A load power real-time control method considering voltage out-of-limit

Abstract

This invention provides a real-time load power control method considering voltage exceedance, comprising the following steps: S1: Determine the target value of the feeder voltage; S2: Establish photovoltaic reactive power-voltage state space models, wind turbine reactive power-voltage state space models, and real-time voltage regulator reactive power-voltage state space models, and combine these with the system configuration to establish a system reactive power-voltage state space model, and combine this with voltage sensitivity to obtain a feeder voltage prediction model; S3: Construct and solve a rolling optimization model for feeder load power control; S4: Execute the optimization results and perform feeder load power feedback correction; S5: Judge the load power control result, exit the optimization or repeat step S2. The method provided in this invention constructs a system feeder voltage prediction model and achieves real-time correction of the feeder voltage through rolling optimization, eliminating voltage exceedance caused by renewable energy power fluctuations while ensuring the feeder load power control effect.

Description

Technical Field

[0001] This invention belongs to the field of load control, and particularly relates to a real-time load power control method that takes into account voltage over-limit. Background Technology

[0002] With the continuous improvement of the intelligence level of distribution networks and the ongoing development of the electricity market, the load-side regulation potential of distribution networks is becoming increasingly apparent. Exploring load-side regulation capabilities to participate in the consumption of new energy sources aligns with the integrated source-grid-load-storage characteristics of my country's new power system. Feeder load power control technology, also known internationally as Conservation Voltage Reduction (CVR), is a load control technology based on the coupling characteristics of load voltage and power. Through voltage control, it regulates load power and is widely used in long-term applications such as system energy conservation and peak shaving.

[0003] To enhance load power regulation, feeder load power control technology often optimizes node voltages to the system's permissible limits, resulting in greater load power regulation. However, real-time fluctuations in renewable energy power can cause voltage over-limit issues, affecting feeder load power control performance and user power quality.

[0004] To address the power fluctuation problem caused by the integration of renewable energy sources, existing research typically focuses on control timescales on the hourly / minute level. However, renewable energy output is unstable and fluctuates significantly. Current time-scale control strategies are ill-suited for scenarios with rapid fluctuations in renewable energy output. Therefore, there is an urgent need to develop real-time power control strategies for systems with rapidly changing renewable energy output to ensure the safe and reliable operation of the distribution network.

[0005] Chinese invention patent CN110601252A discloses a fast voltage control method for feeder-level distribution networks with distributed photovoltaic (PV) power generation based on MPC (Multi-Process Control). The method includes a series-type voltage regulator (DVR), a PV inverter, and a DPMU (DPentiometer Measurement Unit) installed at the feeder head and some PV access nodes. Based on the fast feedback information from the DPMU, it achieves second-level voltage control by coordinating multiple DVRs and distributed PV inverters over a wide area. Specifically, it first estimates the state of the distribution network with distributed PV and DVR based on DPMU measurement information, then performs online extraction of the optimal control gradient for the feeder voltage, online updating of the feeder voltage tracking reference value, and finally, feeder-level voltage closed-loop control based on MPC. However, this patent does not consider the voltage limit exceeding problem caused by real-time fluctuations in renewable energy power.

[0006] Chinese utility model patent CN104753066B discloses a voltage control method for active distribution network feeders based on parameter fitting, including the following steps: 1. The main control unit acquires the voltage status information of each feeder; 2. When the minimum voltage on the feeder is lower than the lower voltage limit, reactive power regulation or load shedding voltage control based on parameter fitting is performed; 3. When the maximum voltage on the feeder is higher than the upper voltage limit, reactive power regulation or active power regulation based on parameter fitting is performed. This invention can avoid voltage exceeding the limit problem, but the optimization method uses parameter fitting, which cannot achieve real-time dynamic adjustment. Summary of the Invention

[0007] To address the shortcomings of existing technologies, the purpose of this invention is to provide a real-time load power control method that considers voltage over-limit. This method achieves real-time and precise control of feeder voltage while ensuring that node voltage does not exceed limits, thereby enhancing the power regulation effect of the distribution network.

[0008] To achieve the above objectives, the present invention adopts the following technical solution:

[0009] A real-time load power control method considering voltage over-limit, the control method includes the following steps:

[0010] S1: Obtain the voltage regulation amount by adjusting the feeder load power, and determine the target value of the feeder voltage;

[0011] S2: Establish photovoltaic reactive power-voltage state space model, wind turbine reactive power-voltage state space model, and real-time voltage regulator reactive power-voltage state space model. Combine the system composition to establish the system reactive power-voltage state space model. Combine the voltage sensitivity to obtain the feeder voltage prediction model.

