Adaptive voltage control method and system for active power distribution network under strong disturbance conditions

Abstract

The application discloses a self-adaptive voltage control method and system of an active power distribution network under strong disturbance conditions, and the method comprises the following steps: constructing a dynamic evolution model of the power distribution network with node voltage as a state variable; constructing a double-layer control framework; the first layer is based on the dynamic evolution model and type I, and uses a hierarchical strategy, a discrete device life priority strategy, a load time translation peak shifting strategy and a system optimization strategy to realize and output a period optimization step; the second layer is based on an observer model, a prediction optimization model and a constraint model, and uses a feedforward strategy, a time domain rolling strategy, a target weighted rolling strategy and a terminal constraint strategy to realize and output a period rolling optimization step. The system is used for realizing the method. The application aims to effectively decouple the medium and long-term system basic load peak shaving and valley filling and the short-time high-frequency voltage rolling optimization by adopting the double-layer control framework, so that the balance between global optimization and local dynamic compensation is realized.

Description

Technical Field

[0001] This invention relates to the field of distribution automation and active distribution networks in power systems, specifically to an adaptive voltage control method and system for active distribution networks under strong disturbance conditions. Background Technology

[0002] To accelerate the construction of new power systems, the penetration rate of renewable energy sources such as wind and solar power and electric vehicles in distribution networks is constantly increasing. Although this has effectively reduced the carbon emissions of the system, the randomness and volatility of the output of such distributed resources can easily cause the voltage of the distribution network to frequently exceed the limit, making its regulation more difficult. To address this problem, the existing technologies mainly have the following shortcomings: (1) The original discrete voltage regulating equipment in the distribution network (such as on-load tap-changing transformers and parallel capacitor banks) has a slow response speed and a fixed adjustment step size, which makes it difficult to meet the current grid's demand for continuous, real-time and fine voltage regulation. (2) When dealing with flexible resources such as electric vehicles, existing studies mostly describe them as uncertain terms and use complex probability distribution models or robust optimization methods for system regulation. However, the behavior of such users is highly flexible, and the accuracy of the model will directly affect the feasibility of the regulation results. Under drastic fluctuation conditions (such as the disorderly access of a large number of vehicles and the random fluctuation of photovoltaics), the voltage fluctuates drastically, which seriously affects the power supply reliability of the distribution network. Summary of the Invention

[0003] The technical problem to be solved by this invention is to provide an adaptive voltage control method and system for active distribution networks under strong disturbance conditions, addressing the aforementioned problems of the prior art. This invention aims to overcome the slow response speed and fixed adjustment step size of the existing discrete voltage regulation equipment in the distribution network, which makes it difficult to meet the current power grid's demand for continuous, real-time and fine voltage regulation.

[0004] To solve the above-mentioned technical problems, the technical solution adopted by the present invention includes the following steps: S1. Construct a dynamic evolution model of the distribution network with node voltage as the state variable. This model includes a mathematical model of photovoltaic (PV) system participation in voltage regulation, an energy storage system output model, an electric vehicle (EV) mathematical model, and a discrete control equipment mathematical model. The PV system voltage regulation model is used to tap into the remaining reactive power capacity of the inverter and regulate reactive power. The energy storage system output model is used to balance peak-valley differences caused by short-term fluctuations by responding in real-time to changes in source load within the system. The EV mathematical model includes a Type I EV model and a Type II EV model. Type I EVs are fixed loads under discrete control, while Type II EVs are flexible resources whose power can be continuously adjusted within the range [0, P_s_max]. The discrete control equipment mathematical model includes an on-load tap-changing transformer (OLTC) model and a parallel capacitor bank (CB) model. The OLTC adjusts the transformer ratio by changing the transformer winding tap changer position to achieve discrete control of node voltage. The CB provides reactive power support to the grid by controlling the connection and disconnection of different numbers of capacitor banks. Output the dynamic evolution model of the distribution network. S2, constructing a two-layer control architecture; the first layer, based on the aforementioned distribution network dynamic evolution model, employs a hierarchical strategy, a discrete equipment lifespan priority strategy, a load time shifting and peak-shaving strategy, and a system optimization strategy. It performs periodic optimization calculations with the first-layer scheduling cycle as the step size and outputs first-layer control commands, including the energy storage system's charging and discharging plan, the OLTC gear adjustment plan, and the CB switching plan. The system optimization strategy sets the load of Type I and Type II electric vehicles to zero and calculates the optimal power flow under undisturbed conditions based on source load output prediction data. The hierarchical strategy aims to minimize adjustment requirements, optimizing the output of energy storage and traditional discrete equipment to achieve peak-valley balance of the system's base load. The load time shifting and peak-shaving strategy adjusts the charging start time within the maximum allowable delay time while maintaining constant power charging for Type I electric vehicles. The discrete equipment lifespan priority strategy aims to minimize the number of OLTC and CB actions and the voltage deviation. The second layer is based on a disturbance observer model, a predictive optimization model, and a constraint model. It employs a feedforward strategy, a time-domain rolling strategy, a target-weighted rolling strategy, and a terminal constraint strategy. It performs periodic rolling optimization with the second-layer sampling period as the step size and outputs second-layer control commands. These second-layer control commands include reactive power regulation commands for the photovoltaic inverter and power regulation commands for the Type II electric vehicle. The disturbance observer model is designed based on Lyapunov stability theory and includes a state observer and a disturbance observer, used to estimate the system voltage state and unknown disturbances in real time and output disturbance estimates. The feedforward strategy uses these disturbance estimates as known feedforward quantities to input into the predictive optimization model. The time-domain rolling strategy repeatedly solves the optimization problem within the predictive time domain with the second-layer sampling period as the step size. The target-weighted rolling strategy introduces time-varying adjustment costs to reflect dynamic electricity market prices and demand-side peak-valley characteristics. The terminal constraint strategy ensures that the charging needs of the Type II electric vehicle are met before the grid disconnection time. S3, the first-level control command and the second-level control command are sent to the corresponding controllable devices for execution, and the actual voltage output of the system after execution is collected and fed back to the disturbance observer model to update the disturbance estimate for the next cycle, thus forming a closed-loop adaptive voltage control.

