Low-voltage distribution network reactive power optimization method, system and equipment based on improved particle swarm optimization and medium

Abstract

The invention relates to the technical field of reactive power optimization of a power system, and discloses a low-voltage power distribution network reactive power optimization method, system and device based on an improved particle swarm algorithm, and a medium. The problems that when a traditional multi-objective optimization algorithm converts multi-objective into single-objective solution, the optimization result is not accurately reflected, and local optimal solution and premature convergence are likely to happen are effectively solved. According to the method, the reactive power regulation capability of high-proportion photovoltaic power and the dynamic compensation capability of the static var generator are fully considered, a multi-target reactive power optimization mathematical model is constructed by taking the minimum network loss amount, the minimum voltage offset amount and the maximum photovoltaic absorption amount as targets, and the operation requirements of the power distribution network can be reflected more comprehensively and accurately.

Description

Technical Field

[0001] This invention relates to the field of reactive power optimization technology in power systems, and in particular to a method, system, equipment, and medium for reactive power optimization of low-voltage distribution networks based on an improved particle swarm optimization algorithm. Background Technology

[0002] With the widespread integration of distributed photovoltaic (PV) power into low-voltage distribution networks, the power flow direction of the system becomes uncertain, node voltages face the risk of exceeding limits, and network losses increase. Traditional reactive power optimization methods are insufficient to address these issues arising from PV integration, and relying solely on PV output to adjust system reactive power has limited compensation effectiveness. Therefore, it is necessary to incorporate reactive power compensation devices and achieve coordinated optimization to ensure the stable operation of low-voltage distribution networks.

[0003] While some progress has been made in using intelligent optimization algorithms to solve reactive power optimization problems, most existing methods transform multi-objective problems into single-objective problems by using fuzzy theory, weighting methods, etc. This approach does not accurately reflect the optimization results and is prone to getting trapped in local optima, premature convergence, and other problems, which cannot fully meet the actual needs of reactive power optimization in distribution networks. Summary of the Invention

[0004] In view of the aforementioned existing problems, the present invention is proposed.

[0005] Therefore, this invention provides a method, system, equipment, and medium for reactive power optimization of low-voltage distribution networks based on an improved particle swarm optimization algorithm. This method can solve problems such as voltage exceeding limits and increased network losses that occur in distribution networks after distributed photovoltaic access. At the same time, it overcomes the defects of traditional multi-objective optimization algorithms and improves the operational stability and economy of distribution networks.

[0006] To solve the above-mentioned technical problems, the present invention provides the following technical solution: In a first aspect, the present invention provides a reactive power optimization method for low-voltage distribution networks based on an improved particle swarm optimization algorithm, comprising: Obtain operational data of the low-voltage distribution network, including distributed photovoltaic access information, load distribution, and network topology. Based on the operation data of low-voltage distribution networks and the reactive power regulation capability of high-proportion photovoltaics and the dynamic compensation capability of static var generators, a multi-objective reactive power optimization mathematical model is constructed with the objectives of minimizing network loss, minimizing voltage deviation, and maximizing photovoltaic absorption. An improved multi-objective particle swarm optimization algorithm that integrates the Pareto optimal solution mechanism and niche technology is used to solve the multi-objective reactive power optimization mathematical model to obtain a non-dominated solution set; The reactive power control strategy is determined based on the non-dominated solution set, and reactive power optimization control is implemented on the low-voltage distribution network.

[0007] As a preferred embodiment of the low-voltage distribution network reactive power optimization method based on the improved particle swarm optimization algorithm described in this invention, the construction of a multi-objective reactive power optimization mathematical model with the objectives of minimizing network losses, minimizing voltage deviation, and maximizing photovoltaic absorption includes: The reactive power output of the photovoltaic inverter and the reactive power output of the static var generator are used as collaborative decision variables. A multi-objective optimization problem is established, taking the total active power loss of the distribution network, the degree of voltage deviation of each node from the rated value, and the actual amount of electricity consumed by distributed photovoltaic power as optimization objectives.

[0008] This optimized solution fully leverages the synergistic effect of photovoltaic inverters and static var generators (SVA) to achieve precise control of reactive power in low-voltage distribution networks. By using their reactive power output as collaborative decision variables, their characteristics and capabilities can be comprehensively considered, and reactive power output can be dynamically adjusted according to the actual operating conditions of the distribution network, avoiding the limitations of single-device control. Using the total active power loss of the distribution network, the degree of voltage deviation from rated values ​​at each node, and the actual amount of electricity consumed by distributed photovoltaic power as optimization targets, the operating status and performance of the distribution network can be comprehensively measured.

[0009] As a preferred embodiment of the low-voltage distribution network reactive power optimization method based on the improved particle swarm optimization algorithm described in this invention, the step of solving the multi-objective reactive power optimization mathematical model using the improved multi-objective particle swarm optimization algorithm that integrates the Pareto optimal solution mechanism and niche technology includes: During particle evolution, non-dominated solutions are selected based on Pareto dominance, and an archive of external elite solutions is maintained. We introduce niche technology to cluster and control the distance of particle populations, thus avoiding getting trapped in local optima.

