scheduling strategy that ensures the voltage quality of the distribution network meets the requirements and the operating economy is optimal.

[0011] Step 5: Decompose the comprehensive scheduling strategy into specific control commands, and send them to the corresponding photovoltaic inverters, energy storage systems, static var generators and load management terminals for execution through the communication module.

[0012] Furthermore, step two specifically includes:

[0013] 2.1 Establish the reactive voltage sensitivity matrix among distribution network nodes;

[0014] 2.2 Based on the aforementioned reactive voltage sensitivity matrix, define and calculate the electrical distance between any two nodes;

[0015] 2.3 Based on the electrical distance, a clustering algorithm is used to divide the distribution network into multiple regions. The clustering algorithm includes: determining the initial partition center, assigning nodes to the corresponding partitions according to the principle of minimum electrical distance, and iteratively updating the partition center until convergence.

[0016] 2.4 Based on the evaluation indicators of regional node coupling degree and reactive power matching degree within the region, determine the final number of partitions and partitioning results.

[0017] Furthermore, the clustering algorithm described in step 2.3 specifically includes:

[0018] Randomly select K nodes as the initial partition centers;

[0019] Calculate the electrical distance from the remaining nodes to the center of each partition, and assign each node to the partition with the smallest electrical distance;

[0020] Recalculate the sum of electrical distances from all nodes in each partition to other nodes in the same partition, and select the node with the smallest sum of distances as the new center of the partition;

[0021] Determine if the partition center has changed. If it has changed, repeat the above steps of node partitioning and center update until the partition center no longer changes, and the clustering is complete.

[0022] Furthermore, in step three, the objective function of the upper-level scheduling model includes minimizing the partitioned network loss and optimizing the voltage quality index, wherein the voltage quality index includes at least the voltage offset level and the static voltage stability; the constraints of the upper-level scheduling model include power flow constraints and distribution network security constraints.

[0023] Furthermore, in step three, the objective function of the lower-level scheduling model is to minimize the total operating cost of the distribution network. The total cost includes: the cost of the distribution network purchasing electricity from the main grid, the cost of the distribution network selling electricity to the main grid, the operating cost of the distribution network, the demand response cost of the flexible load in the distribution network, and the cost of curtailment of solar power in the distribution network. The constraints of the lower-level scheduling model include the uncertainty constraint of new energy sources and the power interaction constraint between the main grid and the distribution network.

[0024] Furthermore, in step four, the iterative solution process for the two-layer optimization model specifically includes:

[0025] 4.1 Define the initial solutions and iteration parameters for the upper-level scheduling model and the lower-level scheduling model;

[0026] 4.2 Solve the current upper-level scheduling model to obtain the current optimal solution, and pass the corresponding upper-level optimization variables to the lower-level scheduling model as part of its constraints;

[0027] 4.3 Solve the current lower-level scheduling model to obtain the current optimal solution;

[0028] 4.4 Determine whether the difference between the optimal solutions of the upper and lower layer models is less than the preset convergence precision. If yes, output the final scheduling scheme; otherwise, feed back the variables corresponding to the optimal solution of the lower layer model to the upper layer scheduling model to update its constraints, and return to step 4.2 for the next iteration.

[0029] Furthermore, in step 4.2, a multi-objective optimization algorithm based on adaptive ε-constraints is used to solve the upper-level scheduling model in order to obtain the Pareto front solution set that is optimal in terms of both network loss and voltage quality.

[0030] A voltage quality control device for large-scale distributed photovoltaic (PV) grid integration in a distribution network includes:

[0031] The impact analysis module is used to analyze the impact of distributed photovoltaic access on the voltage of distribution network nodes and network losses, and to determine the voltage exceeding the limit and network loss increase caused by the access.

[0032] The partitioning module is used to calculate the reactive voltage sensitivity between nodes of the distribution network based on the impact analysis results determined by the impact analysis module, calculate the electrical distance between any two nodes in the distribution network based on the reactive voltage sensitivity, and partition the distribution network based on the electrical distance using a clustering method to obtain the optimal partitioning scheme.