[0005] Furthermore, the mathematical model formula for the photovoltaic system participating in voltage regulation in step S1 is as follows:

[0006] in and These represent the active and reactive power outputs of photovoltaic power, respectively. This refers to the response time interval of the photovoltaic inverter. The sampling period; Power output for the photovoltaic port; Decision variables for controlling reactive power output; Sampling time; This refers to the capacity of the inverter; The power factor.

[0007] Furthermore, the energy storage system output model formula in step S1 is as follows:

[0008] in , They are nodes The lower and upper limits of the charging power of energy storage devices; , They are nodes The lower and upper limits of the discharge power of the energy storage device; For charging decision variables; For discharge decision variables; for Energy stored in an instantaneous energy storage device; , These refer to the charging and discharging efficiency of the energy storage device; , These are the upper and lower limits of energy storage capacity, respectively. This refers to the actual charging power. This represents the actual discharge power. Stored energy for the previous moment. For time step, This is the lower limit of energy storage capacity. This represents the upper limit of energy storage capacity.

[0009] Furthermore, the electric vehicle mathematical model in step S1 includes a Type I electric vehicle model and a Type II electric vehicle model. The Type I electric vehicle model is a fixed load with discrete control mode; the Type II electric vehicle model is a variable power and flexible energy adjustment model, capable of... The ability to continuously adjust within the interval allows the dispatch center to dynamically adjust its charging and discharging power according to the grid's operating status. The formulas for the Type I and Type II electric vehicle models are as follows:

[0010] in The charging efficiency coefficient; For the battery capacity of electric vehicles; Power for charging cars; For the delay time interval, The charging power at time k, The energy sequence after delay processing, For the original energy sequence in (k d) The actual value at time d, where d is the delay time interval;

[0011] in Efficiency coefficient; , These are the upper and lower limits of the capacity of electric vehicles, respectively. For the power output of Type II electric vehicles, The equivalent energy demand of the Type II electric vehicle at time k after scheduling optimization is given. Let k be the expected energy requirement of the original Type II electric vehicle at time k. To store energy in the battery of the Type II electric vehicle at time k. The sampling period.

[0012] Furthermore, the discrete control equipment mathematical model in step S1 includes an OLTC model and a CB model. The OLTC model adjusts the transformer ratio by changing the tap position of the transformer windings to achieve discrete control of the distribution network node voltage, constraining the range of taps and controlling the total number of operations within the dispatch cycle. The CB model provides reactive power support to the grid by controlling the connection and disconnection of different groups of capacitors. The single connection / disconnection capacity is a fixed value, and the total number of connection / disconnection operations is constrained. The formulas for the OLTC model and the CB model are as follows:

[0013] in for OLTC gear position during the time period For the k-1 time period, the gear position of OLTC; , These are the upper and lower limits of the OLTC gear positions, respectively. This represents the maximum number of actions within an OLTC action cycle.

[0014] in for The reactive power output of CB during the time period; for Number of CB (Between Throws) groups during the time period; This is the reactive power compensation amount for a single CB group; The maximum number of CB throw groups; This represents the maximum number of throws and cuts within the CB action cycle.

[0015] Furthermore, in step S2, the optimal system strategy sets the load of Type I and Type II electric vehicles to zero. Based on the source load output prediction data, the optimal power flow of the system under undisturbed operating conditions is calculated. The hierarchical strategy consists of two layers: the first layer aims to minimize the adjustment demand, optimizing the output of energy storage and traditional discrete equipment to achieve peak-valley balance of the system's base load; the second layer fixes the upper-level equipment operation scheme, utilizes the flexibility adjustment margin of photovoltaic and electric vehicles, and continuously optimizes the voltage deviation caused by short-term fluctuations in new energy sources; the load time shifting and peak-shaving strategy maintains a constant charging power, keeping the total energy demand of Type I electric vehicles constant and adjusting the charging start time within the maximum allowable delay time; the discrete equipment lifespan priority strategy aims to minimize the number of OLTC and CB operation cycles and voltage deviation. The specific optimization problem can be expressed as:

[0016] in For variables The state in the first-level control optimization; To optimize timing; and The weights are positive definite and semi-positive definite, respectively. The matrix represents the rated voltage; and These are the number of Type I and Type II electric vehicles connected to the grid, respectively. The scheduling cycle for the first-layer model, This represents the total number of time periods for the first layer of optimization. To control the weighted quadratic norm of action changes, To control the input matrix, The perturbation input matrix, The vector of perturbation changes considered for the first layer. The energy vector of the energy storage system at the start of the first-level scheduling cycle. This represents the energy vector of the energy storage system at the end of the scheduling cycle. The scheduling cycle for the first-layer model. This represents the power vector of the energy storage system during the last time period.

[0017] Furthermore, in step S2, the feedforward strategy is based on the state observer designed in this paper, which transforms the random fluctuations caused by conventional loads and electric vehicle access into known quantities; the time-domain rolling strategy is based on... The sampling period is in the prediction time domain. The internal rolling optimization is performed; the objective weighted rolling strategy aims to minimize the overall operating cost of the system, introducing time-varying adjustment costs into the model. To reflect dynamic electricity market prices and demand-side peak-valley characteristics; the terminal constraint strategy is to introduce constraints... Terminal constraints at specific times ensure that Type I and Type II electric vehicles are not disconnected from the grid during off-grid periods. Before charging is completed, let the latest time for the currently connected vehicle to leave the network be... The observer model includes a state observer and a disturbance observer to track and estimate the system voltage state. and disturbance Lyapunov functions The formula for estimating random disturbances is:

[0018] in Let be the observer matrix. This is the estimated value of the change in disturbance at time k. This is the actual output of the system. Output an estimated value for the system; The state observer estimation method is as follows:

[0019] in For unit array; Let be the observer matrix. Let be the node voltage estimation vector at time k. To control the input matrix, To control the amount of change in variables, The perturbation input matrix, For actual disturbances, This is the measured voltage output vector. This is the output matrix; Lyapunov function for:

[0020] in It is a positive definite matrix. To augment the error vector, For voltage estimation error, For the disturbance estimation error, For real disturbances, Adjust the observer's gain matrix to obtain the perturbation estimate. make It exhibits a monotonically decreasing trend, that is, it satisfies the following formula:

[0021] in To augment the system matrix, To augment the output matrix; The formula for the prediction optimization model is:

[0022] in For time-varying adjustment costs, For the number of Type II electric vehicles, Let V be the power change of the j-th type II electric vehicle at time t+k, and V be the node voltage vector. T1 is the rated voltage vector, and T2 is the voltage deviation weight matrix; The constraint model formula is:

[0023] in: Control rate; In control state; For the j-th type II electric vehicle at the terminal time Battery energy; The target energy for the j-th type II electric vehicle; Let j be the time when the j-th type II electric vehicle leaves the network; The maximum charging power of the j-th Type II electric vehicle; Charging efficiency for Type II electric vehicles; For the second layer sampling period, The minimum energy constraint of the battery is determined by the maximum permissible depth of discharge for electric vehicles; This refers to the predicted energy of the energy storage system at the terminal moment. The number of electric vehicles that have successfully completed charging; For the first The constant charging power of a Type I electric vehicle at node m at time m; The control power for Type II electric vehicles.

