Voltage reactive power collaborative optimization control method under participation of distributed power supply

Abstract

The invention discloses a voltage reactive power collaborative optimization control method under the participation of a distributed power supply, and the method comprises the steps: obtaining the operation data and prediction data of a power grid, and obtaining the voltage fluctuation trend of a future optimization period according to the prediction data; constructing a comprehensive objective function taking the voltage deviation, the system network loss and the energy storage operation income as optimization objectives, and establishing a collaborative optimization model containing constraints; solving the collaborative optimization model by adopting an improved particle swarm optimization algorithm to obtain an optimal collaborative control strategy including reactive power output set values of the photovoltaic inverters, charging and discharging power of the energy storage systems and an SOC curve; and decomposing the optimal cooperative control strategy into control instructions of different time scales, and issuing the control instructions to the corresponding photovoltaic inverter and the energy storage system for execution so as to realize cooperative regulation and control of the power grid voltage. According to the method, the voltage out-of-limit phenomenon can be effectively eliminated, the high efficiency and economical efficiency of voltage control are ensured, and the stability of a power grid and the new energy consumption level are improved.

Description

Technical Field

[0001] This application relates to the field of power technology, and in particular to a voltage and reactive power coordinated optimization control method with the participation of distributed power sources. Background Technology

[0002] Existing grid voltage regulation technologies primarily rely on single-type regulation methods, such as photovoltaic (PV) reactive power regulation or energy storage active power regulation, to achieve voltage regulation. For example, PV reactive power regulation relies solely on PV inverters to provide reactive power support, while energy storage active power regulation relies solely on distributed energy storage (ESS) for active power charging and discharging. The basic idea behind these traditional solutions is:

[0003] PV reactive power regulation strategy: Under the condition that the photovoltaic inverter has a certain reactive power output capacity, the reactive power is adjusted to support or absorb reactive power, thereby alleviating the problem of node voltage exceeding the upper or lower limit. The advantage of this method is that it is simple to implement and can support voltage without reducing the active power output of photovoltaics.

[0004] Energy storage active power regulation strategy: Utilizing the characteristic of distributed energy storage systems to charge during off-peak hours and discharge during peak hours, voltage levels can be raised or lowered through active power regulation. Simultaneously, energy storage can also generate profits from peak shaving and valley filling, as well as electricity price arbitrage.

[0005] Existing voltage regulation systems generally include two core components: 1. Regulation target module: setting the node voltage to be kept within the allowable range (e.g., 0.95~1.05pu); 2. Execution device module: driving the photovoltaic inverter to perform reactive power adjustment or controlling the energy storage to perform active power charging and discharging based on the voltage deviation signal. Figure 1 A typical structural diagram of the prior art is shown. The working principle of the prior art has the following shortcomings:

[0006] (1) Strategy relying solely on PV reactive power regulation: The reactive power regulation capability of PV inverters is constrained by both the inverter's rated capacity and the lower limit of the power factor. When the photovoltaic output is large, its reactive power regulation margin is significantly limited. In scenarios with a high proportion of photovoltaic grid connection, relying solely on PV reactive power support often cannot completely eliminate the problem of node voltage exceeding the limit.

[0007] (2) Strategy relying solely on active power regulation through energy storage: Energy storage can regulate voltage to some extent through active charging and discharging, but its effect is limited by energy storage capacity, charging and discharging power limits, and SOC (state of charge) boundaries. Especially when the SOC is close to its extreme value or the energy storage capacity is insufficient, energy storage cannot continuously provide voltage support.

[0008] Existing voltage regulation methods, while effective under certain conditions, have inherent limitations that prevent them from meeting the complex requirements of high-proportion distributed power supply integration. Specifically:

[0009] (1) Limited control methods and insufficient robustness: Existing voltage control methods often rely on a single means, such as using reactive power regulation of PV inverters to support voltage, or using active power charging and discharging of energy storage systems to raise or lower voltage. Since PV reactive power is limited by inverter capacity and the lower limit of the power factor, its reactive power margin is severely insufficient when PV output is high; energy storage is limited by the SOC boundary and charging / discharging power capability, and cannot continuously provide voltage support when capacity is insufficient or operating time is limited. This single-reliance approach often fails to guarantee system voltage stability under limited equipment capacity or extreme operating conditions. It has the following disadvantages: reduced control capability under equipment constraints; lack of redundancy support when a certain type of resource fails; and the grid is prone to voltage over-limiting or uncontrolled fluctuations under high-penetration PV conditions.

[0010] (2) Lack of multi-entity coordination and insufficient resource utilization: In existing technologies, PV and energy storage are often controlled independently: PV reactive power is used for voltage support, while energy storage active power is used for peak shaving or arbitrage. This separate control method does not coordinate actions under a unified optimization framework, resulting in the incomplete exploitation of the regulation potential of different resources. For example, when the PV reactive power margin is insufficient in certain periods, energy storage fails to participate in compensation in a timely manner, resulting in insufficient voltage regulation margin in the system. It has the following disadvantages: lack of coordination between different entities leads to insufficient voltage regulation margin; low regulation efficiency and low overall resource utilization; voltage regulation effect is limited by the state of a single device and lacks global optimality.

[0011] (3) Inability to comprehensively solve voltage fluctuation problems: Under high-proportion photovoltaic (PV) grid connection conditions, voltage fluctuations exhibit frequent and asymmetrical characteristics: voltage is prone to exceeding the upper limit during the day when sunlight is strong, and may fall below the lower limit at night when there is high load. Traditional single control methods often only work for certain types of problems. For example, PV reactive power regulation is mainly used to alleviate overvoltage, while energy storage discharge is suitable for dealing with undervoltage. Therefore, a single method is difficult to simultaneously address the problem of exceeding the upper and lower limits, and voltage fluctuations are difficult to completely eliminate. It has the following disadvantages: limited effect on voltage quality improvement; inability to address both overvoltage and undervoltage problems; and insufficient grid stability.

[0012] (4) Difficulty in balancing economic efficiency and safety: In existing schemes, the main objective of PV reactive power support is voltage safety, but economic benefits are not considered; while energy storage active power regulation often prioritizes electricity price arbitrage, ignoring voltage stability requirements. This fragmented optimization approach makes it difficult to balance economic efficiency and safety: pursuing economic benefits may neglect the risk of voltage exceeding limits, while emphasizing voltage control may sacrifice economic benefits. It has the following disadvantages: lack of unified safety-economic modeling; dispatch instructions may be biased towards a single objective, reducing the overall operational efficiency of the system; and it is difficult to simultaneously meet the dispatcher's dual requirements for voltage quality and cost-effectiveness in engineering applications. Summary of the Invention

[0013] The purpose of this application is to provide a voltage and reactive power coordinated optimization control method with the participation of distributed power sources. By incorporating the regulation capabilities of photovoltaic reactive power, energy storage active power and other distributed power sources into a unified framework, and combining an improved particle swarm optimization algorithm to achieve fast and global search for control schemes, the method significantly improves the stability, economy and adaptability of the system, and provides an innovative solution for voltage control of power systems with a high proportion of new energy access.