[0033] The two-layer optimization model construction module is used to establish an upper-layer scheduling model with the optimization objectives of minimizing distribution network losses and optimizing voltage quality, and a lower-layer scheduling model with the optimization objective of optimizing distribution network operation economy, based on the partitioning scheme, thereby constructing a two-layer optimization model.

[0034] The model solving module is used to iteratively solve the two-layer optimization model. Through the alternating solving and information interaction between the upper-layer scheduling model and the lower-layer scheduling model, a comprehensive scheduling strategy that ensures the voltage quality of the distribution network is qualified and the operation is economically optimal is obtained.

[0035] The strategy execution module is used to decompose the comprehensive scheduling strategy into specific control commands and send them to the corresponding photovoltaic inverters, energy storage systems, static var generators and load management terminals for execution through the communication module.

[0036] A voltage quality control system for large-scale distributed photovoltaic (PV) grid integration in a distribution network includes: a computer-readable storage medium and a processor;

[0037] The computer-readable storage medium is used to store executable instructions;

[0038] The processor is used to read executable instructions stored in the computer-readable storage medium and execute the voltage quality control method for large-scale distributed photovoltaic grid integration.

[0039] A non-transitory computer-readable storage medium storing a computer program that, when executed by a processor, implements the voltage quality control method for large-scale distributed photovoltaic (PV) grid integration in a distribution network.

[0040] The present invention has the following beneficial effects:

[0041] 1) The scheduling scheme obtained by alternating iterative solution through hierarchical partitioning optimization model can minimize operating costs while ensuring optimal network loss and power quality, thereby ensuring the safe and economical operation of the distribution network;

[0042] 2) Through a hierarchical control architecture, full coverage regulation of voltage deviation from mild to severe is achieved, with voltage deviation controlled within ±2%, to meet the power supply needs of sensitive loads;

[0043] 3) Equipping distribution networks with energy storage devices can achieve multiple benefits, significantly improving network operation, reducing operating costs, minimizing energy loss, improving voltage quality, and enhancing the reliability and safety of network operation. In terms of voltage management, energy storage devices possess outstanding regulatory advantages. Through flexible charge and discharge control, they can quickly smooth voltage fluctuations, improve voltage deviations, stabilize node voltage levels, and effectively solve voltage quality problems caused by load fluctuations and distributed power source integration, providing crucial voltage support for the safe and stable operation of the distribution network. Regarding reducing network losses, energy storage devices can optimize power flow distribution, avoid unreasonable operating states such as reverse flow and overload, reduce energy loss during transmission, and improve the energy utilization efficiency of the distribution network. Attached Figure Description

[0044] Figure 1 This is a flowchart of the voltage quality control method for large-scale distributed photovoltaic (PV) grid integration in the distribution network according to the present invention. Detailed Implementation

[0045] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0046] Please see Figure 1 The first aspect of this invention provides a voltage quality control method for large-scale distributed photovoltaic (PV) grid integration in a distribution network, comprising the following steps:

[0047] Step 1: Analyze the impact of distributed photovoltaic (PV) grid connection on distribution network node voltage and network losses, and determine the voltage exceeding limits and increased network losses caused by the connection.

[0048] (1) Impact on node voltage

[0049] For a having For a distribution network with n nodes, the node voltage equations can be expressed as:

[0050]

[0051] In the formula: It is the nodal admittance matrix; It is the node voltage vector; It is the node injected current vector.

[0052] In the above formula, the admittance matrix and the node injection current have a decisive influence on the node voltage, while the node injection current... The magnitude and phase of the active and reactive power directly determine the amplitude and phase of the node voltage. When distributed photovoltaic (PV) systems are connected to the distribution network, their output active and reactive power will change the node injection current. If the distributed PV system outputs a large amount of active power, the injection current at the connection point will increase. If the reactive power is not properly regulated, the node voltage amplitude will exceed the allowable fluctuation range, causing the voltage to exceed the upper limit. Conversely, if the distributed PV system is limited by sunlight conditions and its output is insufficient, while the load demand remains unchanged or even continues to increase, the injection current amplitude at the connection node will decrease accordingly. The power drawn by the node from the grid will increase accordingly, and the voltage drop across the line impedance will increase, which may cause the node voltage amplitude to fall below the rated lower limit, resulting in a low voltage problem.