[0010] As a preferred embodiment of the low-voltage distribution network reactive power optimization method based on the improved particle swarm optimization algorithm described in this invention, the niche technology calculates the target spatial distance between individual particles to eliminate or guide overly dense areas, thereby improving the uniformity and coverage of the solution set distribution.

[0011] As a preferred embodiment of the low-voltage distribution network reactive power optimization method based on the improved particle swarm optimization algorithm described in this invention, wherein: the non-dominated solution set determines the reactive power control strategy, including: Based on operator preferences, weight allocation, or equivalent trade-offs, a final implementation solution is selected from the non-dominated solution set. The final implementation solution is then converted into specific photovoltaic reactive power commands and static var generator control commands.

[0012] As a preferred embodiment of the low-voltage distribution network reactive power optimization method based on the improved particle swarm optimization algorithm described in this invention, the operating data further includes node voltage limits, line parameters, reactive power compensation device capacity and its response characteristics.

[0013] As a preferred embodiment of the low-voltage distribution network reactive power optimization method based on the improved particle swarm optimization algorithm described in this invention, the method further includes: after implementing reactive power optimization control, verifying the effect of distribution network voltage level, network loss and photovoltaic absorption, and adjusting the model parameters or algorithm strategy online based on the verification results.

[0014] Secondly, the present invention provides a low-voltage distribution network reactive power optimization system based on an improved particle swarm optimization algorithm, comprising: The data acquisition module is used to collect information on the operating status of the low-voltage distribution network and the access information of distributed energy resources. The model building module is used to establish a multi-objective reactive power optimization model that considers the coordinated control of photovoltaic and static var generators. The optimization solution module is used to call the improved particle swarm optimization algorithm that integrates Pareto optimality and niche mechanism to solve the model; The instruction generation and execution module is used to generate and issue reactive power control instructions to field equipment based on the optimization results.

[0015] Thirdly, the present invention provides an electronic device including a memory and a processor, wherein the memory stores a computer program, and the processor executes the computer program to implement the steps of the method described above.

[0016] Fourthly, the present invention provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the steps of the method described above.

[0017] Compared with existing technologies, the beneficial effects of this invention are as follows: This invention proposes a reactive power optimization method for low-voltage distribution networks based on an improved particle swarm optimization algorithm. By integrating the Pareto optimal solution mechanism and niche technology, it effectively solves the problems of inaccurate reflection of optimization results, easy getting trapped in local optima, and premature convergence that exist in traditional multi-objective optimization algorithms when converting multi-objectives into single-objective solutions. This method fully considers the reactive power regulation capability of high-proportion photovoltaics and the dynamic compensation capability of static var generators. It constructs a multi-objective reactive power optimization mathematical model with the objectives of minimizing network losses, minimizing voltage deviation, and maximizing photovoltaic absorption, which can more comprehensively and accurately reflect the operational needs of the distribution network. Attached Figure Description

[0018] To more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings used in the following description of the embodiments will be briefly introduced. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0019] Figure 1 The present invention provides a flowchart of a method for reactive power optimization in low-voltage distribution networks based on an improved particle swarm optimization algorithm, as an embodiment of the present invention.

[0020] Figure 2 The image shows the voltage fluctuation results at 1 hour for a low-voltage distribution network reactive power optimization method based on an improved particle swarm optimization algorithm, as provided in an embodiment of the present invention.

[0021] Figure 3 The image shows the 12-hour voltage fluctuation results of a low-voltage distribution network reactive power optimization method based on an improved particle swarm optimization algorithm, as provided in an embodiment of the present invention.

[0022] Figure 4 The image shows the 24-hour voltage fluctuation results of a low-voltage distribution network reactive power optimization method based on an improved particle swarm optimization algorithm, as provided in an embodiment of the present invention.

[0023] Figure 5 The negative network loss result diagram is provided for a low-voltage distribution network reactive power optimization method based on an improved particle swarm optimization algorithm, as an embodiment of the present invention.

[0024] Figure 6 The photovoltaic absorption result diagram is provided by a low-voltage distribution network reactive power optimization method based on an improved particle swarm optimization algorithm, according to an embodiment of the present invention.

[0025] Figure 7 This is an internal structure diagram of an electronic device for a low-voltage distribution network reactive power optimization method based on an improved particle swarm optimization algorithm, as provided in one embodiment of the present invention. Detailed Implementation

[0026] To make the above-mentioned objects, features, and advantages of the present invention more apparent and understandable, specific embodiments of the present invention will be described in detail below with reference to the accompanying drawings. Obviously, the described embodiments are only a part of the embodiments of the present invention, and not all of them. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort should fall within the protection scope of the present invention.