[0014] In a second aspect of this application, a voltage-reactive power coordinated optimization control method with the participation of distributed power sources is provided, comprising:

[0015] The system acquires grid operation data, which includes at least real-time voltage at each node, photovoltaic active power output, and state of charge (SOC) of the energy storage system. Based on the operation data, the system obtains prediction data, which includes at least photovoltaic power output prediction, load prediction, and time-of-use electricity price curves for the future optimization period. Based on the prediction data, the system obtains voltage fluctuation trends for the future optimization period.

[0016] A comprehensive objective function is constructed with voltage deviation, system network loss and energy storage operation revenue as optimization objectives, and a collaborative optimization model is established that includes reactive power output constraints of photovoltaic inverters, charging and discharging power constraints of energy storage systems and SOC constraints of energy storage systems.

[0017] An improved particle swarm optimization algorithm is used to solve the cooperative optimization model to obtain the optimal cooperative control strategy, which includes the reactive power output setpoints of each photovoltaic inverter, the charging and discharging power of each energy storage system, and the SOC curve; wherein, the improved particle swarm optimization algorithm adopts an asymmetric learning factor and / or mutation mechanism.

[0018] The optimal collaborative control strategy is decomposed into control commands at different time scales and sent to the corresponding photovoltaic inverters and energy storage systems for execution, so as to achieve collaborative regulation of grid voltage.

[0019] In a preferred embodiment, the integrated objective function is expressed as follows:

[0020]

[0021] Where, ΔV t P represents the voltage deviation at each node at time t. loss,t R represents the system network loss at time t. ESS,t Let t represent the energy storage operation revenue at time t, and α, β, and γ represent the weighting coefficients of the voltage deviation, system network loss, and energy storage operation revenue, respectively.

[0022] In a preferred embodiment, the reactive power output constraint of the photovoltaic inverter is expressed as follows: Where S PV P is the rated apparent power of the photovoltaic inverter. PV The active power output of the photovoltaic inverter is given by [formula missing]; the charging and discharging power constraint of the energy storage system is expressed as [formula missing]. in These represent the charging power and discharging power of the energy storage system, respectively; the SOC constraint of the energy storage system is expressed as SOC. min ≤SOC t ≤SOC max SOC t The State of Charge (SOC) of the energy storage system at time t represents the state of charge of the system. min The state of charge (SOC) of the energy storage system represents the minimum state of charge. max This indicates the maximum state of charge of the energy storage system.

[0023] In a preferred embodiment, the step of solving the cooperative optimization model using an improved particle swarm optimization algorithm further includes:

[0024] The decision variables of the collaborative optimization model are mapped to the position vectors of particles in the particle swarm;

[0025] The positions and velocities of the particle swarm are randomly initialized within the feasible region defined by the physical operation constraints. The parameters of the improved particle swarm optimization algorithm are set, including the asymmetric learning factor.

[0026] In each iteration, the fitness corresponding to the position vector of each particle is calculated. The fitness is calculated based on the comprehensive objective function of the collaborative optimization model and includes penalties for violating constraints. The individual historical best position of each particle and the collective historical best position of the particle swarm are updated according to the fitness.

[0027] Based on the asymmetric learning factor, the individual historical best position, and the global historical best position, the position of the particle in the next iteration is calculated according to the particle swarm's velocity update formula and position update formula.

[0028] When the maximum number of iterations or fitness convergence condition is met, the iteration is terminated and the final group historical best position is decoded as the optimal cooperative control strategy.

[0029] In a preferred embodiment, the velocity update formula is expressed as follows: The location update formula is expressed as follows: in Let be the velocity vectors of the i-th particle in the k-th iteration. Let be the position vector of the i-th particle in the k-th iteration, ω be the inertia weight, c1(k) be the individual learning factor in the k-th iteration, c2(k) be the group learning factor in the k-th iteration, and r1 and r2 be random numbers uniformly distributed in the interval [0,1]. Let g be the vector of the best historical positions of the i-th particle before the k-th iteration. k Let c1(k) be the globally optimal position vector found by the particle swarm before the kth iteration. In the early stage of the iteration, the value of the individual learning factor c1(k) is greater than the value of the swarm learning factor c2(k). In the later stage of the iteration, the value of the swarm learning factor c2(k) is greater than or equal to the value of the individual learning factor c1(k).

[0030] In a preferred embodiment, the parameters include mutation triggering conditions, and the step of solving the cooperative optimization model using an improved particle swarm optimization algorithm further includes:

[0031] After calculating the individual's historical best position and the group's historical best position, it is determined whether the preset mutation triggering condition is met. If the mutation triggering condition is met, a random mutation operation is performed on the position vector of the selected particle according to the preset probability, and the particle that exceeds the physical operation constraint after mutation is constrained and repaired, and projected back into the feasible domain.

[0032] In a preferred embodiment, the mutation triggering condition includes one or more of the following groups:

[0033] The improvement in fitness of the global optimum is less than a preset threshold in M ​​consecutive iterations;

[0034] The current iteration count satisfies k / K ≥ β, where K is the maximum iteration count and β is a preset coefficient; and

[0035] Diversity is determined based on the variance of particle swarm positions or the average distance. If the diversity of the particle swarm is lower than a set threshold, the diversity is determined.

[0036] In a preferred embodiment, the step of decomposing the optimal cooperative control strategy into control instructions at different time scales further includes:

[0037] Based on the local voltage measurement value of the photovoltaic inverter, the photovoltaic inverter is controlled to perform rapid reactive power adjustment in seconds according to the preset reactive power droop control or interval segmentation control strategy.

[0038] If the voltage still exceeds the limit after the second-level rapid adjustment of reactive power, the energy storage system is controlled to perform active power compensation adjustment on a minute-level time scale.

[0039] Based on the predicted data, the energy storage charging and discharging reference power and SOC reference trajectory of the energy storage system within the future prediction period are solved on an hourly basis, and a power adjustment margin is reserved for the minute-level coordinated control.

[0040] In a second aspect of this application, an electronic device is provided, comprising: a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the computer program to implement the method described above.

[0041] In a third aspect of this application, a non-transitory computer-readable storage medium is provided, the non-transitory computer-readable storage medium comprising a computer program that, when executed by a processor, implements the steps of the method as described above.

[0042] Compared with existing traditional voltage control methods that rely solely on PV reactive power regulation or energy storage active power regulation, the voltage and reactive power coordinated optimization control method proposed in this invention, involving distributed power sources, has at least the following significant improvements:

[0043] i) Improved robustness of regulation: Traditional methods are difficult to guarantee voltage stability when equipment capacity is limited or a single means fails. This invention, through the joint optimization of PV and energy storage, forms multi-level and multi-means redundant support, which significantly enhances the robustness of the system under various operating conditions.

[0044] ii) Multi-entity collaborative optimization: Traditional PV and energy storage often operate separately and independently, resulting in low resource utilization. This invention achieves coordinated control of multiple entities through a unified optimization framework, thereby improving overall regulation margin and grid resource utilization efficiency.

[0045] iii) Comprehensive management of voltage fluctuations: Traditional methods are difficult to deal with overvoltage and undervoltage problems at the same time. This invention can effectively suppress frequent and asymmetrical voltage fluctuations and improve voltage quality through the hierarchical synergy of photovoltaic reactive power rapid response and energy storage active power regulation.