[0053] YbusV=Ii

[0054] (2) Impact on network losses

[0055] In traditional power distribution networks, power flows unidirectionally from the substation bus to the load end. Power loss on the lines is mainly due to the thermal effect of current on resistance, which can be expressed as:

[0056]

[0057] In the formula: This represents the total active power loss of the system; Line current; This represents the line resistance.

[0058] After large-scale distributed photovoltaic (PV) power is integrated into the distribution network, it changes the direction, magnitude, and distribution characteristics of power flow, breaking the inherent unidirectional power flow pattern of the traditional distribution network. Reverse power flow and power circulation phenomena occur frequently, and line current increases due to reverse power transmission and internal circulation. Line losses show a continuous upward trend with increasing penetration rate. When the penetration rate exceeds a critical value, line losses may exceed the level when no PV power is integrated, leading to economic deterioration.

[0059] Step 2: Based on the impact analysis results determined in Step 1, calculate the reactive voltage sensitivity between each node in the distribution network. Based on the reactive voltage sensitivity, calculate the electrical distance between any two nodes in the distribution network. Based on the electrical distance, partition the distribution network using a clustering method to obtain the optimal partitioning scheme.

[0060] Reactive power voltage sensitivity is a core indicator for quantifying the impact of reactive power changes on node voltage in a distribution network. Taking the PQ node, which accounts for the vast majority in medium and low voltage distribution networks, as an example, the process of solving the distribution network voltage sensitivity is as follows:

[0061]

[0062] In the formula: For any complex number; The series is the number of levels that can be expanded.

[0063] Therefore, the node and nodes reactive voltage sensitivity It can be derived from the following formula:

[0064]

[0065] In the formula: For nodes The amount of reactive power injected.

[0066] Based on reactive power voltage sensitivity, a clustering method is used to partition the distribution network. Node and The reactive voltage sensitivity may differ between elements, but clustering requires that the distance between two elements be the same. Therefore, based on reactive voltage sensitivity, nodes are defined. and Electrical distance:

[0067]

[0068] In the formula: For nodes The voltage; For nodes The voltage; For nodes The amount of reactive power injected; For nodes The amount of reactive power injected.

[0069] The steps for clustering and partitioning based on electrical distance are as follows:

[0070] 1) Random election Each node serves as the center of the partition cluster, denoted as region center 1, region center 2, ...;

[0071] 2) Calculate the distances from all nodes except the region center to the destination. The electrical distance between the centers of each zone is used as the standard to assign the remaining nodes to each zone until all nodes are assigned to a zone.

[0072] 3) Calculation The electrical distances from each node in a partition to other nodes in the partition are summed, and the node with the smallest sum of distances in the partition is selected as the new center of the partition.

[0073] 4) Compare whether the center of the region has changed. If it has changed, repeat step 2. If it has not changed, end the clustering and obtain the partitioning results.

[0074] The number of partitions is determined by the node coupling degree and reactive power matching degree within the region, as shown in the following formulas:

[0075]

[0076] In the formula: The degree of node coupling within the region; For partitioning The set of reactive power sources within; For partitioning Number of internal nodes; For reactive power sources within the zone Maximum reactive power capacity; For nodes within the partition The reactive load. The smaller this indicator, the better the reactive power matching within the zone.

[0077]

[0078] In the formula: The degree of reactive power matching within the region; This represents the maximum electrical distance between nodes in a distribution network. This indicator indicates the region. The ratio of the average electrical distance between internal nodes to the maximum electrical distance of the entire network indicates a stronger electrical coupling between nodes within the region. The reactive power matching degree and intra-regional node coupling degree are calculated sequentially for each number of zones, and the optimal ratio is selected as the final number of zones.