[0027] Example 1, referring to Figure 1 This is the first embodiment of the present invention, which provides a reactive power optimization method for low-voltage distribution networks based on an improved particle swarm optimization algorithm, comprising: Existing technologies have several limitations. Traditional reactive power optimization methods struggle to cope with the uncertainties in power flow direction in distribution networks after large-scale integration of distributed photovoltaic systems, failing to effectively address issues such as voltage exceeding limits and increased network losses. Furthermore, existing intelligent optimization algorithms, when solving multi-objective reactive power optimization problems, mostly employ fuzzy theory and weighted methods to transform multiple objectives into single-objective solutions. This approach not only inaccurately reflects the optimization results but also easily gets trapped in local optima, leading to premature convergence and failing to meet the practical needs of reactive power optimization in distribution networks.

[0028] This invention provides a method that can effectively solve the problems mentioned above. The following will describe in detail how to implement this low-voltage distribution network reactive power optimization method based on the improved particle swarm optimization algorithm, using multiple embodiments. Figure 1 A flowchart of a reactive power optimization method for low-voltage distribution networks based on an improved particle swarm optimization algorithm is shown, including: S101, acquire the operation data of the low-voltage distribution network, including distributed photovoltaic access information, load distribution and network topology; In this embodiment of the invention, the operating data also includes node voltage limits, line parameters, reactive power compensation device capacity and its response characteristics.

[0029] In an optional implementation, the access location, installed capacity, and reactive power regulation capability of each distributed photovoltaic system in the low-voltage distribution network can be obtained. For example, a 50kW photovoltaic system connected to a certain node has a power factor regulation range of ±0.35. The active and reactive power distribution and its time-series variation characteristics of each load node can be obtained. For example, the load in the residential area shows obvious evening peak characteristics from 18:00 to 22:00. In one optional implementation, the network topology of the distribution network is obtained, including line connection relationships, switch status, and the hierarchical structure of main lines and branch lines, such as the radial connection from node 1 to node 33 in the IEEE 33-node system; the voltage limits of each node are also obtained, such as the range of 0.95 to 1.1 per-unit values ​​specified by national standards; line parameters are obtained, such as the resistance, reactance, and length data of each feeder segment; and the capacity and response characteristics of the reactive power compensation device are obtained, for example, the rated capacity of a static var generator (SVG) is ±100 kvar, and the response time is less than 20 ms.

[0030] It should be noted that this step provides a complete and accurate physical and operational foundation for subsequent modeling and optimization. Only by understanding the location and capacity of photovoltaic grid connection, the spatiotemporal distribution characteristics of load, and network connectivity can the system power flow, voltage distribution, and reactive power potential be accurately characterized. This ensures that the constructed optimization model truly reflects the actual operating state of the distribution network and avoids infeasible optimization results or performance degradation due to missing or distorted data.

[0031] S102, based on the operation data of low-voltage distribution network and the reactive power regulation capability of high-proportion photovoltaic and the dynamic compensation capability of static var generator, a multi-objective reactive power optimization mathematical model is constructed with the objectives of minimizing network loss, minimizing voltage deviation and maximizing photovoltaic absorption. In this embodiment of the invention, the multi-objective reactive power optimization mathematical model, which aims to minimize grid loss, minimize voltage deviation, and maximize photovoltaic absorption, includes: The reactive power output of the photovoltaic inverter and the reactive power output of the static var generator are used as collaborative decision variables. A multi-objective optimization problem is established, taking the total active power loss of the distribution network, the degree of voltage deviation of each node from the rated value, and the actual amount of electricity consumed by distributed photovoltaic power as optimization objectives.

[0032] In one optional implementation, the reactive power output of the photovoltaic inverter can be set as a continuously adjustable variable, the adjustment range of which is constrained by the inverter's rated capacity and current active power output. For example, a photovoltaic inverter with a rated capacity of 30kVA and a current active power output of 20kW has a maximum reactive power output of ±22.36kvar. The reactive power output of the static var generator (SVG) can be set as an independent controllable variable, the value range of which is determined by the device's rated capacity. For example, an SVG with a rated capacity of ±50kvar can be dynamically adjusted between -50kvar and +50kvar. In one optional implementation, the total active power loss of all branches of the distribution network is used as the network loss target. The active power loss of the lines is accumulated segment by segment through power flow calculation. For example, the total I²R loss of 32 branches in the IEEE 33-node system. The sum of the absolute deviations of the voltage amplitude of each node from the 1.0 per-unit value is used as the voltage offset target. For example, the accumulated deviation of the voltage of each node from 1.0 pu in the 33 nodes. The active power of distributed photovoltaic power actually accepted by the grid is used as the photovoltaic absorption target. For example, the total amount of power actually output by all photovoltaic nodes that has not been curtailed during the 24-hour dispatch cycle. Through the above three objective functions and two types of collaborative decision variables, a multi-objective reactive power optimization mathematical model is constructed, forming an optimization problem that includes nonlinear constraints and multi-objective conflicts.