[0046] iv) Balancing economy and safety: Existing technologies present a contradiction between economy and voltage stability. This invention introduces dual constraints in the optimization objective, balancing electricity price arbitrage and voltage quality, thereby improving the economic benefits of energy storage while ensuring the safe operation of the power grid.

[0047] It should be understood that, within the scope of this invention, the above-described technical features of this invention and the technical features specifically described below (such as in the embodiments) can be combined with each other to form new or preferred technical solutions. Due to space limitations, they will not be described in detail here. Attached Figure Description

[0048] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the accompanying drawings used in the description of the embodiments or the prior art will be briefly introduced below. It should be understood that the accompanying drawings described below are merely some implementation examples of the present invention, and those skilled in the art can obtain other implementation examples based on these drawings without creative effort.

[0049] Figure 1 This illustrates a technology roadmap for existing technologies that rely on PV reactive power regulation or energy storage active power regulation.

[0050] Figure 2 This is a flowchart illustrating a voltage and reactive power coordinated optimization control method with the participation of distributed power sources according to one embodiment of this application.

[0051] Figure 3 This is a system block diagram of a voltage and reactive power coordinated optimization control method with the participation of distributed power sources according to one embodiment of this application. Detailed Implementation

[0052] In the following description, many technical details are presented to help the reader better understand this application. However, those skilled in the art will understand that the technical solutions claimed in this application can be implemented even without these technical details and various variations and modifications based on the following embodiments.

[0053] With the large-scale integration of new energy sources, especially distributed photovoltaic (PV), into the power grid, the problem of system voltage fluctuations has become increasingly prominent. In scenarios with high PV penetration, traditional methods relying solely on PV reactive power regulation or energy storage active power regulation can alleviate voltage exceedances to some extent. However, due to their limited scope and insufficient resource utilization, they are ill-suited to the operating environment of new power systems characterized by multi-source uncertainties and complex operating conditions. Existing solutions often fail to simultaneously address voltage safety and economic efficiency, lacking multi-entity collaboration and cross-time-domain comprehensive optimization capabilities. Therefore, through extensive and in-depth research, the inventors have proposed a voltage and reactive power collaborative optimization control method based on distributed power source participation. This method fully integrates the regulation capabilities of PV, energy storage, and other distributed power sources, achieving collaborative optimization across multiple objectives and time scales. It not only improves the accuracy and robustness of voltage regulation but also ensures the safe and stable operation of the system under complex conditions, providing more reliable technical support for the development of new power systems.

[0054] This application has at least the following beneficial effects and advantages:

[0055] This invention departs from reliance on a single entity and proposes for the first time a joint optimization control method for photovoltaic reactive power and energy storage active power. It integrates multiple resources, including photovoltaics, energy storage, static var generators (SVG), static var compensators (SVC), capacitor banks, and on-load tap changers (OLTC), to participate in regulation, forming a multi-entity joint redundant system. When the capacity of a single device is insufficient, other resources can compensate, ensuring that the voltage remains within a safe range under different operating conditions, thereby significantly improving system robustness.

[0056] This invention, within a unified collaborative optimization framework, jointly optimizes PV reactive power and energy storage active power, and further introduces multiple types of distributed power sources. Through a hierarchical-regional control architecture, resource usage priorities and switching logic are set to achieve collaborative complementarity among resources. During voltage fluctuations, reactive power resources are prioritized, and energy storage active power is only utilized when reactive power is insufficient, maximizing resource utilization and ensuring the effectiveness and efficiency of voltage regulation.

[0057] This invention constructs a collaborative control chain across time scales by introducing a strategy of "reactive power priority and minimum necessary active power coordination." At the second-level time scale, PV reactive power and droop control respond rapidly; at the minute-level, energy storage active power is optimized in coordination with other regulating devices; and at the hour-level, regulating tasks are allocated through daily rolling optimization. This mechanism enables voltage regulation to both quickly suppress short-term fluctuations and address upper and lower limit exceedance issues globally, achieving a comprehensive solution to voltage fluctuations.

[0058] This invention constructs an explicit multi-objective function in the optimization model, unifying safety objectives such as "minimizing voltage deviation and maximizing VAR reserve margin" with economic objectives such as "minimizing grid loss, minimizing energy storage lifetime loss, and maximizing electricity price revenue" in the model, and achieving global optimization through hierarchical weight coordination. This improves operational economy while ensuring voltage stability, balancing safety and efficiency, and enhancing the overall control level of the power system.

[0059] The above analysis reveals significant shortcomings in existing voltage regulation technologies, including limited control methods, lack of multi-entity coordination, inability to comprehensively address voltage fluctuations, and difficulty in balancing economic efficiency and safety. These limitations make it challenging to meet the operational demands of new power systems with high-proportion distributed generation (DG) integration. These shortcomings can be effectively addressed by the voltage and reactive power collaborative optimization control method with DG participation proposed in this invention. The technical solution of this invention fully utilizes the regulation capabilities of PV, energy storage, and various types of DG to construct a cross-timescale, multi-objective collaborative optimization mechanism. This mechanism enables rapid suppression and global optimization of voltage deviations, improving not only the robustness and accuracy of voltage regulation but also ensuring grid safety and stability while maintaining operational economics. This provides reliable technical support for the high-quality development of new power systems.

[0060] To make the objectives, technical solutions, and advantages of this application clearer, the embodiments of this application will be described in further detail below with reference to the accompanying drawings.

[0061] The first embodiment of this application relates to a voltage and reactive power coordinated optimization control method involving distributed power sources, the process of which is as follows: Figure 2 As shown, the method includes the following steps:

[0062] Step (a) involves acquiring grid operation data, which includes at least the real-time voltage of each node, photovoltaic active power output, and the state of charge (SOC) of the energy storage system. Then, based on the operation data, predictive data is obtained, which includes at least the photovoltaic output forecast, load forecast, and time-of-use pricing curve for the future optimization period. Further, the voltage fluctuation trend for the future optimization period is obtained based on the predictive data.

[0063] Step (b) involves constructing a comprehensive objective function with voltage deviation, system network loss, and energy storage operation revenue as optimization objectives, and establishing a collaborative optimization model that includes reactive power output constraints of photovoltaic inverters, charging and discharging power constraints of energy storage systems, and SOC constraints of energy storage systems.

[0064] Step (c) involves solving the collaborative optimization model using an improved particle swarm optimization (MPSO) algorithm to obtain the optimal collaborative control strategy. This optimal collaborative control strategy includes the optimal collaborative control strategies for the reactive power output setpoints of each photovoltaic inverter, the charging and discharging power of each energy storage system, and the SOC curve. The improved particle swarm optimization algorithm employs an asymmetric learning factor and / or mutation mechanism.

[0065] In a preferred embodiment, the MPSO, based on the standard particle swarm optimization "velocity-position" iterative framework, introduces an asymmetric learning factor and a mutation-repair mechanism triggered by the search state: First, by designing the individual learning factor to decrease with the number of iterations and the swarm learning factor to increase with the number of iterations, particles are biased towards individual exploration in the early stages of iteration and towards swarm convergence in the later stages. Second, when a preset stagnation criterion or insufficient diversity criterion is met, random mutation is performed on the position vectors of some particles, and constraints are repaired (boundary truncation or projection back to the feasible region) for particles that exceed the physical constraints (including photovoltaic inverter capacity constraints, energy storage charging and discharging power constraints, SOC constraints, and node voltage upper and lower limits constraints). At the same time, an over-limit penalty term is introduced into the fitness function to ensure that the search process always converges around a high-quality solution near the feasible region, thus possessing both global optimization capability and stable convergence capability in non-convex, multi-constraint collaborative optimization problems. The asymmetric learning factor and mutation mechanism will be described in more detail below.