[0079] Step 3: Based on the aforementioned partitioning scheme, establish an upper-level scheduling model with the optimization objectives of minimizing distribution network losses and optimizing voltage quality, and a lower-level scheduling model with the optimization objective of optimizing distribution network operation economy, thus constructing a two-layer optimization model.

[0080] To meet the dual requirements of distribution network stability and economy, a dual optimization model is adopted, with objective functions designed for different optimization objectives. Combined with the actual operating characteristics of the distribution network, this approach prioritizes ensuring diversified power quality while minimizing operating costs, thus implementing a two-layer optimization scheduling strategy for a distribution network with a high proportion of distributed photovoltaic power. The upper-layer scheduling objective is to improve the power quality of the distribution network, while the lower-layer scheduling objective is to utilize various resources in coordination to minimize voltage deviation and network losses.

[0081] (1) Upper-level scheduling model

[0082] The uncertainty and volatility of renewable energy output can lead to irregular voltage variations at various nodes in the distribution network. Reducing the impact of high-proportion distributed photovoltaic (PV) grid integration on voltage level fluctuations and improving user satisfaction with voltage quality are important metrics for evaluating distribution network operational efficiency. Therefore, the upper-level scheduling model aims to minimize distribution network losses and optimize voltage quality at key nodes in each zone. Voltage quality evaluation indicators include node voltage offset levels and static voltage stability. The objective function of the upper-level scheduling model, on a zone-by-zone basis, is expressed as:

[0083]

[0084]

[0085] In the formula: This refers to the number of nodes in the distribution network. For the region Internal network loss function, kWh; For the region Internal voltage quality level; As a voltage offset level weight; For the region Internal voltage offset level, pu; For the region Internal static voltage stability; This represents the weighting for static voltage stability.

[0086]

[0087] In the formula: The scheduling cycle duration is in hours (h). for Area in the distribution network at all times Inland flow line The current value on it, kA; For the region Internal lines The resistance value.

[0088]

[0089] In the formula: For the distribution network area The number of nodes contained within; for Time of the first The per-unit value of the node voltage at each node, pu; Used as a reference value for determining node voltage offset.

[0090]

[0091]

[0092] In the formula: for Time zone Internal lines Static voltage stability; for Time zone Inner The net active power load of each node, in MW; for Time zone Inner Net reactive load value of each node, in MW; For the region Internal lines The reactance value; For the distribution network area Inside Time of the first Each node voltage value, pu; For the region The maximum value among all feeders, in kV.

[0093] Constraints:

[0094] Current constraints are:

[0095]

[0096] In the formula: , for time The active and reactive power flowing through the branch, in MW and Mvar; , for At every moment Injected active and reactive power, MW and Mvar.

[0097] The safety constraints of the power distribution network are:

[0098]

[0099] In the formula: This represents the upper limit of the square value of the line current. , These are the upper and lower limits of the squared value of the node voltage.

[0100] (2) Lower-level scheduling model

[0101] The lower-level scheduling model has a time granularity of 1 hour and a total scheduling duration of 24 hours. This is based on photovoltaic forecast data, load demand data, and electricity... Based on price parameters and other information, the actual output of high-proportion distributed photovoltaic power and various dispatch resources in the distribution network is determined to ensure the economic minimization of the distribution network operation. The objective function of the lower-level dispatch model is:

[0102]

[0103] In the formula: for The cost of electricity purchased from the main grid by the distribution network at any given time, in ten thousand yuan; for Cost of electricity sold from the distribution network to the main grid at any given time, in ten thousand yuan; for Real-time distribution network operating cost, in ten thousand yuan; The cost of flexible load demand response in the distribution network at all times is 10,000 yuan. for Cost of curtailment of solar power in the distribution network at any given time: 10,000 yuan.