[0033] It should be noted that this step formalizes the physical system characteristics and operational objectives into a computable mathematical expression, providing a clear problem definition for subsequent algorithmic solutions. By incorporating the coordinated control of the photovoltaic inverter and the static var generator into the decision variables, and simultaneously considering economic efficiency (grid loss), safety (voltage), and renewable energy utilization efficiency (consumption), the optimization problem acquires multi-dimensional coordination capabilities, laying the model foundation for generating high-quality control strategies that take into account multiple objectives.

[0034] S103, an improved multi-objective particle swarm optimization algorithm that integrates Pareto optimal solution mechanism and niche technology is used to solve the multi-objective reactive power optimization mathematical model to obtain a non-dominated solution set; In this embodiment of the invention, the improved multi-objective particle swarm optimization algorithm, which integrates the Pareto optimal solution mechanism and niche technology, is used to solve the multi-objective reactive power optimization mathematical model, including: During particle evolution, non-dominated solutions are selected based on Pareto dominance, and an archive of external elite solutions is maintained. We introduce niche technology to cluster and control the distance of particle populations, thus avoiding getting trapped in local optima.

[0035] In an optional implementation, during particle evolution, all individual particles can be compared pairwise based on Pareto dominance to select non-dominated solutions that are not dominated by other solutions. For example, when solution A is superior to or equal to solution B in both network loss and voltage offset, and is strictly superior in at least one objective, A is determined to dominate B, and the undominated solutions are stored in the external elite solution archive. The external elite solution archive is dynamically updated and its capacity is controlled. For example, when the archive exceeds a preset limit of 100, a pruning strategy based on crowding distance is used to retain uniformly distributed non-dominated solutions. Niche techniques are introduced to further refine the solution archive. The particle population is clustered according to the target spatial distance. For example, the niche radius is set to 0.05 per unit, and only the particles with the best fitness are retained within this radius to suppress population aggregation. The global optimal guiding term in the particle velocity update formula is adjusted through a niche sharing mechanism. For example, each particle selects a guiding individual only from the elite solutions in the same niche, thereby enhancing population diversity and avoiding the algorithm from converging to the local Pareto front in the early stage. Finally, at the end of the iteration, a uniformly distributed and well-converged Pareto optimal solution set is output to support multi-objective reactive power optimization decision-making.

[0036] In this embodiment of the invention, the niche technology calculates the target spatial distance between individual particles to eliminate or guide overly dense areas, thereby improving the uniformity and coverage of the solution set distribution.

[0037] In an optional implementation, the Euclidean distance between the vectors formed by the objective function values ​​of individual particles in the target space can be calculated. For example, the Euclidean distance can be calculated for the three-dimensional vectors formed by the grid loss, voltage offset, and photovoltaic absorption of two particles, respectively. When the distance is less than a preset niche radius threshold, the two particles are determined to be in an overly dense region. For example, if the niche radius is set to 0.03 per unit, and the distance between the target vectors of the two particles is 0.025, it is considered dense. Particles in the dense region are processed by a elimination mechanism or a guidance mechanism. For example, particles with better overall performance are retained and the rest are deleted, or the velocity direction of the suboptimal particles is adjusted to move them to sparse regions. This process dynamically maintains the uniformity of the population distribution in the target space. For example, in the high curvature region of the Pareto front, excessive concentration of solutions is avoided, while in the flat region, solution loss is prevented. This improves the coverage and distribution breadth of the final non-dominated solution set to the true Pareto front.

[0038] It should be noted that this step effectively handles conflicts and nonlinear constraints between objectives through advanced intelligent algorithms, avoiding information loss or local optimum traps caused by traditional weighted methods. The Pareto mechanism ensures that the solution set covers the true optimal frontier, and the niche technique improves the diversity and uniformity of solution distribution, thereby providing a set of high-quality and comparable candidate solutions for subsequent decision-making, enhancing the flexibility and robustness of the control strategy.

[0039] S104 determines the reactive power control strategy based on the non-dominated solution set and implements reactive power optimization control on the low-voltage distribution network.

[0040] In embodiments of the present invention, determining the reactive power control strategy using the non-dominated solution set includes: Based on operator preferences, weight allocation, or equivalent compromise principles, a final implementation solution is selected from the non-dominated solution set. The final implementation solution is then converted into specific photovoltaic reactive power commands and static var generator control commands.

[0041] In one alternative implementation, based on the operator's preference for prioritizing voltage stability, the solution with the smallest voltage deviation can be selected as the final implementation solution from the set of non-dominated solutions. For example, in the Pareto front, the solution with a voltage deviation below 0.02 per unit and a grid loss increase of no more than 5% can be selected. Alternatively, the three objectives can be weighted according to a preset weighting scheme. For example, grid loss can be weighted by 0.4, voltage deviation by 0.4, and photovoltaic absorption by 0.2. The comprehensive score of each non-dominated solution can be calculated, and the solution with the highest score can be selected. Alternatively, an equivalent compromise principle can be adopted to select the objective function values ​​that are close to the ideal. For example, by calculating the normalized Euclidean distance from each non-dominated solution to the ideal point (minimum network loss, minimum voltage deviation, maximum absorption), the solution with the smallest distance is selected. The reactive power output value of the photovoltaic inverter in the selected final implementation solution is converted into the corresponding power factor command or reactive power setpoint. For example, +15kvar reactive power output is converted into a power factor 0.95 lag command and sent to the photovoltaic inverter controller. At the same time, the reactive power output value of the static var generator is converted into specific current or reactive power control commands. For example, a +30kvar reactive power compensation command is sent to the SVG to achieve dynamic voltage support.