[0012] S3: Construct and solve the feeder load power control rolling optimization model;

[0013] S4: Execute the optimization results and perform feeder load power feedback correction;

[0014] S5: Determine the load power control result. If the load has been controlled, end the adjustment. If the load has not been controlled, repeat step S2.

[0015] Furthermore, in step S1, the feeder load power control is achieved by adjusting the feeder voltage to control the feeder load power, and the feeder load power needs to be adjusted by ΔP. i At that time, the feeder voltage regulation is ΔP i / K i K i Let V0 be the voltage-power coupling characteristic coefficient of the feeder load, then the target value of the feeder voltage is V0 = V - ΔP. i / K i V0 is the target voltage of the feeder.

[0016] Furthermore, in step S2, the photovoltaic reactive power-voltage state space model includes a photovoltaic reactive power-voltage state space model of PQ control type and a photovoltaic reactive power-voltage state space model of PV control type, and the wind turbine reactive power-voltage state space model includes a wind turbine reactive power-voltage state space model of PQ control type and a wind turbine reactive power-voltage state space model of PV control type.

[0017] Furthermore, the PQ-controlled photovoltaic reactive-voltage state-space model is expressed by the following formula:

[0018] in, ΔQ PV Let ΔQ be the change in photovoltaic reactive power. PV,ref Let ΔQ be the change in the target reactive power of the photovoltaic system. PV,ref =Q PV,ref -Q PV Q PV,ref With Q PV These are the target reactive power and the initial reactive power of the photovoltaic system, respectively. The time constant for photovoltaic reactive power control;

[0019] The photovoltaic reactive power-voltage state-space model for PV control type is expressed by the following formula:

[0020] in, ΔV PV,ref ΔV represents the photovoltaic target voltage deviation. PV,ref =V PV,ref -V PV V PV,ref With V PV These represent the photovoltaic target voltage and initial voltage, respectively; K PV This is the voltage-reactive power control droop coefficient for PV type photovoltaic systems.

[0021] Furthermore, the reactive power-voltage state-space model for PQ control type wind turbines is as follows:

[0022] in, The time constant for reactive power control of the wind turbine;

[0023] The reactive power-voltage state-space model for PV-controlled wind turbines is as follows:

[0024] in, K WT This refers to the voltage-reactive power control droop coefficient for PV type wind turbines.

[0025] Furthermore, the reactive power-voltage state-space model of the real-time voltage regulator is as follows:

[0026] Among them, A VT =-1 / T VT B VT =1 / T VT ΔV out ΔV represents the change in output voltage of the voltage regulator. VT,ref T represents the change in the target voltage of the voltage regulator. VT This is the voltage control time constant of the voltage regulator.

[0027] Furthermore, in step S2, establishing the system reactive power-voltage state space model based on the system configuration specifically involves: establishing the system reactive power-voltage state space model based on the number of PV-controlled photovoltaics and PQ-controlled photovoltaics, and the number of PV-controlled wind turbines and PQ-controlled wind turbines within the system. For a system with one real-time voltage regulator and N... PQPV Taiwan PQ type photovoltaic, N PVPV Taiwan PV type photovoltaic, N PQWT PQ type wind turbine and N PVWT The reactive power-voltage state-space model of the power distribution network for a PV type wind turbine is as follows:

[0028]

[0029]

[0030]

[0031]

[0032]

[0033] Furthermore, step S2 also includes applying the formulas in the system reactive-voltage state-space model. Discretize into x(k+1)=A d x(k)+B d u(k), the system reactive-voltage state-space model combined with voltage sensitivity yields the system feeder voltage prediction model:

[0034] In the formula, V(k) is the system node voltage at step k, V(0) is the initial system voltage; ΔV0(k) is the change in the output voltage of the real-time voltage regulator; ΔP(k) and ΔQ(k) are the changes in the net active power and net reactive power of the node, respectively.

[0035] Furthermore, based on the system feeder voltage prediction model obtained in step S2, a rolling optimization model for feeder load power control is constructed: Where V(k) is the feeder voltage at time k; V0 is the target value of the feeder voltage within the corresponding time period; N p Let N be the set of sampling steps within the prediction period. p =[1,…,N p ] T N p N is the number of sampling steps in the prediction period. p =T p / ΔT p T p For the prediction period, ΔT p Using the sampling period as the basis, the rolling optimization model is solved to obtain the voltage at each node, the action of the reactive power regulating equipment, and the power regulation of the feeder load within the control period.