[0104] The various cost functions are as follows:

[0105]

[0106] In the formula: — The unit electricity price of the distribution network purchasing electricity from the main grid at any given time, in yuan / kWh; — The power purchased by the distribution network from the main grid at any given time, in MW; — The unit electricity price at which the distribution network sells electricity to the main grid at any given time, in yuan / kWh; — Power sold from the distribution network to the main grid at any given time, in MW; — Distributed power source operating costs; — Operating costs of instantaneous energy storage devices; — Real-time SVG running cost; — The unit demand response cost coefficient for load at any given time; — Load response power value at any given time, in MW; — The cost coefficient of curtailment of solar power per unit of electricity in the power distribution network at any time; — Distributed photovoltaic curtailment power at any given time, in MW.

[0107] Constraints:

[0108] New energy uncertainty constraints:

[0109]

[0110]

[0111] In the formula: For the first Distributed photovoltaic in Active power injected at any given time, in MW; For the first Distributed photovoltaic in Predicted output deviation at any given moment; The confidence level for the probability that the constraint condition is true; For the first Distributed photovoltaic in The reactive power output at any given time, Mvar; For the first Distributed photovoltaic in The power factor angle at time t.

[0112] Power interaction constraints between main and distribution networks

[0113]

[0114]

[0115] In the formula: The maximum power purchased by the distribution network from the main grid per unit time, expressed in MW; This represents the maximum power output (MW) that the distribution network sells to the main grid per unit time.

[0116] Step 4: Solving the two-layer model

[0117] In this invention, the upper-level scheduling model contains two objective functions. To avoid the inability of the traditional weighted method to balance the importance of the two objective functions, the traditional particle swarm optimization algorithm is improved, and a multi-objective algorithm based on adaptive ε-constraints is constructed to solve the upper-level optimized power quality model. The original dual-objective problem of the upper-level scheduling model can be specifically expressed as:

[0118]

[0119] In the formula: , —Two objective functions in the upper-level model; —Reference variables for the upper-level scheduling model; , —Relevant constraints. Solving for both objectives separately determines the range of the Pareto front. Specifically, it is expressed as:

[0120]

[0121] From the above, we can see that the Pareto front boundaries are respectively ( )and( The corresponding constraint values ​​can be obtained through the two boundaries. for:

[0122]

[0123] In the formula: — Minimum value The value; — The minimum value during the solution process; — The number of segments in the range division can be adjusted to improve the accuracy of the Pareto front solution set.

[0124] Combining the above equation, the upper-level bi-objective problem can be transformed into:

[0125]

[0126] For the bi-level optimization model, a column-and-constraint algorithm is selected for iterative solution. The column-and-constraint algorithm can solve the bi-level optimization scheduling model quickly and effectively. In each iteration, the solution of the outer optimization problem is first updated by the cross-entropy method, and then the solution of the inner optimization problem is updated by the cross-gradient method until the convergence condition is met.

[0127] The compact form of the upper-level scheduling model is specifically represented as follows:

[0128]

[0129] Based on the objective function expression of the lower-level model, the compact form of the lower-level scheduling model can be obtained as follows:

[0130]

[0131] In the formula: As a reference variable for the lower-level scheduling model; , These are constraints.

[0132] The solution for the two-layer model is as follows:

[0133] 1) Set initial values ​​for the upper-level scheduling model Initial values ​​of the lower-level scheduling model Set the initial value of the number of iterations. Convergence accuracy Set it to 0.01.

[0134] 2) Solve the upper-level scheduling model to obtain the optimal solution for this iteration. The upper-level scheduling model reference variables The data is passed to the lower-level scheduling model, and the constraints of the lower-level model are updated.

[0135] 3) Solve the lower-level scheduling model and update the lower-level optimal solution. .

[0136] 4) Calculate the absolute value of the difference between the optimal solutions of the upper and lower layer scheduling models. If If the solution is found, output the result; otherwise, proceed to the next step.

[0137] 5) Variables The data is passed to the upper-level scheduling model to update the constraints and iteration count. Then proceed to step 2) for the next iteration until the convergence accuracy is met.