[0042] In this embodiment of the invention, after implementing reactive power optimization control, the method further includes verifying the effects on the distribution network voltage level, network losses, and photovoltaic absorption, and adjusting the model parameters or algorithm strategy online based on the verification results. This can be achieved through the following steps: Step 1.1: Collect real-time operating data after the reactive power optimization control is executed, including the voltage amplitude of each node, the active power of the branch, and the actual output value of the photovoltaic system, in order to evaluate the control effect.

[0043] Step 1.2: Calculate the total active power loss of the current distribution network, count the number and degree of voltage over-limit nodes, and calculate the actual power consumption of distributed photovoltaic power to form the quantitative results of three key indicators.

[0044] Specifically, the active power of each branch can be measured in real time and accumulated segment by segment to obtain the total active power loss of the entire distribution network. For example, in the IEEE 33-node system, smart meters collect current and voltage data of each branch, calculate I²R losses, and add up the losses of all branches. At the same time, the voltage amplitude of each node is monitored to identify voltage-limited nodes that are below 0.95 or above 1.1 per unit, and record their number and specific degree of exceeding the limit. For example, the voltage of a node is 1.12 pu, exceeding the upper limit by 0.02 pu. For the actual electricity consumed by distributed photovoltaic power, the actual amount of electricity connected to the grid can be determined by comparing the inverter output records with the data from the grid dispatch center. For example, if a photovoltaic power station generates a total of 100 MWh of electricity in a day, of which 95 MWh is accepted by the grid and the remaining 5 MWh is reduced due to regulation needs, these quantitative results will serve as key indicators for evaluating the effectiveness of reactive power optimization control.

[0045] Step 1.3: Compare the above quantitative results with the baseline values ​​and preset operating standards before optimization to determine whether the expected goals of improving voltage qualification rate, reducing grid loss and increasing photovoltaic consumption have been achieved.

[0046] Specifically, the current network loss can be compared with the baseline value before optimization. For example, if the network loss before optimization was 0.54MW, and the current network loss has decreased to 0.38MW, it indicates that the network loss has been reduced by approximately 29.63%. Similarly, the voltage qualification rate, i.e., the proportion of nodes with voltage within the range of 0.95 to 1.1 per-unit values, can be statistically analyzed. For example, if the voltage qualification rate was 90% before optimization and increased to 95% after optimization, it indicates that voltage stability has been improved. For photovoltaic (PV) absorption, the proportion of actual absorbed electricity to total power generation can be calculated and compared with the preset target. For example, if the preset target is a PV absorption rate of 85%, and the actual absorption rate after optimization is 85.71%, it indicates that the expected target has been achieved. If any of the above indicators fails to meet the preset standard, further analysis of the reasons is required, such as adjusting model parameters or improving algorithm strategies, to ensure that subsequent optimization iterations can better meet the operational needs of the distribution network. In this process, through precise quantification and comparative analysis of various indicators, the actual effectiveness of reactive power optimization control can be clearly defined, providing a scientific basis for subsequent adjustments. Among them, the benchmark value refers to the actual value of each performance indicator before optimization, and the preset operation standard refers to the target value set according to the grid operation requirements, including voltage qualification rate, grid loss reduction rate and photovoltaic consumption ratio, etc.

[0047] Step 1.4: If the verification results do not meet the operational requirements or have significant deviations, adjust the weight coefficients, niche radius, or inertia factor of the particle swarm optimization algorithm in the multi-objective optimization model online, and trigger the next round of optimization iteration.

[0048] Based on the above embodiments, step 1.4 can be implemented in the following ways: Step 1.41: Analyze the specific types of indicators that did not meet the standards and the degree of deviation in the verification results, and identify the key factors that led to insufficient performance, such as the number of voltage over-limit nodes still being higher than the threshold or the photovoltaic absorption rate not meeting the expected target.

[0049] Specifically, a detailed data collection and statistical analysis of each key performance indicator can be conducted first. For example, the voltage level of each node can be obtained through a real-time monitoring system, and the number and severity of nodes exceeding voltage limits can be calculated. Simultaneously, the changes in network losses before and after optimization can be compared, and the actual photovoltaic power consumption can be recorded and compared with preset target values. If the voltage compliance rate is found to be below 95% or the photovoltaic power consumption rate is below 85%, further in-depth analysis can be conducted to determine the reasons for these indicators not meeting the targets, such as whether it is due to insufficient reactive power regulation capability of the photovoltaic inverter, response delay of the static var generator, or limitations of the network topology. Through this detailed analysis process, key factors affecting system performance can be accurately identified, providing a clear direction for subsequent adjustments.