[0024] This invention also provides an adaptive voltage control system for an active distribution network under strong disturbance conditions, comprising: The modeling module is used to construct a dynamic evolution model of the distribution network with node voltage as the state variable. The dynamic evolution model of the distribution network includes a mathematical model of the photovoltaic system participating in voltage regulation, a power output model of the energy storage system, a mathematical model of electric vehicles, and a mathematical model of discrete control equipment. The first-level optimization module is used to perform periodic optimization calculations based on the dynamic evolution model of the distribution network, using a hierarchical strategy, a discrete equipment lifespan priority strategy, a load time shifting and peak-shaving strategy, and a system optimal strategy, with the first-level scheduling cycle as the step size, and output the first-level control commands. The first-level control commands include the charging and discharging plan of the energy storage system, the tap adjustment plan of the on-load tap changer (OLTC), and the switching plan of the parallel capacitor bank (CB). The disturbance observation module is used for disturbance observers designed based on Lyapunov stability theory to estimate the system voltage state and unknown disturbances in real time and output disturbance estimates. The second-layer optimization module is used to perform rolling optimization based on the disturbance estimate, prediction optimization model and constraint model, using feedforward strategy, time-domain rolling strategy, target-weighted rolling strategy and terminal constraint strategy, with the second-layer sampling period as the step size, and output the second-layer control command, which includes the reactive power regulation command of the photovoltaic inverter and the power regulation command of the Type II electric vehicle. The instruction issuance and feedback module is used to issue the first-level control instruction and the second-level control instruction to the corresponding controllable devices for execution, collect the actual voltage output of the system after execution and feed it back to the disturbance observation module to form a closed-loop adaptive voltage control.

[0025] The present invention also provides a computer-readable storage medium storing a computer program, which, when executed by a processor, implements the steps of the above-described adaptive voltage control method for an active distribution network under strong disturbance conditions.

[0026] The present invention also provides a computer program product, including a computer program that, when executed by a processor, implements the steps of the above-described adaptive voltage control method for an active distribution network under strong disturbance conditions.

[0027] Compared with the prior art, the present invention has the following beneficial effects: (1) By designing a disturbance observer based on Lyapunov stability theory, it is possible to convert complex external unknown power disturbances (such as large-scale random access or disconnection of electric vehicles) into known feedforward quantities required by the closed-loop control system in real time and accurately, thus overcoming the limitation of control accuracy on the randomness of the external environment and significantly improving the robustness of voltage control under strong disturbance conditions.

[0028] (2) A dual-layer control architecture is adopted to effectively decouple the life optimization of slow discrete equipment (OLTC, CB) from the high-frequency voltage regulation of fast continuous equipment (photovoltaic inverter, electric vehicle): the first layer optimizes the peak shaving and valley filling of the global base load with a longer cycle, and the second layer compensates for the voltage deviation caused by short-term fluctuations in real time with a shorter cycle, thus achieving a balance between global optimization and local dynamic compensation.

[0029] (3) Fully explore the adjustment potential of various flexible resources: use the remaining reactive capacity of photovoltaic inverters to provide reactive support, use the continuous power regulation capability of Type II electric vehicles for fine voltage control, use energy storage systems for peak-valley balance, and use multiple resources to improve the voltage control effect.

[0030] (4) The second layer adopts a time-domain rolling optimization strategy. At each sampling time, the optimization problem is resolved based on the latest disturbance observation and system state. It can adaptively track time-varying disturbances and avoid the performance degradation problem of traditional open-loop prediction methods when the model is mismatched. Attached Figure Description

[0031] Figure 1 This is the improved IEEE 33-node power distribution system in this invention.

[0032] Figure 2 This refers to photovoltaic prediction data in this invention.

[0033] Figure 3 This describes the tracking situation in scenario 1 of this invention.

[0034] Figure 4 This describes the tracking situation in scenario 2 of this invention.

[0035] Figure 5 This refers to the power of the Type I electric vehicle in this invention.

[0036] Figure 6 This refers to the power of the Type II electric vehicle in this invention.

[0037] Figure 7 This describes the output status of the ESS in this invention.

[0038] Figure 8 This is a comparison of voltage deviations in this invention.

[0039] Figure 9 This is the basic flowchart of the present invention. Detailed Implementation

[0040] To enable those skilled in the art to better understand the technical solutions of the present invention, the technical solutions of the present invention will be further described in detail below with reference to the accompanying drawings in the embodiments of the present invention.