[0066] Step (d) decomposes the optimal collaborative control strategy into control commands at different time scales and sends them to the corresponding photovoltaic inverters and energy storage systems for execution, so as to achieve collaborative regulation of grid voltage.

[0067] In a preferred embodiment, the optimal collaborative control strategy is decomposed into "second-level rapid suppression - minute-level collaborative correction - hour-level rolling plan": the second-level layer uses node voltage deviation as a trigger signal, and the photovoltaic inverter performs reactive power droop or segmented control based on the local measured voltage and reference voltage to achieve rapid suppression of short-term voltage disturbances; the minute-level layer is triggered when the voltage still has the risk of exceeding the limit or the inverter's reactive power is saturated (reaching the capacity constraint boundary), and performs minimum deviation correction on the active power charging and discharging power of energy storage to eliminate the limit without significantly deviating from the hour-level plan; the hour-level layer solves the energy storage reference power and SOC reference trajectory based on the predicted data and reserves power adjustment margin for the minute-level layer to call, thereby forming a closed-loop collaborative control link across time scales.

[0068] To address the complex voltage control issues arising from the integration of a high proportion of distributed power sources, this invention proposes a co-evolutionary mechanism based on time-varying strategies and conditional triggering. By designing an asymmetric learning factor that dynamically adjusts with the iteration process (strengthening individual exploration in the early stage and focusing on group convergence in the later stage), and introducing a directional mutation and constraint repair mechanism based on intelligent triggering of the search state, multiple types of distributed power sources, such as photovoltaics and energy storage, are coordinated within a unified multi-objective optimization framework. This enables adaptive closed-loop control that simultaneously achieves voltage safety and stability and optimal operational economy across multiple time scales, including seconds, minutes, and hours.

[0069] To better understand the technical solution of this application, a specific example is provided below. The details listed in this example are mainly for ease of understanding and are not intended to limit the scope of protection of this application.

[0070] This invention proposes a voltage and reactive power coordinated optimization control method involving distributed power sources. Unlike existing single-control methods that rely solely on PV reactive power or energy storage active power, this invention integrates the reactive power regulation capabilities of PV, energy storage, and other distributed power sources to construct a hierarchical coordinated architecture. Furthermore, it achieves cross-timescale voltage stability control through rapid local PV reactive power response and system-level integration of energy storage and multi-entity optimization. This method explicitly balances voltage quality and operational economy, ensuring rapid suppression of voltage exceedances while improving resource utilization and system robustness, providing effective support for the safe and stable operation of the power grid under high-proportion renewable energy integration.

[0071] (1) System Structure Overview

[0072] The voltage-reactive power coordinated optimization control method of the present invention consists of five core parts: monitoring, prediction, optimization, execution, and feedback. This voltage-reactive power coordinated optimization control method achieves this through, as follows: Figure 3 The system implementation is shown. Specifically, the system includes the modules described below, and the modules achieve closed-loop coordination through data interfaces and control commands.

[0073] 1) Voltage monitoring and data acquisition module: used to collect real-time operating data such as voltage, current, photovoltaic output, load level and energy storage SOC of each node, as the basic input for system optimization.

[0074] 2) Prediction and Status Awareness Module: This module is used to perceive and predict voltage fluctuation trends in the future by utilizing photovoltaic power output forecasting, load forecasting, and electricity price information, combined with network power flow and operational constraints.

[0075] 3) Distributed power source collaborative optimization module: Under a unified framework, the reactive power of PV, the active power of energy storage and other distributed regulation devices are jointly optimized to form a collaborative control strategy with multiple subjects and multiple objectives.

[0076] 4) Optimization and Control Decision Module: Based on the improved particle swarm optimization (MPSO) algorithm, it quickly solves optimization problems involving voltage constraints, inverter capacity limitations, and energy storage SOC boundaries, and generates time-segmented control schemes.

[0077] 5) Execution and Feedback Module: The optimization results are sent to the PV inverter, energy storage system and reactive power regulation device to perform voltage regulation in real time, and the model prediction and control effect is corrected through the feedback mechanism.

[0078] (2) Voltage and reactive power co-optimization modeling method

[0079] a. Optimize target modeling

[0080] This invention constructs a collaborative optimization model with the dual objectives of system voltage quality and operational economy. Its optimization objective function is:

[0081] Where ΔV t V represents the voltage V at each node at time t. i,t Compared with reference value V ref The deviation, ΔV t =V i,t -V ref ;P loss,t R represents the network loss of the system at time t; ESS,t This represents the economic benefit of energy storage at time t under the peak-valley electricity price difference; α, β, and γ are the weighting coefficients of each component, used to balance voltage stability and economic benefits.

[0082] b. Control object modeling

[0083] Reactive power output constraints of photovoltaic inverters: Where S PV P is the rated capacity of the photovoltaic inverter. PV They have contributed to the photovoltaic industry.

[0084] Energy storage charging and discharging and SOC constraint: and SOC min ≤SOC t ≤SOC max Among them, SOC t State of charge (SOC) represents the energy storage state at time t. min State of charge (SOC) represents the minimum state of charge of an energy storage system. max Indicates the maximum state of charge of the energy storage system. These represent charging power and discharging power, respectively.

[0085] c. Collaborative optimization strategy modeling

[0086] i. When the measured voltage V at node i i (t) exceeds the allowed range [V] min V max When [the photovoltaic inverter] is in operation, it prioritizes regulation through reactive power, and its segmented control law can be expressed as:

[0087]

[0088] And satisfy the inverter capacity constraint:

[0089]

[0090] in: V represents the reactive power output of the i-th photovoltaic inverter at time t (injection is positive, absorption is negative); i (t) represents the node voltage; V max V min These are the upper and lower voltage limits (e.g., 1.05 pu and 0.95 pu); k q This is the reactive power droop coefficient; The rated apparent power of the inverter; It provides active power to the inverter.

[0091] ii. Energy storage active power compensation strategy: When the reactive power of photovoltaic power is insufficient to maintain the voltage, energy storage compensates for the voltage through charging and discharging.

[0092]

[0093] Where V max V minThese are the upper and lower limits of the voltage, for example, 1.05 pu and 0.95 pu respectively. ch (t), P dis (t) represents the charging power and the discharging power, respectively.

[0094] (3) Optimization algorithms and unified solution framework

[0095] In a preferred embodiment, the present invention employs an improved particle swarm optimization (MPSO) algorithm to solve the voltage-reactive power co-optimization model. The MPSO algorithm, based on the traditional particle swarm optimization algorithm, introduces an asymmetric learning factor and a mutation mechanism to enhance local search in the early stages of algorithm iteration and ensure global convergence in the later stages, thus balancing global optimization capability and convergence stability.