[0050] Step 1.42: Dynamically adjust the target weight coefficients in the multi-objective optimization model according to the type of deviation. If the voltage problem is prominent, increase the weight of the voltage offset target. If the reduction in network loss is insufficient, increase the priority of the network loss target.

[0051] Specifically, based on the key problem types identified in step 2.41, the target weight coefficients in the multi-objective optimization model can be adjusted accordingly. For example, when voltage exceedance is severe, the weight of the voltage deviation target can be increased from 0.4 to 0.6 to emphasize the improvement of voltage stability; if the reduction in network loss is not significant, the weight of the network loss target can be increased, such as from 0.3 to 0.5, to strengthen the energy-saving and consumption-reducing target. During the adjustment process, ensure that the sum of the weights of each target remains constant, typically 1. Furthermore, a dynamic weight adjustment mechanism can be introduced to gradually fine-tune the weights based on the results of each iteration, making the optimization process more flexible and better suited to actual operational needs. In this way, the optimization model can focus more on the most pressing problems, thereby improving the effectiveness of the overall control strategy.

[0052] Step 1.43: Simultaneously adjust and improve the control parameters of the particle swarm optimization algorithm, including reducing or increasing the habitat radius to improve the solution set distribution density, or modifying the inertia factor to balance the algorithm's global exploration and local exploitation capabilities.

[0053] Specifically, the control parameters of the Improved Particle Swarm Optimization (MOPSO) algorithm can be dynamically adjusted based on the solution set distribution and convergence speed from the previous optimization round. For example, if the solution set is found to be too dense, it indicates that the algorithm may be trapped in a local optimum. In this case, the niche radius should be appropriately increased, such as from 0.03 per unit to 0.05 per unit, to encourage particles to migrate to a wider search space, thereby increasing the diversity of the solution set. Conversely, if the solution set is too scattered, the niche radius should be decreased, such as from 0.05 to 0.03, to encourage particles to gather near better solutions, forming a more compact Pareto front. Furthermore, the inertia factor can be adjusted to balance the ability to explore globally and exploit locally. For example, a larger inertia factor, such as 0.9, can be set in the initial stage to promote global search, while the inertia factor can be gradually decreased in the later stages, such as to 0.4, to better refine local optima. Through this refined parameter adjustment strategy, the adaptability and robustness of the algorithm can be significantly improved, ensuring that it can still efficiently solve multi-objective optimization problems in complex power system environments. Among them, the niche radius is a threshold used to measure the distance between particles, and the inertia factor is a parameter that controls the inertia of particle motion. The two together affect the convergence characteristics and solution quality of the particle swarm algorithm.

[0054] Step 1.44: Load the updated model parameters and algorithm configuration into the optimization solution module, immediately start a new round of optimization iteration, generate a new non-dominated solution set that adapts to the current running state, and regenerate control instructions for execution.

[0055] It should be noted that this step transforms the algorithm output into executable device instructions, realizing a closed-loop implementation from theoretical optimization to actual operation. Based on the selection mechanism of non-dominated solution set, such as operating preferences or trade-off principles, it ensures that the final strategy meets the needs of on-site operation. The precisely issued photovoltaic reactive power and SVG control instructions can directly improve voltage levels, reduce losses, and increase photovoltaic utilization, thus transforming all the aforementioned efforts in data acquisition, modeling, and solving into actual operational benefits.

[0056] In summary, this invention proposes a reactive power optimization method for low-voltage distribution networks based on an improved particle swarm optimization algorithm. By integrating the Pareto optimal solution mechanism and niche technology, it effectively solves the problems of inaccurate reflection of optimization results, susceptibility to local optima, and premature convergence that exist in traditional multi-objective optimization algorithms when transforming multi-objective solutions into single-objective solutions. This method fully considers the reactive power regulation capability of high-proportion photovoltaic power and the dynamic compensation capability of static var generators. It constructs a multi-objective reactive power optimization mathematical model with the objectives of minimizing network losses, minimizing voltage deviation, and maximizing photovoltaic absorption, thus more comprehensively and accurately reflecting the operational needs of the distribution network.

[0057] Example 2, refer to Figures 2-6The present invention can be implemented by designing the following specific operating steps based on the method in Example 1: A coordinated control strategy of high-proportion photovoltaic reactive power output and static var generator is adopted to establish a multi-objective optimization mathematical model with the objectives of minimizing grid loss, minimizing voltage deviation, and maximizing high-proportion photovoltaic absorption. The objective function is solved using a multi-objective particle swarm optimization algorithm that combines Pareto optimality and niche techniques. The control strategy can effectively reduce network losses, improve voltage stability and photovoltaic absorption, and is verified by simulation of the IEEE 33-node distribution network system. The improved algorithm has significant advantages in solving multi-objective reactive power optimization problems and can provide reliable technical support for reactive power optimization of distribution networks with distributed photovoltaic access.