[0138] Calculation example

[0139] 1. Example parameters

[0140] Taking a 10kV distribution network as an example, the parameters are as follows:

[0141] Distribution network structure: radial feeders, total length 8km, conductor type JKLYJ-10-240, resistance 0.1Ω / km, reactance 0.3Ω / km, total distribution network load 3~4MW;

[0142] Distributed photovoltaic: 3 grid connection points, total installed capacity of 5MW, accounting for about 60% of the distribution transformer capacity, with output fluctuation range of 2~5MW;

[0143] Control equipment: Photovoltaic inverter reactive power regulation range ±0.4pu, energy storage system capacity 2MWh, power 1MW, SVG capacity 2Mvar.

[0144] 2. Test Scenario

[0145] Scenario 1: During the midday period from 12:00 to 13:00, the photovoltaic output increased from 3MW to 5MW, an increase of approximately 67%, while the load remained stable at 3.5MW;

[0146] Scenario 2: During the cloudy period from 14:00 to 15:00, the photovoltaic output fluctuates by ±30%, with a fluctuation frequency of 3 times / minute. The load increases from 3MW to 4MW, an increase of approximately 33%.

[0147] 3. Comparison of control effects, see Table 1:

[0148] Table 1

[0149]

[0150] A second aspect of the present invention provides a voltage quality control device for large-scale distributed photovoltaic (PV) grid integration in a distribution network, comprising:

[0151] The impact analysis module is used to analyze the impact of distributed photovoltaic access on the voltage of distribution network nodes and network losses, and to determine the voltage exceeding the limit and network loss increase caused by the access.

[0152] The partitioning module is used to calculate the reactive voltage sensitivity between nodes of the distribution network based on the impact analysis results determined by the impact analysis module, calculate the electrical distance between any two nodes in the distribution network based on the reactive voltage sensitivity, and partition the distribution network based on the electrical distance using a clustering method to obtain the optimal partitioning scheme.

[0153] The two-layer optimization model construction module is used to establish an upper-layer scheduling model with the optimization objectives of minimizing distribution network losses and optimizing voltage quality, and a lower-layer scheduling model with the optimization objective of optimizing distribution network operation economy, based on the partitioning scheme, thereby constructing a two-layer optimization model.

[0154] The model solving module is used to iteratively solve the two-layer optimization model. Through the alternating solving and information interaction between the upper-layer scheduling model and the lower-layer scheduling model, a comprehensive scheduling strategy that ensures the voltage quality of the distribution network is qualified and the operation is economically optimal is obtained.

[0155] The strategy execution module is used to decompose the comprehensive scheduling strategy into specific control commands and send them to the corresponding photovoltaic inverters, energy storage systems, static var generators and load management terminals for execution through the communication module.

[0156] Another aspect of the present invention provides a voltage quality control system for large-scale distributed photovoltaic (PV) grid integration in a distribution network, comprising: a computer-readable storage medium and a processor;

[0157] The computer-readable storage medium is used to store executable instructions;

[0158] The processor is used to read executable instructions stored in the computer-readable storage medium and execute the voltage quality control method for large-scale distributed photovoltaic access in the distribution network as described in the first aspect.

[0159] In another aspect, the present invention provides a non-transitory computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the voltage quality control method for large-scale distributed photovoltaic access in a distribution network as described in the first aspect.

[0160] Those skilled in the art will understand that embodiments of the present invention can be provided as methods, systems, or computer program products. Therefore, the present invention can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, the present invention can take the form of a computer program product embodied on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0161] This invention is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of the invention. It will be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, generate instructions for implementing the flowchart illustrations and / or block diagrams. Figure 1 One or more processes and / or boxes Figure 1 A device that provides the functions specified in one or more boxes.

[0162] These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing device to function in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means, which are implemented in a process Figure 1 One or more processes and / or boxes Figure 1 The function specified in one or more boxes.