[0096] i. Particle Encoding and Basic Update Formula

[0097] In this embodiment, each particle represents a set of control decision variables to be optimized, such as the photovoltaic reactive power setpoint and the energy storage active power charging and discharging power during each control period. Let the particle swarm size be N. p The position vector of the i-th particle at the k-th iteration is denoted as... The velocity vector is denoted as The update formulas for its velocity and position are as follows:

[0098]

[0099] in Let be the velocity vectors of the i-th particle in the k-th iteration. Let be the position vector of the i-th particle in the k-th iteration, ω be the inertia weight, c1(k) be the individual learning factor in the k-th iteration, c2(k) be the group learning factor in the k-th iteration, and r1 and r2 be random numbers uniformly distributed in the interval [0,1]. Let g be the vector of the best historical positions of the i-th particle before the k-th iteration. k Let be the globally optimal position vector found by the particle swarm before the k-th iteration.

[0100] ii. Asymmetric learning factor design

[0101] To achieve the optimization strategy of "emphasizing individual exploration in the early stage and focusing on group convergence in the later stage", this implementation designes c1(k) and c2(k) as time-varying parameters that change with the number of iterations, specifically adopting the following linear variation form:

[0102]

[0103] Where: K is the preset maximum number of iterations; k is the current iteration number, k = 0, 1, 2…K; c 1,max c 1,min These represent the maximum and minimum values ​​of the individual learning factor, respectively; c 2,max c 2,min These are the maximum and minimum values ​​of the group learning factor, respectively.

[0104] Through the above asymmetric time-varying design, in the early stage of algorithm iteration, c1(k) > c2(k), and particles tend to move closer to their own historical optimal positions, thereby enhancing the local exploration of the solution space; in the later stage of iteration, c2(k) > c1(k), and particles tend to gather near the global optimum, thereby improving the convergence speed and stability.

[0105] In a preferred embodiment, the boundary between the early and late stages of the iteration can be determined as follows:

[0106] (1) Fixed boundary according to iteration ratio: when k / K≤β, it is defined as the early stage of iteration, and when k / K>β, it is defined as the late stage of iteration, where k is the current iteration number, K is the maximum iteration number, and β is a preset threshold (e.g., 0.5~0.7).

[0107] (2) Adaptive boundary based on search state: When the improvement of the global optimal fitness in M ​​consecutive iterations is less than the threshold ε, it is determined to enter the later stage of iteration and the proportion of the group learning factor is increased to accelerate convergence.

[0108] iii. Mutation mechanisms and their triggering conditions

[0109] To avoid particle swarm optimization getting trapped in local optima in complex non-convex optimization problems, this implementation introduces a mutation mechanism into the MPSO algorithm. The triggering conditions for the mutation mechanism can be, but are not limited to, any one or more of the following:

[0110] 1) The improvement in global optimal fitness is less than a preset threshold ε over M consecutive iterations;

[0111] 2) The current iteration number satisfies k / K≥β (e.g., β=0.7), and the mutation probability p is preset. m Trigger;

[0112] 3) Determine diversity based on the variance of particle swarm position or average distance, and trigger mutation when the population diversity is below a set threshold.

[0113] Mutation operations can randomly perturb the position vector of selected particles. For example, mutation operations can be represented by the following formula:

[0114]

[0115] in: σ is the mutated particle position vector; σ is the perturbation amplitude coefficient; η is a random vector uniformly distributed in the interval [-1,1]; ⊙ represents the element-wise multiplication operator; x max x min These are the upper and lower bound vectors for the position variables, respectively.

[0116] After particle mutation, if physical operating constraints are violated (such as photovoltaic inverter capacity limits, energy storage power limits, SOC limits, voltage limits, etc.), boundary truncation or projection methods can be used to map the out-of-bounds variables back into the feasible region, thereby ensuring that the mutated solution still satisfies the engineering constraints.

[0117] By combining the asymmetric time-varying learning factor and mutation mechanism described above, the MPSO algorithm in this embodiment significantly enhances the global optimization capability and the ability to escape local optima while ensuring the convergence speed.

[0118] (4) Hierarchical-partition collaborative optimization control strategy

[0119] In a preferred embodiment, the present invention divides the power grid into several voltage control areas and constructs a three-layer collaborative control architecture of "regional level - feeder level - source level" to achieve voltage and reactive power collaborative optimization control with the participation of distributed power sources.

[0120] i. Partitioning Strategy

[0121] In this embodiment, the power grid can be partitioned according to one or more of the following principles: 1) Based on the topology, nodes within the same feeder or the same distribution area are grouped into the same region; 2) Based on cluster analysis of voltage sensitivity to reactive power, nodes with high electrical coupling are grouped into the same region; 3) Based on measurement and communication conditions, the region where complete operational data can be obtained is designated as a control partition. Each partition corresponds to a feeder-level optimization control unit, which is responsible for the coordinated control of various distributed power sources and reactive power equipment within that partition.

[0122] ii. Source-level (second-level) fast voltage suppression control

[0123] Source-level control primarily operates on locally controllable devices such as photovoltaic inverters, enabling rapid, second-level response to voltage fluctuations. In this embodiment, it can be based on the local voltage measurement value V. i With a given voltage reference value V ref The reactive power droop control or interval segmented control strategy is adopted, as follows:

[0124] 1) When the node voltage V i >V max At the same time, control the photovoltaic inverter to absorb reactive power;

[0125] 2) When the node voltage V i <V min At that time, control the photovoltaic inverter to inject reactive power;

[0126] 3) When V min ≤V i ≤V max At the same time, maintain the current reactive power output or adjust it in small steps.

[0127] The reactive power output Q of the photovoltaic inverter PV The reactive power output constraint of the photovoltaic inverter mentioned above must be met, V max V min These are the upper and lower limits of the voltage, respectively.

[0128] iii. Feeder-level (minute-level) collaborative compensation control

[0129] If voltage over-limit or insufficient reactive power still exists after rapid adjustment at the source level, the feeder-level control unit coordinates the energy storage system, OLTC, capacitor bank, SVG / SVC and other equipment on a minute-level time scale.

[0130] The input signals of the feeder-level control unit include the voltage of each node in the zone, power flow distribution, photovoltaic reactive power margin, and energy storage state of charge (SOC), etc., and the output is the energy storage charging and discharging power setpoint. In addition to other reactive power equipment operation commands, priority should be given to using energy storage to support undervoltage nodes or to alleviate overvoltage problems by appropriately reducing voltage, while keeping the SOC within a safe range.

[0131] iv. Regional (hourly) rolling optimization scheduling

[0132] Regional-level control primarily operates on an hourly timescale, addressing longer-term issues of balancing economic efficiency and security. This implementation uses the aforementioned MPSO algorithm to perform rolling optimization based on given load forecasts, photovoltaic output forecasts, and electricity price curves to obtain a baseline plan {P} for energy storage charging and discharging within future optimization cycles (e.g., several hours). ESS,base (t)}、SOC reference trajectory{SOC ref (t)} and reactive power regulation strategy, providing coordination constraints and regulation margin ΔP(t) for feeder stage and source stage.

[0133] Through the aforementioned partitioned and hierarchical control structure, this implementation method organically combines "rapid voltage stabilization, local coordination, and global economy," enabling various distributed power sources, such as distributed photovoltaics and energy storage, to participate in voltage and reactive power control in a coordinated manner.