[0041] The method in this embodiment specifically includes: S1. Construct a dynamic evolution model of the distribution network with node voltage as the state variable. This model includes a mathematical model of photovoltaic (PV) system participation in voltage regulation, an energy storage system output model, an electric vehicle (EV) mathematical model, and a discrete control equipment mathematical model. The PV system voltage regulation model is used to tap into the remaining reactive power capacity of the inverter and regulate reactive power. The energy storage system output model is used to balance peak-valley differences caused by short-term fluctuations by responding in real-time to changes in source load within the system. The EV mathematical model includes a Type I EV model and a Type II EV model. Type I EVs are fixed loads under discrete control, while Type II EVs are flexible resources whose power can be continuously adjusted within the range [0, P_s_max]. The discrete control equipment mathematical model includes an on-load tap-changing transformer (OLTC) model and a parallel capacitor bank (CB) model. The OLTC adjusts the transformer ratio by changing the transformer winding tap changer position to achieve discrete control of node voltage. The CB provides reactive power support to the grid by controlling the connection and disconnection of different numbers of capacitor banks. Output the dynamic evolution model of the distribution network. S2, constructing a two-layer control architecture; the first layer, based on the aforementioned distribution network dynamic evolution model, employs a hierarchical strategy, a discrete equipment lifespan priority strategy, a load time shifting and peak-shaving strategy, and a system optimization strategy. It performs periodic optimization calculations with the first-layer scheduling cycle as the step size and outputs first-layer control commands, including the energy storage system's charging and discharging plan, the OLTC gear adjustment plan, and the CB switching plan. The system optimization strategy sets the load of Type I and Type II electric vehicles to zero and calculates the optimal power flow under undisturbed conditions based on source load output prediction data. The hierarchical strategy aims to minimize adjustment requirements, optimizing the output of energy storage and traditional discrete equipment to achieve peak-valley balance of the system's base load. The load time shifting and peak-shaving strategy adjusts the charging start time within the maximum allowable delay time while maintaining constant power charging for Type I electric vehicles. The discrete equipment lifespan priority strategy aims to minimize the number of OLTC and CB actions and the voltage deviation. The second layer is based on a disturbance observer model, a predictive optimization model, and a constraint model. It employs a feedforward strategy, a time-domain rolling strategy, a target-weighted rolling strategy, and a terminal constraint strategy. It performs periodic rolling optimization with the second-layer sampling period as the step size and outputs second-layer control commands. These second-layer control commands include reactive power regulation commands for the photovoltaic inverter and power regulation commands for the Type II electric vehicle. The disturbance observer model is designed based on Lyapunov stability theory and includes a state observer and a disturbance observer, used to estimate the system voltage state and unknown disturbances in real time and output disturbance estimates. The feedforward strategy uses these disturbance estimates as known feedforward quantities to input into the predictive optimization model. The time-domain rolling strategy repeatedly solves the optimization problem within the predictive time domain with the second-layer sampling period as the step size. The target-weighted rolling strategy introduces time-varying adjustment costs to reflect dynamic electricity market prices and demand-side peak-valley characteristics. The terminal constraint strategy ensures that the charging needs of the Type II electric vehicle are met before the grid disconnection time. S3, the first-level control command and the second-level control command are sent to the corresponding controllable devices for execution, and the actual voltage output of the system after execution is collected and fed back to the disturbance observer model to update the disturbance estimate for the next cycle, thus forming a closed-loop adaptive voltage control.

[0042] In a specific application example, the mathematical model formula for the photovoltaic system participating in voltage regulation in step S1 is as follows:

[0043] in and These represent the active and reactive power outputs of photovoltaic power, respectively. This refers to the response time interval of the photovoltaic inverter. The sampling period; Power output for the photovoltaic port; Decision variables for controlling reactive power output; Sampling time; This refers to the capacity of the inverter; The power factor.

[0044] In a specific application example, the energy storage system output model formula in step S1 is:

[0045] in , They are nodes The lower and upper limits of the charging power of energy storage devices; , They are nodes The lower and upper limits of the discharge power of the energy storage device; For charging decision variables; For discharge decision variables; for Energy stored in an instantaneous energy storage device; , These refer to the charging and discharging efficiency of the energy storage device; , These are the upper and lower limits of energy storage capacity, respectively. This refers to the actual charging power. This represents the actual discharge power. Stored energy for the previous moment. For time step, This is the lower limit of energy storage capacity. This represents the upper limit of energy storage capacity.

[0046] In a specific application example, the electric vehicle mathematical model in step S1 includes a Type I electric vehicle model and a Type II electric vehicle model. The Type I electric vehicle model represents a fixed load under discrete control mode; the Type II electric vehicle model features variable power and flexible energy adjustment, and is capable of... The ability to continuously adjust within the interval allows the dispatch center to dynamically adjust its charging and discharging power according to the grid's operating status. The formulas for the Type I and Type II electric vehicle models are as follows:

[0047] in The charging efficiency coefficient; For the battery capacity of electric vehicles; Power for charging cars; For the delay time interval, The charging power at time k, The energy sequence after delay processing, For the original energy sequence in (k d) The actual value at time d, where d is the delay time interval;

[0048] in Efficiency coefficient; , These are the upper and lower limits of the capacity of electric vehicles, respectively. For the power output of Type II electric vehicles, The equivalent energy demand of the Type II electric vehicle at time k after scheduling optimization is given. Let k be the expected energy requirement of the original Type II electric vehicle at time k. To store energy in the battery of the Type II electric vehicle at time k. The sampling period.

[0049] In a specific application example, the discrete control equipment mathematical model in step S1 includes an OLTC model and a CB model. The OLTC model adjusts the transformer ratio by changing the tap position of the transformer winding, thereby achieving discrete control of the voltage at the distribution network nodes, constraining the range of taps, and controlling the total number of operations within the dispatch cycle. The CB model provides reactive power support to the grid by controlling the connection and disconnection of different groups of capacitors. The single switching capacity is a fixed value, and the total number of switching operations is constrained. The formulas for the OLTC model and the CB model are as follows:

[0050] in for OLTC gear position during the time period For the k-1 time period, the gear position of OLTC; , These are the upper and lower limits of the OLTC gear positions, respectively. This represents the maximum number of actions within an OLTC action cycle.

[0051] in for The reactive power output of CB during the time period; for Number of CB (Between Throws) groups during the time period; This is the reactive power compensation amount for a single CB group; The maximum number of CB throw groups; This represents the maximum number of throws and cuts within the CB action cycle.

[0052] In a specific application example, the optimal system strategy in step S2 sets the load of Type I and Type II electric vehicles to zero. Based on the source load output prediction data, the optimal power flow of the system under undisturbed operating conditions is calculated. The hierarchical strategy consists of two layers: the first layer aims to minimize the adjustment demand, optimizing the output of energy storage and traditional discrete equipment to achieve peak-valley balance of the system's base load; the second layer fixes the upper-level equipment operation scheme, utilizes the flexibility adjustment margin of photovoltaic and electric vehicles, and continuously optimizes the voltage deviation caused by short-term fluctuations in new energy sources; the load time shifting and peak-shaving strategy maintains a constant charging power, keeping the total energy demand of Type I electric vehicles constant and adjusting the charging start time within the maximum allowable delay time; the discrete equipment lifespan priority strategy aims to minimize the number of OLTC and CB operation cycles and voltage deviation. The specific optimization problem can be expressed as:

[0053] in For variables The state in the first-level control optimization; To optimize timing; and The weights are positive definite and semi-positive definite, respectively. The matrix represents the rated voltage; and These are the number of Type I and Type II electric vehicles connected to the grid, respectively. The scheduling cycle for the first-layer model, This represents the total number of time periods for the first layer of optimization. To control the weighted quadratic norm of action changes, To control the input matrix, The perturbation input matrix, The vector of perturbation changes considered for the first layer. The energy vector of the energy storage system at the start of the first-level scheduling cycle. This represents the energy vector of the energy storage system at the end of the scheduling cycle. The scheduling cycle for the first-layer model. This represents the power vector of the energy storage system during the last time period.