[0058] Preferably, a multi-objective particle swarm optimization algorithm based on Pareto optimality and niche technology is used to solve the objective function.

[0059] Preferably, its implementation can be achieved by calling the following formula: In the formula, This represents a threshold value; when the interparticle distance is less than this threshold, the habitat is added. This represents the number of particles in an individual within a niche. , Each represents any two particles.

[0060] Depend on Figure 2 , 3 As shown in section 4, the voltage fluctuation results before and after adopting the method in this paper are significantly different. According to the IEEE standard, the critical voltage range is set to 0.95~1.1 per unit. The node voltage fluctuation results at 1 hour, 12 hours and 24 hours are compared according to the "before-middle-after" principle. After the algorithm in this paper is optimized, the voltage does not exceed the limit, while the network without algorithm optimization has 9 times, 8 times and 13 times respectively.

[0061] Depend on Figure 5 It can be seen that the network loss without algorithm optimization is 0.54 MW, while the network loss after optimization by the algorithm in this paper is 0.38 MW, and the network loss reduction rate is 29.63%.

[0062] Depend on Figure 6 It can be seen that the total photovoltaic capacity connected to the network is 0.21 MW, the absorption capacity is 0.18 MW, and the absorption rate is increased to 85.71%.

[0063] Therefore, this invention adopts a low-voltage distribution network reactive power optimization method based on an improved particle swarm optimization algorithm as described in Embodiment 1 above. First, a coordinated control strategy of high-proportion photovoltaic reactive power output and static var generators is employed to establish a multi-objective optimization mathematical model with the objectives of minimizing network losses, minimizing voltage deviation, and maximizing high-proportion photovoltaic absorption. Second, a multi-objective particle swarm optimization algorithm combining Pareto optimality and niche technology is used to solve the objective function. Finally, simulation verification using an IEEE 33-node distribution network system demonstrates that this control strategy can effectively reduce network losses, improve voltage stability, and enhance photovoltaic absorption. The improved algorithm shows significant advantages in solving multi-objective reactive power optimization problems and can provide reliable technical support for reactive power optimization in distribution networks with distributed photovoltaic access.

[0064] Example 3, referring to Figure 7 This embodiment also provides a low-voltage distribution network reactive power optimization system based on an improved particle swarm optimization algorithm, including: The data acquisition module is used to collect information on the operating status of the low-voltage distribution network and the access information of distributed energy resources. The model building module is used to establish a multi-objective reactive power optimization model that considers the coordinated control of photovoltaic and static var generators. The optimization solution module is used to call the improved particle swarm optimization algorithm that integrates Pareto optimality and niche mechanism to solve the model; The instruction generation and execution module is used to generate and issue reactive power control instructions to field equipment based on the optimization results.

[0065] The above-mentioned unit modules can be embedded in the processor of the electronic device in hardware form or independent of it, or they can be stored in the memory of the electronic device in software form, so that the processor can call and execute the corresponding operations of the above modules.

[0066] This embodiment also provides an electronic device, which can be a terminal, and its internal structure diagram can be as follows: Figure 7As shown, the electronic device includes a processor, memory, communication interface, display screen, and input devices connected via a system bus. The processor provides computing and control capabilities. The memory includes a non-volatile storage medium and internal memory. The non-volatile storage medium stores the operating system and computer programs. The internal memory provides an environment for the operation of the operating system and computer programs stored in the non-volatile storage medium. The communication interface is used for wired or wireless communication with external terminals; wireless communication can be achieved through Wi-Fi, carrier networks, NFC (Near Field Communication), or other technologies. When the computer program is executed by the processor, it implements a reactive power optimization method for low-voltage distribution networks based on an improved particle swarm optimization algorithm. The display screen can be an LCD screen or an e-ink screen. The input devices can be a touch layer covering the display screen, buttons, a trackball, or a touchpad on the device's casing, or an external keyboard, touchpad, or mouse.

[0067] This embodiment also provides a computer-readable storage medium on which a computer program is stored, and when the computer program is executed by a processor, it performs the following steps: Obtain operational data of the low-voltage distribution network, including distributed photovoltaic access information, load distribution, and network topology. Based on the operation data of low-voltage distribution networks and the reactive power regulation capability of high-proportion photovoltaics and the dynamic compensation capability of static var generators, a multi-objective reactive power optimization mathematical model is constructed with the objectives of minimizing network loss, minimizing voltage deviation, and maximizing photovoltaic absorption. An improved multi-objective particle swarm optimization algorithm that integrates the Pareto optimal solution mechanism and niche technology is used to solve the multi-objective reactive power optimization mathematical model and obtain the non-dominated solution set. The reactive power control strategy is determined based on the non-dominated solution set, and reactive power optimization control is implemented on the low-voltage distribution network.