[0163] These computer program instructions may also be loaded onto a computer or other programmable data processing equipment to cause a series of operational steps to be performed on the computer or other programmable equipment to produce a computer-implemented process, thereby providing instructions that execute on the computer or other programmable equipment for implementing the process. Figure 1 One or more processes and / or boxes Figure 1 The steps of the function specified in one or more boxes.

[0164] This invention has the following features and effects:

[0165] 1. A clustering and partitioning method based on reactive voltage sensitivity and electrical distance is proposed to achieve reasonable partitioning of the distribution network and enhance the reactive self-balancing capability and collaborative control efficiency within the region.

[0166] 2. Construct a two-layer optimization model with the upper layer focusing on voltage quality and network loss, and the lower layer focusing on economic operation. Through alternating iterative solutions, minimize operating costs while ensuring voltage compliance.

[0167] 3. An adaptive ε-constrained multi-objective optimization algorithm is used to process the upper-level model, effectively obtaining the Pareto front solution set and improving the coordination and adaptability of the scheduling scheme.

[0168] 4. Through zoned collaborative control and coordinated scheduling of multiple types of equipment (photovoltaic inverters, energy storage, SVG, loads), millisecond-level dynamic response is achieved, and voltage deviation can be controlled within ±2%, improving the adaptability and operational safety of the distribution network to high proportion of photovoltaic access.

[0169] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and not to limit it. Although the present invention has been described in detail with reference to the above embodiments, those skilled in the art should understand that modifications or equivalent substitutions can still be made to the specific implementation of the present invention. Any modifications or equivalent substitutions that do not depart from the spirit and scope of the present invention should be covered within the scope of protection of the claims of the present invention.

Claims

1. A voltage quality control method for large-scale distributed photovoltaic (PV) grid integration in a distribution network, characterized in that, Includes the following steps: Step 1: Analyze the impact of distributed photovoltaic (PV) grid connection on distribution network node voltage and network losses, and determine the voltage exceeding limits and increased network losses caused by the connection. Step 2: Based on the impact analysis results determined in Step 1, calculate the reactive voltage sensitivity between each node of the distribution network, calculate the electrical distance between any two nodes in the distribution network based on the reactive voltage sensitivity, and partition the distribution network using a clustering method based on the electrical distance to obtain the optimal partitioning scheme; Step 3: Based on the aforementioned partitioning scheme, establish an upper-level scheduling model with the optimization objectives of minimizing distribution network losses and optimizing voltage quality, and a lower-level scheduling model with the optimization objective of optimizing distribution network operation economy, thus constructing a two-layer optimization model; Step 4: Iteratively solve the two-layer optimization model. Through alternating solution and information exchange between the upper-layer scheduling model and the lower-layer scheduling model, obtain a comprehensive scheduling strategy that ensures the voltage quality of the distribution network meets the requirements and the operating economy is optimal. Step 5: Decompose the comprehensive scheduling strategy into specific control commands, and send them to the corresponding photovoltaic inverters, energy storage systems, static var generators and load management terminals for execution through the communication module.

2. The method according to claim 1, characterized in that, Step two specifically includes: 2.1 Establish the reactive voltage sensitivity matrix among distribution network nodes; 2.2 Based on the aforementioned reactive voltage sensitivity matrix, define and calculate the electrical distance between any two nodes; 2.3 Based on the electrical distance, a clustering algorithm is used to divide the distribution network into multiple regions. The clustering algorithm includes: determining the initial partition center, assigning nodes to the corresponding partitions according to the principle of minimum electrical distance, and iteratively updating the partition center until convergence. 2.4 Based on the evaluation indicators of regional node coupling degree and reactive power matching degree within the region, determine the final number of partitions and partitioning results.

3. The method according to claim 2, characterized in that, The clustering algorithm described in step 2.3 specifically includes: Randomly select K nodes as the initial partition centers; Calculate the electrical distance from the remaining nodes to the center of each partition, and assign each node to the partition with the smallest electrical distance; Recalculate the sum of electrical distances from all nodes in each partition to other nodes in the same partition, and select the node with the smallest sum of distances as the new center of the partition; Determine if the partition center has changed. If it has changed, repeat the above steps of node partitioning and center update until the partition center no longer changes, and the clustering is complete.