[0054] In a specific application example, the feedforward strategy in step S2 is based on the state observer designed in this paper, which transforms the random fluctuations caused by conventional loads and electric vehicle access into known quantities; the time-domain rolling strategy is based on... The sampling period is in the prediction time domain. The internal rolling optimization is performed; the objective weighted rolling strategy aims to minimize the overall operating cost of the system, introducing time-varying adjustment costs into the model. To reflect dynamic electricity market prices and demand-side peak-valley characteristics; the terminal constraint strategy is to introduce constraints... Terminal constraints at specific times ensure that Type I and Type II electric vehicles are not disconnected from the grid during off-grid periods. Before charging is completed, let the latest time for the currently connected vehicle to leave the network be... The observer model includes a state observer and a disturbance observer to track and estimate the system voltage state. and disturbance Lyapunov functions The formula for estimating random disturbances is:

[0055] in Let be the observer matrix. This is the estimated value of the change in disturbance at time k. This is the actual output of the system. Output an estimated value for the system; The state observer estimation method is as follows:

[0056] in For unit array; Let be the observer matrix. Let be the node voltage estimation vector at time k. To control the input matrix, To control the amount of change in variables, The perturbation input matrix, For actual disturbances, This is the measured voltage output vector. This is the output matrix; Lyapunov function for:

[0057] in It is a positive definite matrix. To augment the error vector, For voltage estimation error, For the disturbance estimation error, For real disturbances, Adjust the observer's gain matrix to obtain the perturbation estimate. make It exhibits a monotonically decreasing trend, that is, it satisfies the following formula:

[0058] in To augment the system matrix, To augment the output matrix; The formula for the prediction optimization model is:

[0059] in For time-varying adjustment costs, For the number of Type II electric vehicles, Let V be the power change of the j-th type II electric vehicle at time t+k, and V be the node voltage vector. T1 is the rated voltage vector, and T2 is the voltage deviation weight matrix; The constraint model formula is:

[0060] in: Control rate; In control state; For the j-th type II electric vehicle at the terminal time Battery energy; The target energy for the j-th type II electric vehicle; Let j be the time when the j-th type II electric vehicle leaves the network; The maximum charging power of the j-th Type II electric vehicle; Charging efficiency for Type II electric vehicles; For the second layer sampling period, The minimum energy constraint of the battery is determined by the maximum permissible depth of discharge for electric vehicles; This refers to the predicted energy of the energy storage system at the terminal moment. The number of electric vehicles that have successfully completed charging; For the first The constant charging power of a Type I electric vehicle at node m at time m; The control power for Type II electric vehicles.

[0061] This embodiment also includes an adaptive voltage control system for an active distribution network under strong disturbance conditions, including: The modeling module is used to construct a dynamic evolution model of the distribution network with node voltage as the state variable. The dynamic evolution model of the distribution network includes a mathematical model of the photovoltaic system participating in voltage regulation, a power output model of the energy storage system, a mathematical model of electric vehicles, and a mathematical model of discrete control equipment. The first-level optimization module is used to perform periodic optimization calculations based on the dynamic evolution model of the distribution network, using a hierarchical strategy, a discrete equipment lifespan priority strategy, a load time shifting and peak-shaving strategy, and a system optimal strategy, with the first-level scheduling cycle as the step size, and output the first-level control commands. The first-level control commands include the charging and discharging plan of the energy storage system, the tap adjustment plan of the on-load tap changer (OLTC), and the switching plan of the parallel capacitor bank (CB). The disturbance observation module is used for disturbance observers designed based on Lyapunov stability theory to estimate the system voltage state and unknown disturbances in real time and output disturbance estimates. The second-layer optimization module is used to perform rolling optimization based on the disturbance estimate, prediction optimization model and constraint model, using feedforward strategy, time-domain rolling strategy, target-weighted rolling strategy and terminal constraint strategy, with the second-layer sampling period as the step size, and output the second-layer control command, which includes the reactive power regulation command of the photovoltaic inverter and the power regulation command of the Type II electric vehicle. The instruction issuance and feedback module is used to issue the first-level control instruction and the second-level control instruction to the corresponding controllable devices for execution, collect the actual voltage output of the system after execution and feed it back to the disturbance observation module to form a closed-loop adaptive voltage control.

[0062] This embodiment also includes a computer-readable storage medium storing a computer program that, when executed by a processor, implements the steps of the above-described adaptive voltage control method for an active distribution network under strong disturbance conditions.

[0063] This embodiment also includes a computer program product, which includes a computer program that, when executed by a processor, implements the steps of the above-described adaptive voltage control method for an active distribution network under strong disturbance conditions.

[0064] This embodiment is based on a simulation verification of an improved IEEE 33-node distribution system. The system parameters are configured as follows: reference voltage V_B = 12.66 kV, reference capacity S_B = 100 MVA, and the allowable range of node voltage is 0.93 ~ 1.07 pu. Energy storage systems with a capacity of 1000 kWh are installed at nodes 6, 10, and 17. Photovoltaic systems with a capacity of 1.5 MW are connected to nodes 15, 18, and 22. Node 13 is designated as the grid connection point for Type I electric vehicles, and node 21 as the grid connection point for Type II electric vehicles. The first-level control step size (scheduling cycle) is 0.5 h, and the second-level sampling period is 1 min. The topology of the improved IEEE 33-node distribution system is as follows: Figure 1 As shown, the photovoltaic output forecast data is as follows: Figure 2 As shown.