[0036] Furthermore, the rolling optimization model includes the following constraints: voltage limit exceedance constraint, reactive power voltage regulation equipment control quantity constraint, reactive power voltage regulation equipment control quantity remains unchanged within the same control cycle, and voltage sensitivity remains unchanged within the prediction cycle.

[0037] Compared with the prior art, the real-time load power control method considering voltage over-limit provided by the present invention has the following beneficial effects:

[0038] With the increasing severity of global energy, environmental, and climate issues, accelerating the development and utilization of renewable energy to promote economic development has become a consensus across society. Among existing renewable energy development technologies, photovoltaic and wind power are becoming increasingly mature and widely used. In recent years, new energy power generation has developed rapidly, with wind and solar power surpassing nuclear power to become the third largest power source after coal and hydropower. As new energy power generation continues to be integrated into the grid, it is necessary to explore load-side regulation capabilities to participate in the consumption of new energy.

[0039] In existing technologies, feeder load power control technology is generally used to regulate load power through voltage control, and it is widely used in long-term applications such as system energy saving and peak shaving. To enhance the load power regulation effect, the node voltage is often optimized to the system's allowable boundary, such as 0.93 and 1.07 as specified in national standards. However, the power flow changes caused by real-time fluctuations in renewable energy power can easily lead to voltage limit violations. For example, high renewable energy output may cause power flow backflow, leading to voltage rise or even exceeding the upper voltage limit; while a sudden drop in renewable energy output may cause the voltage at the feeder end to exceed the lower limit. Voltage limit violations are not only a power quality issue, but also affect the reliability of the distribution network and the utilization rate of renewable energy. If not controlled, they may lead to serious problems such as renewable energy disconnection from the grid, affecting the feeder load power control effect and the power quality for users.

[0040] Many power system professionals have studied this problem. However, existing research usually implements feeder load power control technology on a longer time scale, which is difficult to meet the requirements of new power systems for the speed of load-side power regulation. On the other hand, although some studies have considered feeder load power control technology on a real-time time scale and improved the feeder load power control effect by optimizing the voltage to the boundary, they have ignored the fact that at this time, the fluctuation of new energy power is very likely to cause voltage over-limit problems, and the distribution network has the risk of voltage over-limit.

[0041] The real-time load power control method considering voltage over-limit provided by this invention is a load control technology based on the load voltage-power coupling characteristics. By controlling the voltage, the load power is adjusted, the voltage adjustment amount is determined according to the power load adjustment amount, and the target voltage is obtained.

[0042] This invention analyzes the reactive power regulation characteristics of different types of photovoltaic (PV) systems, wind turbines, and real-time voltage regulators (RTDs), establishing reactive power-voltage state-space models for various types of PV systems, wind turbines, and RTDs. Considering the structure of the power grid system, where RTDs are typically installed at the feeder head as relaxation nodes to flexibly adjust system voltage, while PV and wind turbines are distributed across different nodes, controlling system voltage through reactive power injection, and taking into account the characteristic that there is usually one RTD and multiple PV and wind turbines, a system reactive power-voltage state-space model applicable to most power networks is established. After discretization and incorporating voltage sensitivity, a feeder voltage prediction model is obtained. Feeder voltage is related to the initial voltage, the change in net active power at the wind turbine / PV node, the change in output voltage of the RTD, and the change in net reactive power at the wind turbine / PV node. The feeder voltage prediction model constructed in this invention has a wide applicability, applicable to most power systems, and the relevant indicators within the model are clear, specific, and calculable.

[0043] This invention constructs a rolling optimization model, specifically a rolling optimization model based on predictive control. The optimization of predictive control is not performed offline once, but repeatedly online as the sampling time progresses. At each sampling time, the change in control action is recalculated based on the current prediction error. The rolling calculation is continuously performed, and the deviation at each sampling time is repeatedly optimized. This allows for timely correction of various complex situations that occur during the control process.