4. The method according to claim 1, characterized in that, In step three, the objective function of the upper-level scheduling model includes minimizing the partitioned network loss and optimizing the voltage quality index. The voltage quality index includes at least the voltage offset level and the static voltage stability. The constraints of the upper-level scheduling model include power flow constraints and distribution network security constraints.

5. The method according to claim 4, characterized in that, In step three, the objective function of the lower-level scheduling model is to minimize the total operating cost of the distribution network. The total cost includes: the cost of the distribution network purchasing electricity from the main grid, the cost of the distribution network selling electricity to the main grid, the operating cost of the distribution network, the demand response cost of the flexible load in the distribution network, and the cost of curtailment of solar power in the distribution network. The constraints of the lower-level scheduling model include the uncertainty constraint of new energy sources and the power interaction constraint between the main grid and the distribution network.

6. The method according to claim 5, characterized in that, Step four, the iterative solution process for the two-layer optimization model specifically includes: 4.1 Define the initial solutions and iteration parameters for the upper-level scheduling model and the lower-level scheduling model; 4.2 Solve the current upper-level scheduling model to obtain the current optimal solution, and pass the corresponding upper-level optimization variables to the lower-level scheduling model as part of its constraints; 4.3 Solve the current lower-level scheduling model to obtain the current optimal solution; 4.4 Determine whether the difference between the optimal solutions of the upper and lower layer models is less than the preset convergence precision. If yes, output the final scheduling scheme; otherwise, feed back the variables corresponding to the optimal solution of the lower layer model to the upper layer scheduling model to update its constraints, and return to step 4.2 for the next iteration.

7. The method according to claim 6, characterized in that, In step 4.2, a multi-objective optimization algorithm based on adaptive ε-constraints is used to solve the upper-level scheduling model in order to obtain the Pareto front solution set that is optimal in terms of both network loss and voltage quality.

8. A voltage quality control device for large-scale distributed photovoltaic (PV) grid integration in a distribution network, characterized in that, include: The impact analysis module is used to analyze the impact of distributed photovoltaic access on the voltage of distribution network nodes and network losses, and to determine the voltage exceeding the limit and network loss increase caused by the access. The partitioning module is used to calculate the reactive voltage sensitivity between nodes of the distribution network based on the impact analysis results determined by the impact analysis module, calculate the electrical distance between any two nodes in the distribution network based on the reactive voltage sensitivity, and partition the distribution network based on the electrical distance using a clustering method to obtain the optimal partitioning scheme. The two-layer optimization model construction module is used to establish an upper-layer scheduling model with the optimization objectives of minimizing distribution network losses and optimizing voltage quality, and a lower-layer scheduling model with the optimization objective of optimizing distribution network operation economy, based on the partitioning scheme, thereby constructing a two-layer optimization model. The model solving module is used to iteratively solve the two-layer optimization model. Through the alternating solving and information interaction between the upper-layer scheduling model and the lower-layer scheduling model, a comprehensive scheduling strategy that ensures the voltage quality of the distribution network is qualified and the operation is economically optimal is obtained. The strategy execution module is used to decompose the comprehensive scheduling strategy into specific control commands and send them to the corresponding photovoltaic inverters, energy storage systems, static var generators and load management terminals for execution through the communication module.

9. A voltage quality control system for large-scale distributed photovoltaic (PV) grid integration in a distribution network, comprising: Computer-readable storage media and processors; The computer-readable storage medium is used to store executable instructions; The processor is used to read executable instructions stored in the computer-readable storage medium and execute the voltage quality control method for large-scale distributed photovoltaic access in the distribution network as described in any one of claims 1-7.

10. A non-transitory computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the voltage quality control method for large-scale distributed photovoltaic access in a distribution network as described in any one of claims 1-7.

Images