[0065] 1. Verification of observer tracking performance To verify the effectiveness of the disturbance observer in this invention under strong disturbance conditions, two extreme scenarios were set for verification analysis: Scenario 1 is large-scale electric vehicle access (positive power surge), and Scenario 2 is large-scale electric vehicle disconnection (negative power drop). The observer's scheduling period was set to 1 hour, and the disturbance around 12:00 noon was taken as the observation period. The tracking performance of the observer in the two scenarios is as follows: Figure 3 and Figure 4 As shown.

[0066] During the undisturbed steady-state period from 0 to 20 minutes, both the observed and actual values ​​stabilize near zero. When a strong disturbance is introduced at t = 20 minutes, regardless of the direction of the disturbance (positive or negative), the observed values ​​eventually achieve a high degree of overlap with the actual unknown nonlinear disturbance. This indicates that the disturbance observer designed based on Lyapunov stability can effectively overcome the limitations of the randomness of the external environment and accurately convert it into the known feedforward quantity required by the closed-loop control system.

[0067] In scenario 1 ( Figure 3 In the data, at time t = 30 min, the actual disturbance exhibited a localized sharp spike, and the observed value showed a brief downward deviation. This is the result of a trade-off between computational efficiency and accuracy by the observer: to prevent the control signal from becoming overly sensitive and causing frequent invalid actions of the discrete voltage regulator, the observer filtered out some extreme high-frequency glitches and then quickly self-corrected within a very short time, re-locking onto the main trend.

[0068] In scenario 2 ( Figure 4 In the initial abrupt change interval of t = 20 ~ 30 min, the actual power rapidly drops to a negative value. During this phase, the observer maintains a high initial gain to quickly eliminate the initial state error, leading to transient overshoot in the control logic. However, based on the Lyapunov stability design, the gain matrix is ​​constrained and exhibits a monotonically decreasing trend. After t = 30 min, as the gain rapidly converges, the observer effectively suppresses divergence and re-tracks the actual disturbance.

[0069] 2. Scheduling Result Analysis The power distribution of the Type I electric vehicle at node 13 is as follows: Figure 5 As shown, once a Type I electric vehicle is connected to the power distribution network, it charges at a constant power, and its total energy demand remains fixed in the grid dispatch. Since the charging power after connection is not adjustable, this type of vehicle is treated as a conventional fixed load. The system can only optimize its grid connection period to avoid peak electricity demand, that is, shift the charging start time within its maximum allowable delay time, but cannot change the power amplitude during a single charging process.

[0070] The charging power distribution of Type II electric vehicles at node 21 is as follows: Figure 6 As shown. Unlike Type I electric vehicles, Type II electric vehicles have continuously dynamic adjustment characteristics in their charging power. The dispatch center can continuously and dynamically adjust the charging and discharging power of the vehicles based on the strong external disturbances and real-time operating status faced by the active distribution network, thereby more effectively mitigating system voltage over-limits. For Type II electric vehicles, the model predictive control strategy does not need to maintain a constant operating power; it only needs to ensure that the vehicle meets the total energy demand at the end of the entire dispatch cycle and reaches a fully charged state before leaving the grid, under the terminal constraints.

[0071] The output of the energy storage system is as follows Figure 7 As shown, ESS1 and ESS2 maintain positive active power output (i.e., discharge state) for most of the day, mainly undertaking the task of supporting the overall load demand of the system and raising the voltage of local nodes; while ESS3 exhibits frequent switching between charging and discharging states, showing significant negative charging output during specific periods, indicating that the model is flexibly utilizing it to absorb excess power in local areas or cope with short-term drastic fluctuations. Through this differentiated dynamic coordination, the first-level model's control benchmark for peak shaving and valley filling of the system's base load is effectively implemented, while mitigating the risk of voltage exceeding limits caused by strong disturbances from renewable energy and flexible loads.

[0072] 3. Analysis of scheduling effects in uncertain scenarios To verify the superiority of the model of this invention, the voltage optimization effects of traditional model predictive control, robust optimization control, and the strategy of this invention were compared. The voltage comparison results are as follows: Figure 8 As shown.

[0073] Depend on Figure 8It can be seen that the voltage management effect of the strategy of the present invention is the best compared with the other two models. The specific analysis is as follows: (1) Traditional model predictive control only relies on source load prediction and lacks a dynamic observation and compensation mechanism for unknown strong disturbances such as "plug and play" electric vehicles or sudden load. This leads to a decrease in model tracking accuracy when dealing with severe nonlinear fluctuations. The largest voltage drop occurs at the middle nodes of the distribution network (such as nodes 15 to 18). (2) Although robust optimization control ensures system safety by setting the worst uncertainty boundary at the mathematical model level, this modeling approach inevitably brings about the conservatism of the control strategy, which limits the full utilization of the system's flexibility resource adjustment margin. Therefore, although its voltage distribution performance is better than that of traditional model predictive control, it still fails to reach the optimal state. (3) The strategy of the present invention introduces a random disturbance observer based on Lyapunov stability design into the control model, which can accurately convert complex external unknown power disturbances into known feedforward inputs required by the closed-loop control system in real time. Based on a two-layer control architecture, this model not only establishes a globally optimal control benchmark in the first layer, but also utilizes these precise observation states in the second layer for high-frequency model rolling optimization and adaptive error correction. This closed-loop feedback mechanism enables the invention to effectively overcome the physical limitations caused by random fluctuations in the external environment, keeping the voltage of each node tightly controlled near the optimal benchmark (1.0 pu) even under extreme operating conditions, demonstrating extremely strong anti-interference capability and control robustness.

[0074] 4. Overall Method Flow like Figure 9 As shown, the overall process of the method of the present invention includes: First, constructing a dynamic evolution model of the distribution network with node voltage as the state variable. This model integrates the mathematical models of photovoltaic systems, energy storage systems, electric vehicles, and discrete control equipment. Then, constructing a two-layer control architecture. The first layer is based on the dynamic evolution model and source-load prediction data, and adopts a hierarchical strategy, a discrete equipment lifespan priority strategy, a load time shifting and peak-shaving strategy, and a system optimal strategy to output energy storage charging and discharging plans, OLTC gear plans, and CB switching plans with a relatively long scheduling cycle. The second layer is based on a disturbance observer model, a predictive optimization model, and a constraint model, and adopts a feedforward strategy, a time-domain rolling strategy, a target-weighted rolling strategy, and a terminal constraint strategy to perform rolling optimization with a relatively short sampling cycle, and outputs photovoltaic reactive power regulation commands and Type II electric vehicle power regulation commands. Finally, the two layers of control commands are sent to the controllable equipment for execution, and the actual voltage is collected and fed back to the disturbance observer to form a closed-loop control.