[0044] Rolling optimization models utilize predictive models and historical system data, along with future inputs, to predict the future output of the system. Feedback correction control is obtained by optimizing a specific performance index within a rolling, finite time interval. Unlike traditional global optimization, rolling optimization optimizes the performance index only for a finite future time from that moment onwards, while simultaneously advancing this optimization time to the next moment, continuously performing online optimization. At each moment, a set of future control actions is obtained, but only the control action for that moment is implemented. A new set of control actions is predicted and optimized for the next moment, also implementing only a new control action; each step involves feedback correction. Predictive control gains foresight, and rolling optimization and feedback correction can better adapt to actual systems, exhibiting stronger robustness. This method enables closed-loop control, allowing operators to adjust the operating points of the system's reactive power regulating equipment according to the results of the rolling optimization model.

[0045] In this invention, the rolling optimization model includes the following constraints: voltage limit exceedance constraint, reactive power voltage regulation equipment control quantity constraint, and the control quantity of reactive power voltage regulation equipment remaining constant within the same control cycle, while voltage sensitivity remains constant within the prediction cycle. This can prevent voltage limit exceedance and make the regulation process compatible with the reactive power regulation equipment within the power system.

[0046] In summary, the real-time load power control method considering voltage over-limit provided by this invention utilizes real-time voltage regulators, photovoltaics, and wind turbines as reactive power voltage regulation devices, establishes a feeder voltage prediction model, and continuously optimizes the feeder voltage. This solves the problems of load power control accuracy and voltage over-limit caused by renewable energy power fluctuations during feeder load power control, and enhances the power regulation effect of the distribution network. Attached Figure Description

[0047] The present invention will now be described in further detail with reference to the accompanying drawings.

[0048] Figure 1 The flowchart shows the real-time load power control method considering voltage over-limit provided by the present invention. Detailed Implementation

[0049] The present invention will be further described below with reference to embodiments. Obviously, the described embodiments are only some, not all, of the embodiments of the present invention. All other embodiments obtained by those skilled in the art based on the embodiments of the present invention without creative effort are within the scope of protection of the present invention.

[0050] Please refer to Figure 1 , Figure 1 This is a flowchart illustrating the real-time load power control method considering voltage over-limit provided by the present invention.

[0051] The real-time load power control method considering voltage over-limit provided by this invention includes the following steps:

[0052] S1: Obtain the voltage regulation amount by adjusting the feeder load power, and determine the target value of the feeder voltage;

[0053] S2: Establish photovoltaic reactive power-voltage state space model, wind turbine reactive power-voltage state space model, and real-time voltage regulator reactive power-voltage state space model. Combine the system composition to establish the system reactive power-voltage state space model. Combine the voltage sensitivity to obtain the feeder voltage prediction model.

[0054] S3: Construct and solve the feeder load power control rolling optimization model;

[0055] S4: Execute the optimization results and perform feeder load power feedback correction;

[0056] S5: Determine the load power control result. If the load has been controlled, end the adjustment. If the load has not been controlled, repeat step S2.

[0057] In some preferred embodiments, in step S1, the feeder load power control is achieved by adjusting the feeder voltage to control the feeder load power, and the feeder load power needs to be adjusted by ΔP. i At that time, the feeder voltage regulation is ΔP i / K i K i Let V0 be the voltage-power coupling characteristic coefficient of the feeder load, then the target value of the feeder voltage is V0 = V - ΔP. i / K i V0 is the target voltage of the feeder.

[0058] In step S2, the photovoltaic reactive power-voltage state space model includes a photovoltaic reactive power-voltage state space model of PQ control type and a photovoltaic reactive power-voltage state space model of PV control type, and the wind turbine reactive power-voltage state space model includes a wind turbine reactive power-voltage state space model of PQ control type and a wind turbine reactive power-voltage state space model of PV control type.

[0059] For the PQ control type photovoltaic reactive-voltage state-space model, the photovoltaic inverter adopts PQ decoupling control. Since the millisecond-level transient adjustment process inside the photovoltaic controller is not considered, the photovoltaic regulation model of the PQ control type is approximately represented by a first-order inertial element as shown in equation (1).

[0060]

[0061] In the formula, ΔQ PV ΔQ represents the change in photovoltaic reactive power. PV,ref ΔQ represents the change in the target reactive power of the photovoltaic system.PV,ref =Q PV,ref -Q PV Q PV,ref With Q PV These are the target reactive power and the initial reactive power of the photovoltaic system, respectively. This is the photovoltaic reactive power control time constant.

[0062] Accordingly, the state-space model of photovoltaic reactive power control under PQ control is shown in equation (2).