[0134] (5) Cross-timescale collaborative control

[0135] In a preferred embodiment, the present invention employs a cross-timescale collaborative control strategy, which organically couples hourly rolling optimization, minute-level local correction, and second-level rapid reactive power suppression to form a top-down hierarchical control closed loop, thereby achieving the comprehensive goal of satisfying both voltage safety and operational economy.

[0136] i. Hourly rolling optimization layer (planning layer)

[0137] The hourly rolling optimization layer uses a relatively long time window (e.g., the next 24 hours) as the scheduling cycle. It uses load forecast sequence, photovoltaic output forecast sequence and time-of-use electricity price information to solve the economic operation benchmark plan of the energy storage system and the corresponding SOC reference curve.

[0138] The output includes the energy storage charge / discharge reference power {P} at each hour or finer time interval. ESS,base (t)} and SOC reference value SOC ref (t), while reserving a certain upper and lower adjustment margin ΔP(t). The actual energy storage power is constrained to P. ESS (t)∈[P ESS,base (t)-ΔP(t),P ESS,base [(t)+ΔP(t)], where P ESS (t) represents the actual charging and discharging power, and ΔP(t) represents the power regulation margin reserved for voltage regulation.

[0139] ii. Minute-level local correction layer (coordination layer)

[0140] Minute-level control mainly focuses on "minimum deviation correction" based on the hourly-level baseline plan. Its goal is to quickly eliminate voltage over-limits while minimizing disruption to the economic baseline.

[0141] When a voltage over-limit or near-over-limit situation is detected within a certain zone, the feeder-level control unit operates within the constraint range [P]. ESS,base (t)-ΔP(t),P ESS,base Within the range [(t)+ΔP(t)], the energy storage power is corrected. The primary objective is to restore the voltage at each node to [V]. min V max Within the interval, the secondary objective is to make P ESS (t) as close to P as possible ESS,base (t), thereby reducing the disturbance to the hourly economic optimization results.

[0142] iii. Second-level rapid reactive power suppression layer (execution layer)

[0143] Second-level control is performed by the photovoltaic inverter and other local reactive power devices, providing a rapid response to short-term voltage fluctuations. The photovoltaic inverter adjusts its reactive power output in real time based on the local voltage deviation, performing second-level control according to the principle of "absorbing reactive power when the voltage is high and injecting reactive power when the voltage is low." When the inverter's reactive power output approaches its capacity limit, it sends a "reactive power saturation" signal to the upper-level control layer, prompting the minute-level control layer to consider calling upon energy storage or other reactive power resources to further support the voltage.

[0144] In summary, this implementation achieves cross-timescale collaborative control through the following mechanisms: the hourly rolling optimization layer provides the economic operating benchmark and reserved adjustment margin for energy storage, taking into account both economy and safety globally; the minute-level local correction layer fine-tunes the energy storage output based on the real-time voltage status within the margin allowed by the benchmark plan, eliminating residual voltage exceedances locally; the second-level rapid reactive power suppression layer directly applies voltage to the system through devices such as photovoltaic inverters, providing the fastest voltage stability support and feeding back saturation information to the upper layer; the three timescales cooperate with each other through a closed-loop link of "benchmark plan - local correction - rapid execution," enabling this invention to maximize new energy consumption and improve the overall economy and operational robustness of the system while ensuring voltage safety in scenarios with a high proportion of distributed photovoltaic access.

[0145] (6) Algorithm Flow

[0146] Step 1: Data preparation and constraint modeling, including: collecting photovoltaic power output forecasts, load forecasts, electricity price curves, and grid parameters; establishing constraints, as shown below:

[0147]

[0148] Q max2 =P·tan(cos -1 ψ)

[0149] Q max =min{Q max1 Q max2}

[0150] -P max ≤P i ≤P max

[0151] SOC min ≤SOC≤SOC max

[0152] Q max1 The maximum reactive power output determined by the rated apparent power capacity constraint of the photovoltaic inverter. pv Q is the rated apparent power capacity of the photovoltaic inverter. P is the active power output of the photovoltaic inverter at the current moment. max2The maximum reactive power output of a photovoltaic inverter, determined by power factor constraints. ψ is the minimum allowable power factor (power factor limit) for the photovoltaic inverter. max It is the maximum available reactive power output of the photovoltaic inverter under the current operating conditions, taken as the smaller of the capacity constraint and the power factor constraint. P i P represents the active power output of the i-th energy storage system at the current moment, where discharge is represented by positive and charging by negative. max It is the maximum allowable charging and discharging power amplitude of the energy storage system.

[0153] The constraints established in Step 1 are consistent with those in "Control Object Modeling" above, both belonging to the physical operational constraint set Ω of the collaborative optimization model. "Control Object Modeling" provides the physical source and mathematical expression of the constraints, while Step 1 organizes the constraints into a unified feasible domain description on the solver side, and is used for feasibility judgment and penalty term construction in fitness calculation.

[0154] Step 2: Objective function setting: Construct a comprehensive objective function to balance voltage stability, grid loss and energy storage benefits.

[0155] Step 3: Particle encoding and initialization: Use PV reactive power, energy storage active power, and the state of the regulating device as particle position vectors; initialize the particle swarm position and velocity, and set the inertial weight and asymmetric learning factor.

[0156] Step 4: Fitness Assessment and Update: Perform power flow calculation and constraint verification. If the voltage exceeds the limit or the SOC is violated, a penalty is added. The improved MPSO is used to update the individual and group optimal formulas.

[0157] Step 5: Convergence and Optimal Solution Output: When the number of iterations reaches the upper limit or the fitness converges, the optimal control strategy is output, including: PV reactive power setpoints for each time period, energy storage charging and discharging power and SOC curves, grid voltage fluctuation suppression effect and economic indicators.

[0158] Step 6: Rolling Optimization and Online Application: Deploy the optimal strategy to the execution end and monitor the effect in real time; based on the latest predictions and operational data, execute steps 1–5 in a rolling manner to achieve dynamic adaptive optimization.

[0159] In summary, the present invention, through a technical solution that improves the MPSO solution by jointly optimizing PV reactive power and energy storage active power, has the following advantages and effects compared to existing technologies that rely solely on PV reactive power or energy storage regulation:

[0160] (1) Significant suppression of voltage over-limit and improvement in voltage quality

[0161] Existing technologies often experience frequent voltage exceedances when photovoltaic penetration is high. Relying solely on independent regulation of PV reactive power or energy storage is insufficient to address overvoltage and undervoltage issues across different operating periods. This invention combines the rapid reactive power regulation characteristics of PV with the flexible active power compensation capabilities of energy storage through joint optimization, comprehensively considering voltage deviations across multiple nodes in the global optimization process. The resulting effects include: in simulations of the IEEE 33-bus system, without optimization, excessive PV output during the day caused five nodes to exceed their voltage limits, while during the evening peak, three nodes experienced voltages below their lower limits. After applying the optimization method of this invention, all node voltages returned to the allowable range (0.95–1.05 pu), voltage fluctuations significantly converged, and the number of exceedances approached zero.

[0162] (2) Photovoltaic power output remains unchanged, and the grid-friendliness of new energy sources is improved.