[0075] The above description is merely a preferred embodiment of the present invention. The scope of protection of the present invention is not limited to the above embodiments. All technical solutions falling within the scope of the present invention's concept are within the scope of protection of the present invention. It should be noted that for those skilled in the art, any improvements and modifications made without departing from the principles of the present invention should also be considered within the scope of protection of the present invention.

Claims

1. An adaptive voltage control method for an active distribution network under strong disturbance conditions, characterized in that, include: S1. Construct a dynamic evolution model of the distribution network with node voltage as the state variable. This model includes a mathematical model of photovoltaic (PV) system participation in voltage regulation, an energy storage system output model, an electric vehicle (EV) mathematical model, and a discrete control equipment mathematical model. The PV system voltage regulation model is used to tap into the remaining reactive power capacity of the inverter and regulate reactive power. The energy storage system output model is used to balance peak-valley differences caused by short-term fluctuations by responding in real-time to changes in source load within the system. The EV mathematical model includes a Type I EV model and a Type II EV model. Type I EVs are fixed loads under discrete control, while Type II EVs are flexible resources whose power can be continuously adjusted within the range [0, P_s_max]. The discrete control equipment mathematical model includes an on-load tap-changing transformer (OLTC) model and a parallel capacitor bank (CB) model. The OLTC adjusts the transformer ratio by changing the transformer winding tap changer position to achieve discrete control of node voltage. The CB provides reactive power support to the grid by controlling the connection and disconnection of different numbers of capacitor banks. Output the dynamic evolution model of the distribution network. S2, constructing a two-layer control architecture; the first layer, based on the aforementioned distribution network dynamic evolution model, employs a hierarchical strategy, a discrete equipment lifespan priority strategy, a load time shifting and peak-shaving strategy, and a system optimization strategy. It performs periodic optimization calculations with the first-layer scheduling cycle as the step size and outputs first-layer control commands, including the energy storage system's charging and discharging plan, the OLTC gear adjustment plan, and the CB switching plan. The system optimization strategy sets the load of Type I and Type II electric vehicles to zero and calculates the optimal power flow under undisturbed conditions based on source load output prediction data. The hierarchical strategy aims to minimize adjustment requirements, optimizing the output of energy storage and traditional discrete equipment to achieve peak-valley balance of the system's base load. The load time shifting and peak-shaving strategy adjusts the charging start time within the maximum allowable delay time while maintaining constant power charging for Type I electric vehicles. The discrete equipment lifespan priority strategy aims to minimize the number of OLTC and CB actions and the voltage deviation. The second layer is based on a disturbance observer model, a predictive optimization model, and a constraint model. It employs a feedforward strategy, a time-domain rolling strategy, a target-weighted rolling strategy, and a terminal constraint strategy. It performs periodic rolling optimization with the second-layer sampling period as the step size and outputs second-layer control commands. These second-layer control commands include reactive power regulation commands for the photovoltaic inverter and power regulation commands for the Type II electric vehicle. The disturbance observer model is designed based on Lyapunov stability theory and includes a state observer and a disturbance observer, used to estimate the system voltage state and unknown disturbances in real time and output disturbance estimates. The feedforward strategy uses these disturbance estimates as known feedforward quantities to input into the predictive optimization model. The time-domain rolling strategy repeatedly solves the optimization problem within the predictive time domain with the second-layer sampling period as the step size. The target-weighted rolling strategy introduces time-varying adjustment costs to reflect dynamic electricity market prices and demand-side peak-valley characteristics. The terminal constraint strategy ensures that the charging needs of the Type II electric vehicle are met before the grid disconnection time. S3, the first-level control command and the second-level control command are sent to the corresponding controllable devices for execution, and the actual voltage output of the system after execution is collected and fed back to the disturbance observer model to update the disturbance estimate for the next cycle, thus forming a closed-loop adaptive voltage control.

2. The adaptive voltage control method for active distribution networks under strong disturbance conditions according to claim 1, characterized in that, The mathematical model formula for the photovoltaic system participating in voltage regulation in step S1 is as follows: in and These represent the active and reactive power outputs of photovoltaic power, respectively. This refers to the response time interval of the photovoltaic inverter. The sampling period; Power output for the photovoltaic port; Decision variables for controlling reactive power output; Sampling time; This refers to the capacity of the inverter; The power factor.

3. The adaptive voltage control method for active distribution networks under strong disturbance conditions according to claim 1, characterized in that, The formula for the energy storage system output model in step S1 is: in , They are nodes The lower and upper limits of the charging power of energy storage devices; , They are nodes The lower and upper limits of the discharge power of the energy storage device; For charging decision variables; For discharge decision variables; for Energy stored in an instantaneous energy storage device; , These refer to the charging and discharging efficiency of the energy storage device; , These are the upper and lower limits of energy storage capacity, respectively. This refers to the actual charging power. This represents the actual discharge power. Stored energy for the previous moment. For time step, This is the lower limit of energy storage capacity. This represents the upper limit of energy storage capacity.

4. The adaptive voltage control method for an active distribution network under strong disturbance conditions according to claim 1, characterized in that, The electric vehicle mathematical model in step S1 includes a Type I electric vehicle model and a Type II electric vehicle model. The Type I electric vehicle model is a fixed load with discrete control mode; the Type II electric vehicle model is a variable power and flexible energy adjustment model, capable of... The ability to continuously adjust within the interval allows the dispatch center to dynamically adjust its charging and discharging power according to the grid's operating status. The formulas for the Type I and Type II electric vehicle models are as follows: in The charging efficiency coefficient; For the battery capacity of electric vehicles; Power for charging cars; For the delay time interval, The charging power at time k, The energy sequence after delay processing, For the original energy sequence in (k d) The actual value at time d, where d is the delay time interval; in Efficiency coefficient; , These are the upper and lower limits of the capacity of electric vehicles, respectively. For the power output of Type II electric vehicles, The equivalent energy demand of the Type II electric vehicle at time k after scheduling optimization is given. Let k be the expected energy requirement of the original Type II electric vehicle at time k. To store energy in the battery of the Type II electric vehicle at time k. The sampling period.