[0063]

[0064] In the formula,

[0065] For the photovoltaic reactive power-voltage state-space model of PV control type, the photovoltaic of PV control type takes voltage as the control target. The photovoltaic uses a local droop controller to calculate the equivalent target reactive power based on the deviation between the target voltage and the actual voltage at the grid connection point, as shown in Equation (3), and then completes the target reactive power through reactive power control.

[0066] ΔQ PV,ref =K PV ΔV PV,ref (3)

[0067] In the formula, ΔV PV,ref ΔV represents the photovoltaic target voltage deviation. PV,ref =V PV,ref -V PV V PV,ref With V PV These represent the photovoltaic target voltage and initial voltage, respectively; K PV This is the voltage-reactive power control droop coefficient for PV type photovoltaic systems.

[0068] Accordingly, the state-space model of photovoltaic reactive power control of the PV control type is shown in equation (4).

[0069]

[0070] In the formula,

[0071] For the reactive power-voltage state-space model of wind turbines, the VSC control logic of wind turbines and photovoltaics is similar, but the power control time constant of wind turbines is generally much larger than that of photovoltaics. Accordingly, the state-space model for reactive power control of PQ-type wind turbines is as follows:

[0072]

[0073] In the formula, This is the time constant for reactive power control of the wind turbine.

[0074] The state-space model for reactive power control of PV-controlled wind turbines is as follows:

[0075]

[0076] In the formula, K WT This refers to the voltage-reactive power control droop coefficient for PV type wind turbines.

[0077] This paper addresses the reactive power-voltage state-space model of a real-time voltage regulator. Real-time voltage regulators include intelligent transformers and dynamic voltage restorers, and generally employ VSC (Voltage Control System) for voltage control. Taking a dynamic voltage restorer as an example, it typically uses a dual-loop control mode of voltage and current. After obtaining the target compensation voltage value, the compensation voltage is tracked in real time through dual-loop voltage and current control. VSC control is then completed via PWM modulation to output the corresponding compensation voltage, thus achieving voltage control. For a real-time voltage regulator, its voltage regulation external characteristics can be represented by a first-order inertial element.

[0078] ΔV out =ΔV VT,ref / (1+sT VT (7)

[0079] In the formula, ΔV out ΔV represents the change in output voltage of the voltage regulator. VT,ref T represents the change in the target voltage of the voltage regulator. VT This is the voltage control time constant of the voltage regulator.

[0080] Accordingly, the state-space model of the real-time voltage regulator is:

[0081]

[0082] In the formula, A VT =-1 / T VT B VT =1 / T VT .

[0083] In step S2, the reactive power-voltage state-space model of the system is established by considering the number of PV-controlled and PQ-controlled photovoltaic (PV) generators, as well as the number of PV-controlled and PQ-controlled wind turbines. The system typically contains a small number of real-time voltage regulators and a large number of photovoltaic and wind power generators. Real-time voltage regulators are generally installed at the feeder head, acting as relaxation nodes to flexibly adjust the system voltage. The photovoltaic and wind turbine generators are distributed across different nodes, and the system voltage is controlled by injecting reactive power.

[0084] For a device with 1 voltage regulator, N PQPVTaiwan PQ type photovoltaic, N PVPV Taiwan PV type photovoltaic, N PQWT PQ type wind turbine and N PVWT The distribution network of a PV type wind turbine, including the system state-space model of all control variables, is as follows:

[0085]

[0086]

[0087]

[0088]

[0089]

[0090] Since the above system state-space model is a continuous model, but discrete control is used in actual engineering, for ease of control, formula (9) is expressed in terms of ΔT. P To discretize the step size, we obtain formula (14).

[0091] x(k+1)=A d x(k)+B d u(k) (14) The reactive-voltage state-space model of the system is combined with the voltage sensitivity to obtain the system feeder voltage prediction model:

[0092] In the formula, V(k) is the system node voltage at step k, V(0) is the initial system voltage; ΔV0(k) is the change in relaxed node voltage (i.e., the change in output voltage of the real-time voltage regulator); ΔP(k) and ΔQ(k) are the changes in net active power and net reactive power of the node, respectively.

[0093] In step S3, based on the system feeder voltage prediction model obtained in step S2, a feeder load power control rolling optimization model is constructed as shown in formula (16). On the one hand, the objective function makes the feeder load power approach the target value, ensuring the feeder load power control effect; on the other hand, by limiting the node voltage deviation, the voltage over-limit caused by the fluctuation of new energy power is reduced.