[0163] Traditional methods often alleviate overvoltage problems in high-voltage penetration scenarios by reducing photovoltaic (PV) active power, leading to a decrease in renewable energy utilization. This invention, within a framework of synergistic optimization of PV reactive power and energy storage active power, explicitly stipulates that PV active power is not reduced, with the adjustment space primarily derived from PV reactive power and energy storage power. The resulting effects include: optimization simulations show that the system can still maintain the voltage within a acceptable range without reducing any PV active power. This maximizes the utilization of PV power generation, improves renewable energy absorption, and avoids economic losses. The absorption capacity of PV power generation is increased by 15%-20%, ensuring the stable operation of the power grid under high renewable energy penetration and providing technical support for the efficient integration of renewable energy.

[0164] (3) Balancing economic benefits and operational safety

[0165] Traditional PV reactive power regulation focuses solely on voltage safety, while energy storage dispatch often prioritizes economic gains (electricity price arbitrage), lacking a synergy between voltage quality and revenue. This invention introduces grid loss and energy storage operation revenue terms into the objective function, achieving a multi-objective optimization balance between "voltage stability, grid loss minimization, and energy storage economic benefits." The resulting effects include: under time-of-use pricing (peak 1.0 yuan / kWh, flat 0.7 yuan / kWh, valley 0.3 yuan / kWh), energy storage achieves both peak shaving and valley filling, as well as voltage regulation, generating a daily revenue of 36,800 yuan, while simultaneously reducing grid losses by approximately 8% and resulting in more stable voltage operation.

[0166] (4) Enhanced global optimization capability and algorithm robustness

[0167] Traditional Probabilistic Search (PSO) is prone to getting trapped in local optima in complex constrained optimization problems, resulting in insufficient convergence speed. The improved PSO proposed in this invention introduces an asymmetric learning factor into the speed update formula and adds a random mutation mechanism in the later stages of convergence. This design ensures both initial global exploration capability and stable convergence in the later stages. The resulting effects include: in the IEEE 33-node multimodal optimization problem, the improved PSO achieves approximately 40% faster convergence speed than the standard PSO, and the fitness of the final solution is improved by 15%, avoiding the premature convergence defect of traditional algorithms.

[0168] (5) Portability and broad prospects for engineering applications

[0169] The method of this invention has been validated in the IEEE 33-node standard system, and the parameter settings and constraints are generalizable. Because the optimization framework is built upon physical constraints and the characteristics of distributed power sources, it has good reusability. The resulting effects include: the method is not only applicable to the IEEE 33-node test system, but can also be extended to provincial power grids, distribution networks, and microgrid scenarios; it can be reused in power systems of different scales; it can seamlessly interface with SCADA / EMS systems; and it has good potential for industrial application.

[0170] By introducing multi-source decomposition modeling theory and differentiated collaborative prediction technology into power system ramp demand prediction, this invention has significant advantages over existing technologies in the following aspects: Through multi-agent collaborative optimization, it successfully eliminates voltage over-limit phenomena caused by equipment limitations or single regulation methods in traditional methods, achieving stable voltage control; Compared with existing technologies, this invention can fully utilize photovoltaic reactive power and energy storage active power without reducing photovoltaic active power (maximizing photovoltaic output), thus improving the level of new energy consumption; This invention simultaneously optimizes voltage stability and energy storage economic benefits through an objective function, providing higher economic returns for the power grid and improving the economic efficiency of the power grid while ensuring safe operation; By improving the particle swarm optimization (MPSO) algorithm, it significantly improves the convergence speed and global optimization capability under multiple constraints and complex operating environments, avoiding the local optimum problem of traditional methods; This invention has good reusability and engineering scalability, can seamlessly connect with existing power grid SCADA / EMS systems, and adapt to the dispatching needs of power systems of different scales. These advantages enable the present invention to not only overcome many shortcomings of the prior art, but also provide reliable technical support for the safe and stable operation of the power system under the high penetration rate of new energy, and provide innovative solutions for the optimized and economical dispatch of smart grids.

[0171] The second embodiment of this application relates to a voltage and reactive power coordinated optimization control system with distributed generation participation. This system includes: a voltage monitoring and data acquisition module, a prediction and state perception module, a distributed generation coordinated optimization module, an optimization solution and control decision module, and an execution and feedback module. The voltage monitoring and data acquisition module is used to acquire grid operation data, which includes at least the real-time voltage of each node, photovoltaic active power output, and the state of charge (SOC) of the energy storage system. The prediction and state perception module is used to make predictions based on the operation data and obtain prediction data, and to obtain the voltage fluctuation trend for future optimization cycles based on the prediction data. The prediction data includes at least the photovoltaic output prediction, load prediction, and time-of-use electricity price curve for future optimization cycles. The distributed generation coordinated optimization module is used to construct a comprehensive objective function with voltage deviation, system grid loss, and energy storage operation revenue as optimization objectives, and to establish a coordinated optimization model that includes photovoltaic inverter reactive power output constraints, energy storage system charging and discharging power constraints, and energy storage system SOC constraints. The optimization and control decision module uses an improved particle swarm optimization algorithm to solve the collaborative optimization model, obtaining the optimal collaborative control strategy that includes the reactive power output setpoints of each photovoltaic inverter, the charging and discharging power of each energy storage system, and the SOC curve. The improved particle swarm optimization algorithm employs an asymmetric learning factor and / or mutation mechanism. The execution and feedback module decomposes the optimal collaborative control strategy into control commands at different time scales and issues them to the corresponding photovoltaic inverters and energy storage systems for execution, thereby achieving coordinated regulation of the grid voltage.

[0172] On the other hand, the present invention also provides an electronic device, which may include: a processor, a communication interface, a memory, and a communication bus, wherein the processor, the communication interface, and the memory communicate with each other through the communication bus. The processor can call logical instructions in the memory to execute a voltage-reactive power coordinated optimization control method involving distributed power sources.

[0173] Furthermore, the logical instructions in the aforementioned memory can be implemented as software functional units and sold or used as independent products, and can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, or the part that contributes to the prior art, or a part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of the present invention. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory, random access memory, magnetic disks, or optical disks.

[0174] On the other hand, the present invention also provides a computer program product, which includes a computer program stored on a non-transitory computer-readable storage medium. The computer program includes program instructions, and when the program instructions are executed by the computer, the computer is able to execute the voltage and reactive power coordinated optimization control method with the participation of distributed power sources provided by the above methods.

[0175] In another aspect, the present invention also provides a non-transitory computer-readable storage medium storing a computer program thereon, which, when executed by a processor, implements the voltage-reactive power coordinated optimization control methods provided above incorporating distributed power sources.

[0176] Through the above description of the embodiments, those skilled in the art can clearly understand that each embodiment can be implemented by means of software plus necessary general-purpose hardware platforms, and of course, it can also be implemented by hardware. Based on this understanding, the above technical solutions, in essence or the part that contributes to the prior art, can be embodied in the form of a software product. This computer software product can be stored in a computer-readable storage medium, such as ROM / RAM, magnetic disk, optical disk, etc., and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute the methods described in the various embodiments or some parts of the embodiments.