[0068] It should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and are not intended to limit it. Although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can be made to the technical solutions of the present invention without departing from the spirit and scope of the technical solutions of the present invention, and all such modifications or substitutions should be covered within the scope of the claims of the present invention.

[0069] Although preferred embodiments of the invention have been described, those skilled in the art, upon learning the basic inventive concept, can make other changes and modifications to these embodiments. Therefore, the appended claims are intended to be interpreted as including both the preferred embodiments and all changes and modifications falling within the scope of the invention.

[0070] Obviously, those skilled in the art can make various modifications and variations to this invention without departing from its spirit and scope. Therefore, if these modifications and variations fall within the scope of the claims of this invention and their equivalents, this invention also intends to include these modifications and variations.

Claims

1. A reactive power optimization method for low-voltage distribution networks based on an improved particle swarm optimization algorithm, characterized in that, include: Obtain operational data of the low-voltage distribution network, including distributed photovoltaic access information, load distribution, and network topology. Based on the operation data of low-voltage distribution networks and the reactive power regulation capability of high-proportion photovoltaics and the dynamic compensation capability of static var generators, a multi-objective reactive power optimization mathematical model is constructed with the objectives of minimizing network loss, minimizing voltage deviation, and maximizing photovoltaic absorption. An improved multi-objective particle swarm optimization algorithm that integrates the Pareto optimal solution mechanism and niche technology is used to solve the multi-objective reactive power optimization mathematical model to obtain a non-dominated solution set; The reactive power control strategy is determined based on the non-dominated solution set, and reactive power optimization control is implemented on the low-voltage distribution network.

2. The reactive power optimization method for low-voltage distribution networks based on an improved particle swarm optimization algorithm as described in claim 1, characterized in that, The multi-objective reactive power optimization mathematical model, which aims to minimize grid loss, minimize voltage deviation, and maximize photovoltaic absorption, includes: The reactive power output of the photovoltaic inverter and the reactive power output of the static var generator are used as collaborative decision variables. A multi-objective optimization problem is established, taking the total active power loss of the distribution network, the degree of voltage deviation of each node from the rated value, and the actual amount of electricity consumed by distributed photovoltaic power as optimization objectives.

3. The reactive power optimization method for low-voltage distribution networks based on an improved particle swarm optimization algorithm as described in claim 2, characterized in that, The improved multi-objective particle swarm optimization algorithm, which integrates the Pareto optimal solution mechanism and niche technology, is used to solve the multi-objective reactive power optimization mathematical model, including: During particle evolution, non-dominated solutions are selected based on Pareto dominance, and an archive of external elite solutions is maintained. We introduce niche technology to cluster and control the distance of particle populations, thus avoiding getting trapped in local optima.

4. The reactive power optimization method for low-voltage distribution networks based on an improved particle swarm optimization algorithm as described in claim 3, characterized in that, The niche technology calculates the target spatial distance between individual particles, and removes or guides overly dense areas, thereby improving the uniformity and coverage of the solution set distribution.

5. The reactive power optimization method for low-voltage distribution networks based on an improved particle swarm optimization algorithm as described in claim 4, characterized in that, The non-dominated solution set determines the reactive power control strategy, including: Based on operator preferences, weight allocation, or equivalent trade-offs, a final implementation solution is selected from the non-dominated solution set. The final implementation solution is then converted into specific photovoltaic reactive power commands and static var generator control commands.

6. The reactive power optimization method for low-voltage distribution networks based on an improved particle swarm optimization algorithm as described in claim 5, characterized in that, The operational data also includes node voltage limits, line parameters, reactive power compensation device capacity, and its response characteristics.

7. The reactive power optimization method for low-voltage distribution networks based on an improved particle swarm optimization algorithm as described in claim 6, characterized in that, After implementing reactive power optimization control, the process also includes verifying the effects on distribution network voltage levels, network losses, and photovoltaic absorption, and adjusting model parameters or algorithm strategies online based on the verification results.

8. A low-voltage distribution network reactive power optimization system based on an improved particle swarm optimization algorithm, using the method described in any one of claims 1 to 7, characterized in that, include: The data acquisition module is used to collect information on the operating status of the low-voltage distribution network and the access information of distributed energy resources. The model building module is used to establish a multi-objective reactive power optimization model that considers the coordinated control of photovoltaic and static var generators. The optimization solution module is used to call the improved particle swarm optimization algorithm that integrates Pareto optimality and niche mechanism to solve the model; The instruction generation and execution module is used to generate and issue reactive power control instructions to field equipment based on the optimization results.

9. An electronic device comprising a memory and a processor, wherein the memory stores a computer program, characterized in that, When the processor executes the computer program, it implements the steps of any one of claims 1 to 7 for the reactive power optimization method of a low-voltage distribution network based on an improved particle swarm optimization algorithm.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by the processor, it implements the steps of the low-voltage distribution network reactive power optimization method based on the improved particle swarm algorithm as described in any one of claims 1 to 7.

Images