[0094]

[0095] In the formula, V(k) is the feeder voltage at time k; V0 is the target value of the feeder voltage within the corresponding time period; N p Let N be the set of sampling steps within the prediction period. p =[1,…,N p ] T N p N is the number of sampling steps in the prediction period. p=T p / ΔT p T p For the prediction period, ΔT p The sampling period.

[0096] The rolling optimization model uses real-time measurement data as input to perform online correction of the feeder voltage. Based on the target feeder voltage value obtained in step 1, the rolling optimization model considers real-time power fluctuations from renewable energy sources and performs rolling optimization on various reactive power regulation devices. Decision variables include the grid connection point voltage or reactive power of the renewable energy inverter and the output voltage of the real-time voltage regulator. The optimization objective is to minimize the deviation between the system feeder voltage and the target voltage. Solving the rolling optimization model yields the voltage at each node, the action of the reactive power regulation devices, and the power regulation of the feeder load within the control cycle.

[0097] Furthermore, the rolling optimization model includes the following constraints: voltage limit exceedance constraint, reactive power voltage regulation equipment control quantity constraint, reactive power voltage regulation equipment control quantity remains unchanged within the same control cycle, and voltage sensitivity remains unchanged within the prediction cycle.

[0098] Voltage over-limit constraints for the rolling optimization model. To avoid voltage over-limit, the node voltage should be within the allowable range, as shown in equations (17) and (18).

[0099]

[0100]

[0101] In the formula, V i (k) represents the voltage of feeder i at time k, V. max With V min These represent the maximum and minimum values ​​of the feeder voltage, respectively. N is the set of system nodes.

[0102] Operating constraints of reactive power regulating equipment in the rolling optimization model. The reactive power regulating equipment in this invention includes PQ-controlled photovoltaic and PV-controlled photovoltaic systems, PQ-controlled wind turbines and PV-controlled wind turbines, and real-time voltage regulators. The control quantities of the reactive power regulating equipment should be within the allowable range, including the output voltage of the voltage regulator, the reactive power of the photovoltaic and wind turbine inverters, the grid connection point voltage, etc., as shown in equations (19)-(23).

[0103]

[0104]

[0105]

[0106]

[0107]

[0108] In the formula, V VT,ref (k) represents the target output voltage of the voltage regulator at time k; Let i be the target reactive power of the i-th PQ type photovoltaic unit at time k; Let be the target voltage of the i-th PV-type photovoltaic unit at time k; Let be the target reactive power of the i-th PQ type wind turbine at time k; Let N be the target voltage of the i-th PV type wind turbine at time k. VT N represents the set of voltage regulator nodes. PQPV With N PVPV These are sets of photovoltaic nodes of PQ type and PV type, respectively; N PQWT With N PVWT These are sets of wind turbine nodes of PQ type and PV type, respectively.

[0109] The control quantity of the reactive voltage regulator only changes when the control cycle is switched, and remains unchanged within the same control cycle, as shown in equations (24)-(28).

[0110]

[0111]

[0112]

[0113]

[0114]

[0115] In the formula, N c To control the set of sampling steps within a period, N c =[1,…,N c ] T N c =T c / ΔT p T c The control cycle is denoted by m, where m represents the m-th control cycle.

[0116] Voltage sensitivity constraints for the rolling optimization model. The voltage sensitivity remains constant throughout the prediction period, as shown in equations (29)-(58).

[0117]

[0118]

[0119]

[0120] In step S4, the optimization results are executed, and feeder load power feedback correction is performed: the optimization model obtains the reactive power regulating equipment's operating parameters for multiple control cycles, but only the optimization results within the latest control cycle are executed each time. After a single control cycle ends, feedback correction is performed, and the model parameters are adjusted according to the actual operating state of the system. Then, a new model optimization solution and control command issuance are performed again to achieve closed-loop control. Operators adjust the operating point of the system's reactive power regulating equipment according to the rolling optimization model results.

[0121] In step S5, the load power control result is determined. If the load power has been controlled, the adjustment ends. If the load power control has not been completed, step S2 is repeated.