[0177] The various technical features disclosed in the above-described invention, the various technical features disclosed in the following embodiments and examples, and the various technical features disclosed in the accompanying drawings can be freely combined to form various new technical solutions (all of which are considered to have been recorded in this specification), unless such a combination of technical features is technically infeasible. For example, in one example, feature A+B+C is disclosed, and in another example, feature A+B+D+E is disclosed. Features C and D are equivalent technical means that serve the same function, and technically only one needs to be used; it is impossible to use both simultaneously. Feature E can be technically combined with feature C. Therefore, the solution A+B+C+D should not be considered as recorded because it is technically infeasible, while the solution A+B+C+E should be considered as recorded.

[0178] All documents mentioned in this application are considered to be incorporated in their entirety into the disclosure of this application so that they can serve as a basis for modifications if necessary. Furthermore, it should be understood that after reading the foregoing disclosure of this application, those skilled in the art can make various alterations or modifications to this application, and these equivalent forms also fall within the scope of protection claimed in this application.

Claims

1. A voltage and reactive power coordinated optimization control method with the participation of distributed power sources, characterized in that, include: The system acquires grid operation data, which includes at least real-time voltage at each node, photovoltaic active power output, and state of charge (SOC) of the energy storage system. Based on the operation data, the system obtains prediction data, which includes at least photovoltaic power output prediction, load prediction, and time-of-use electricity price curves for the future optimization period. Based on the prediction data, the system obtains voltage fluctuation trends for the future optimization period. A comprehensive objective function is constructed with voltage deviation, system network loss and energy storage operation revenue as optimization objectives, and a collaborative optimization model is established that includes reactive power output constraints of photovoltaic inverters, charging and discharging power constraints of energy storage systems and SOC constraints of energy storage systems. An improved particle swarm optimization algorithm is used to solve the cooperative optimization model to obtain the optimal cooperative control strategy, which includes the reactive power output setpoints of each photovoltaic inverter, the charging and discharging power of each energy storage system, and the SOC curve; wherein, the improved particle swarm optimization algorithm adopts an asymmetric learning factor and / or mutation mechanism. The optimal collaborative control strategy is decomposed into control commands at different time scales and sent to the corresponding photovoltaic inverters and energy storage systems for execution, so as to achieve collaborative regulation of grid voltage.

2. The voltage and reactive power coordinated optimization control method as described in claim 1, characterized in that, The comprehensive objective function is expressed as the following formula: Where, ΔV t P represents the voltage deviation at each node at time t. loss,t R represents the system network loss at time t. ESS,t Let t represent the energy storage operation revenue at time t, and α, β, and γ represent the weighting coefficients of the voltage deviation, system network loss, and energy storage operation revenue, respectively.

3. The voltage and reactive power coordinated optimization control method as described in claim 1, characterized in that: The reactive power output constraint of the photovoltaic inverter is expressed as follows: Where S PV P is the rated apparent power of the photovoltaic inverter. PV The active power output of the photovoltaic inverter; The charging and discharging power constraint of the energy storage system is expressed as follows: in These represent the charging power and discharging power of the energy storage system, respectively. The SOC constraint of the energy storage system is expressed as SOC min ≤SOC t ≤SOC max SOC t The State of Charge (SOC) of the energy storage system at time t represents the state of charge of the system. min The state of charge (SOC) of the energy storage system represents the minimum state of charge. max This indicates the maximum state of charge of the energy storage system.

4. The voltage and reactive power coordinated optimization control method as described in claim 1, characterized in that, The step of solving the cooperative optimization model using an improved particle swarm optimization algorithm further includes: The decision variables of the collaborative optimization model are mapped to the position vectors of particles in the particle swarm; The positions and velocities of the particle swarm are randomly initialized within the feasible region defined by the physical operation constraints. The parameters of the improved particle swarm optimization algorithm are set, including the asymmetric learning factor. In each iteration, the fitness corresponding to the position vector of each particle is calculated. The fitness is calculated based on the comprehensive objective function of the collaborative optimization model and includes penalties for violating constraints. The individual historical best position of each particle and the collective historical best position of the particle swarm are updated according to the fitness. Based on the asymmetric learning factor, the individual historical best position, and the global historical best position, the position of the particle in the next iteration is calculated according to the particle swarm's velocity update formula and position update formula. When the maximum number of iterations or fitness convergence condition is met, the iteration is terminated and the final group historical best position is decoded as the optimal cooperative control strategy.

5. The voltage and reactive power coordinated optimization control method as described in claim 4, characterized in that: The speed update formula is expressed as follows: The location update formula is expressed as follows: in Let be the velocity vectors of the i-th particle in the k-th iteration. Let be the position vector of the i-th particle in the k-th iteration, ω be the inertia weight, c1(k) be the individual learning factor in the k-th iteration, c2(k) be the group learning factor in the k-th iteration, and r1 and r2 be random numbers uniformly distributed in the interval [0,1]. Let g be the vector of the best historical positions of the i-th particle before the k-th iteration. k The vector representing the globally optimal position of the particle swarm before the k-th iteration. In the early stages of iteration, the value of the individual learning factor c1(k) is greater than the value of the group learning factor c2(k); in the later stages of iteration, the value of the group learning factor c2(k) is greater than or equal to the value of the individual learning factor c1(k).

6. The voltage and reactive power coordinated optimization control method as described in claim 4, characterized in that, The parameters include mutation triggering conditions, and the step of solving the cooperative optimization model using an improved particle swarm optimization algorithm further includes: After calculating the individual's historical best position and the group's historical best position, it is determined whether the preset mutation triggering condition is met. If the mutation triggering condition is met, a random mutation operation is performed on the position vector of the selected particle according to the preset probability, and the particle that exceeds the physical operation constraint after mutation is constrained and repaired, and projected back into the feasible domain.

7. The voltage and reactive power coordinated optimization control method as described in claim 6, characterized in that, The mutation triggering conditions include one or more of the following groups: The improvement in fitness of the global optimum is less than a preset threshold in M ​​consecutive iterations; The current iteration number satisfies k / K≥β, where K is the maximum iteration number and β is a preset coefficient; and Diversity is determined based on the variance of particle swarm positions or the average distance. If the diversity of the particle swarm is lower than a set threshold, the diversity is determined.

8. The voltage and reactive power coordinated optimization control method as described in claim 1, characterized in that, The step of decomposing the optimal cooperative control strategy into control instructions at different time scales further includes: Based on the local voltage measurement value of the photovoltaic inverter, the photovoltaic inverter is controlled to perform rapid reactive power adjustment in seconds according to the preset reactive power droop control or interval segmentation control strategy. If the voltage still exceeds the limit after the second-level rapid adjustment of reactive power, the energy storage system is controlled to perform active power compensation adjustment on a minute-level time scale. Based on the predicted data, the energy storage charging and discharging reference power and SOC reference trajectory of the energy storage system within the future prediction period are solved on an hourly basis, and a power adjustment margin is reserved for the minute-level coordinated control.

9. An electronic device, comprising: A memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that the processor, when executing the computer program, implements the method according to any one of claims 1 to 8.

10. A non-transitory computer-readable storage medium, characterized in that, The non-transitory computer-readable storage medium includes a computer program that, when executed by a processor, implements the steps of the method as described in any one of claims 1 to 8.

Images