[0122] The present invention provides a real-time load power control method that considers voltage limit exceedance. In distribution networks with a high proportion of renewable energy, node voltage may exceed limits due to renewable energy power disturbances, affecting real-time power control using voltage control. The method constructs a system feeder voltage prediction model, which fully considers the impact of rapid renewable energy fluctuations on feeder load power control from a real-time operation perspective. Through rolling optimization, it achieves real-time correction of feeder voltage, eliminating voltage limit exceedances caused by renewable energy power fluctuations while ensuring the effectiveness of feeder load power control.

[0123] Therefore, those skilled in the art should recognize that although numerous exemplary embodiments of the present invention have been shown and described in detail herein, many other variations or modifications conforming to the principles of the present invention can be directly determined or derived from the disclosure of the present invention without departing from the spirit and scope of the invention. Thus, the scope of the present invention should be understood and construed as covering all such other variations or modifications.

Claims

1. A load power real-time control method considering voltage excursion, characterized in that, The control method includes the following steps: S1: Obtain the voltage regulation amount by adjusting the feeder load power, and determine the target value of the feeder voltage; S2: Establish photovoltaic reactive power-voltage state space model, wind turbine reactive power-voltage state space model, and real-time voltage regulator reactive power-voltage state space model. Combine the system composition to establish the system reactive power-voltage state space model. Combine the voltage sensitivity to obtain the feeder voltage prediction model. S3: Construct and solve the feeder load power control rolling optimization model; S4: Execute the optimization results and perform feeder load power feedback correction; S5: Determine the load power control result. If the load has been controlled, end the adjustment. If the load has not been controlled, repeat step S2.

2. The load power real-time control method considering voltage excursion according to claim 1, characterized in that, In step S1, the feeder load power control is achieved by adjusting the feeder voltage to control the feeder load power. The feeder load power needs to be adjusted by adjusting ΔP. i At that time, the feeder voltage regulation is ΔP i / K i， Where K i Let V0 be the voltage-power coupling characteristic coefficient of the feeder load, then the target value of the feeder voltage is V0 = V - ΔP. i / K i， V0 is the target voltage for the feeder.

3. The load power real-time control method considering voltage excursion according to claim 1, characterized in that, In step S2, the photovoltaic reactive power-voltage state space model includes a photovoltaic reactive power-voltage state space model of PQ control type and a photovoltaic reactive power-voltage state space model of PV control type. The wind turbine reactive power-voltage state space model includes a wind turbine reactive power-voltage state space model of PQ control type and a wind turbine reactive power-voltage state space model of PV control type.

4. The load power real-time control method considering voltage excursion according to claim 3, characterized in that, The state-space model of photovoltaic reactive power-voltage control type is expressed by the following formula: wherein, ; , AQ PV is the photovoltaic reactive power variation, AQ PV,ref is the photovoltaic target reactive power variation, AQ PV,ref = Q PV,ref - Q PV , Q PV,ref and Q PV are the photovoltaic target reactive power and the initial reactive power, respectively; is the photovoltaic reactive power control time constant; The photovoltaic reactive power-voltage state-space model for PV control type is expressed by the following formula: wherein, ; , AV PV,ref is the photovoltaic target voltage deviation, AV PV,ref = V PV,ref - V PV , V PV,ref and V PV are the photovoltaic target voltage and the initial voltage, respectively; is the PV-type photovoltaic voltage-reactive power control droop coefficient.

5. The load power real-time control method considering voltage excursion according to claim 3, characterized in that, The reactive power-voltage state-space model for PQ control type wind turbines is: wherein, ; ; Kf is a fan reactive power control time constant; AQ WT Kf is a fan reactive power control time constant; AQ WT,ref Kf is a fan reactive power control time constant; AQ The reactive power-voltage state-space model for PV-controlled wind turbines is as follows: wherein, ; ; is the PV type fan voltage-reactive power control droop coefficient; AV WT,ref is the fan target voltage deviation, AV WT,ref = V WV,ref - V WT , V WT,ref and V WT are the fan target voltage and initial voltage, respectively.

6. A real-time load power control method considering voltage over-limit as described in claim 3, characterized in that, The reactive power-voltage state-space model of the real-time voltage regulator is as follows: ,in, ; ΔV out ΔV represents the change in output voltage of the voltage regulator. VT,ref This refers to the change in the target voltage of the voltage regulator. This is the voltage control time constant of the voltage regulator.

7. A real-time load power control method considering voltage over-limit as described in claim 6, characterized in that, In step S2, establishing the system reactive power-voltage state space model based on the system configuration specifically involves: establishing the system reactive power-voltage state space model based on the number of PV-controlled photovoltaics and PQ-controlled photovoltaics, and the number of PV-controlled wind turbines and PQ-controlled wind turbines within the system. For a system with one real-time voltage regulator and N... PQPV Taiwan PQ type photovoltaic, N PVPV Taiwan PV type photovoltaic, N PQWT PQ type wind turbine and N PVWT The reactive power-voltage state-space model of the power distribution network for a PV type wind turbine is as follows: ;in, This represents the change in reactive power of the first photovoltaic unit. For the Nth PQPV The change in reactive power of a PQ type photovoltaic system; For the Nth PVPV The change in reactive power of a PV-type photovoltaic system; This represents the change in reactive power of the first wind turbine. For the Nth PQWT The change in reactive power of a PQ type wind turbine; For the Nth PVWT The change in reactive power of a PV type wind turbine; The deviation between the target reactive power and the initial reactive power of the first photovoltaic unit; For the Nth PQPV The deviation between the target reactive power and the initial reactive power of a PQ type photovoltaic system; For the Nth PVPV The deviation between the target reactive power and the initial reactive power of a PV-type photovoltaic system; The deviation between the target reactive power and the initial reactive power of the first wind turbine; For the Nth PQWT The deviation between the target reactive power and the initial reactive power of a PQ type wind turbine; For N PVWT The deviation between the target reactive power and the initial reactive power of a PV type wind turbine; , , The reactive power control time constant of the voltage regulator; , , The time constant for reactive power control of the first photovoltaic unit. The voltage-reactive power control droop coefficient for the first photovoltaic unit; , , For the Nth PQPV Taiwan PQ type photovoltaic reactive power control time constant, For the Nth PQPV Taiwan PQ type photovoltaic voltage-reactive power control droop coefficient; , , For the Nth PVPV The time constant for reactive power control of PV type photovoltaic systems. For the Nth PVPV PV type photovoltaic voltage-reactive power control droop coefficient; , , The reactive power control time constant for the first wind turbine. The voltage-reactive power control droop coefficient for the first wind turbine; , , For the Nth PQWT The reactive power control time constant of a PQ-class wind turbine. For the Nth PQWT Voltage-reactive power control droop coefficient for PQ type wind turbines; , , For the Nth PVWT Reactive power control time constant of a PV type wind turbine For the Nth PVWT Voltage-reactive power control droop coefficient for PV type wind turbines.

8. A real-time load power control method considering voltage over-limit as described in claim 6, characterized in that, Step S2 also includes applying the formulas in the reactive-voltage state-space model of the system. Discretize into The system reactive-voltage state-space model, combined with voltage sensitivity, yields the system feeder voltage prediction model: ， In the formula, V(k) is the system node voltage at step k, V(0) is the initial system voltage, and ΔV0(k) is the change in the output voltage of the real-time voltage regulator. ΔP(k) and ΔQ(k) are the changes in net active power and net reactive power at the node, respectively; Where V(0) is the initial voltage of the system; V = (V1, V2, V3, ..., V... N ) T V1 is the voltage at the first node, V N Let P be the voltage at the Nth node; P and Q are the active and reactive power matrices of the node, respectively, P = (P1, P2, P3, ..., P...). N ) T Q = (Q1, Q2, Q3, ..., Q) N ) T P1 and Q1 are the injected active and reactive power of the first node, respectively. N and Q N These represent the injected active power and reactive power at the Nth node, respectively. = ; = 。 9. A real-time load power control method considering voltage over-limit as described in claim 8, characterized in that, Based on the system feeder voltage prediction model obtained in step S2, a rolling optimization model for feeder load power control is constructed: Where V(k) is the feeder voltage at time k; V0 is the target value of the feeder voltage within the corresponding time period; N p Let N be the set of sampling steps within the prediction period. p =[1, …, N p ] T ; N p To predict the number of sampling steps in a prediction period, N p =T p / ΔT p , T p is the prediction period, and ΔT p is the sampling period. Solving the rolling optimization model can obtain the voltage of each node, the action of reactive power regulation devices, and the power regulation of feeder loads in the control period.

10. A real-time load power control method considering voltage over-limit as described in claim 9, characterized in that, The rolling optimization model includes the following constraints: voltage over-limit constraint, reactive power voltage regulation equipment control quantity constraint, reactive power voltage regulation equipment control quantity remains unchanged within the same control cycle, and voltage sensitivity remains unchanged within the prediction cycle.

Images