5. The adaptive voltage control method for an active distribution network under strong disturbance conditions according to claim 1, characterized in that, In step S1, the OLTC model adjusts the transformer ratio by changing the tap position of the transformer winding, thereby achieving discrete control of the voltage at distribution network nodes, constraining the range of tap positions, and controlling the total number of operations within the dispatch cycle. The CB model provides reactive power support to the grid by controlling the connection and disconnection of different numbers of capacitors. The single connection / disconnection capacity is a fixed value, and the total number of connection / disconnection operations is constrained. The formulas for the OLTC model and the CB model are as follows: in for OLTC gear position during the time period For the k-1 time period, the gear position of OLTC; , These are the upper and lower limits of the OLTC gear positions, respectively. This represents the maximum number of actions within an OLTC action cycle. in for The reactive power output of CB during the time period; for Number of CB (Between Throws) groups during the time period; This is the reactive power compensation amount for a single CB group; The maximum number of CB throw groups; This represents the maximum number of throws and cuts within the CB action cycle.

6. The adaptive voltage control method for an active distribution network under strong disturbance conditions according to claim 1, characterized in that, In step S2, the optimal system strategy sets the load of Type I and Type II electric vehicles to zero. Based on the source load output prediction data, the optimal power flow of the system under undisturbed operating conditions is calculated. The hierarchical strategy consists of two layers: the first layer aims to minimize the adjustment demand, optimizing the output of energy storage and traditional discrete equipment to achieve peak-valley balance of the system's base load; the second layer fixes the upper-level equipment operation scheme, utilizes the flexibility adjustment margin of photovoltaic and electric vehicles, and continuously optimizes the voltage deviation caused by short-term fluctuations in new energy sources; the load time shifting and peak-shaving strategy maintains a constant charging power, keeping the total energy demand of Type I electric vehicles constant and adjusting the charging start time within the maximum allowable delay time; the discrete equipment lifespan priority strategy aims to minimize the number of OLTC and CB operation cycles and voltage deviation. The specific optimization problem can be expressed as: in For variables The state in the first-level control optimization; To optimize timing; and The weights are positive definite and semi-positive definite, respectively. The matrix represents the rated voltage; and These are the number of Type I and Type II electric vehicles connected to the grid, respectively. The scheduling cycle for the first-layer model, This represents the total number of time periods for the first layer of optimization. To control the weighted quadratic norm of action changes, To control the input matrix, The perturbation input matrix, The vector of perturbation changes considered for the first layer. The energy vector of the energy storage system at the start of the first-level scheduling cycle. This represents the energy vector of the energy storage system at the end of the scheduling cycle. The scheduling cycle for the first-layer model. This represents the power vector of the energy storage system during the last time period.

7. The adaptive voltage control method for an active distribution network under strong disturbance conditions according to claim 1, characterized in that, In step S2, the feedforward strategy is based on the state observer designed in this paper, which transforms the random fluctuations caused by conventional loads and electric vehicle access into known quantities; the time-domain rolling strategy is based on... The sampling period is in the prediction time domain. The internal rolling optimization is performed; the objective weighted rolling strategy aims to minimize the overall operating cost of the system, introducing time-varying adjustment costs into the model. To reflect dynamic electricity market prices and demand-side peak-valley characteristics; the terminal constraint strategy is to introduce constraints... Terminal constraints at specific times ensure that Type I and Type II electric vehicles are not disconnected from the grid during off-grid periods. Before charging is completed, let the latest time for the currently connected vehicle to leave the network be... The observer model includes a state observer and a disturbance observer to track and estimate the system voltage state. and disturbance Lyapunov functions The formula for estimating random disturbances is: in Let be the observer matrix. This is the estimated value of the change in disturbance at time k. This is the actual output of the system. Output an estimated value for the system; The state observer estimation method is as follows: in For unit array; Let be the observer matrix. Let be the node voltage estimation vector at time k. To control the input matrix, To control the amount of change in variables, The perturbation input matrix, For actual disturbances, This is the measured voltage output vector. This is the output matrix; Lyapunov function for: in It is a positive definite matrix. To augment the error vector, For voltage estimation error, For the disturbance estimation error, For real disturbances, Adjust the observer's gain matrix to obtain the perturbation estimate. make It exhibits a monotonically decreasing trend, that is, it satisfies the following formula: in To augment the system matrix, To augment the output matrix; The formula for the prediction optimization model is: in For time-varying adjustment costs, For the number of Type II electric vehicles, Let V be the power change of the j-th type II electric vehicle at time t+k, and V be the node voltage vector. T1 is the rated voltage vector, and T2 is the voltage deviation weight matrix; The constraint model formula is: in: Control rate; In control state; For the j-th type II electric vehicle at the terminal time Battery energy; The target energy for the j-th type II electric vehicle; Let j be the time when the j-th type II electric vehicle leaves the network; The maximum charging power of the j-th Type II electric vehicle; Charging efficiency for Type II electric vehicles; For the second layer sampling period, The minimum energy constraint of the battery is determined by the maximum permissible depth of discharge for electric vehicles; This refers to the predicted energy of the energy storage system at the terminal moment. The number of electric vehicles that have successfully completed charging; For the first The constant charging power of a Type I electric vehicle at node m at time m; The control power for Type II electric vehicles.

8. An adaptive voltage control system for an active distribution network under strong disturbance conditions, characterized in that, include: The modeling module is used to construct a dynamic evolution model of the distribution network with node voltage as the state variable. The dynamic evolution model of the distribution network includes a mathematical model of the photovoltaic system participating in voltage regulation, a power output model of the energy storage system, a mathematical model of electric vehicles, and a mathematical model of discrete control equipment. The first-level optimization module is used to perform periodic optimization calculations based on the dynamic evolution model of the distribution network, using a hierarchical strategy, a discrete equipment lifespan priority strategy, a load time shifting and peak-shaving strategy, and a system optimal strategy, with the first-level scheduling cycle as the step size, and output the first-level control commands. The first-level control commands include the charging and discharging plan of the energy storage system, the tap adjustment plan of the on-load tap changer (OLTC), and the switching plan of the parallel capacitor bank (CB). The disturbance observation module is used for disturbance observers designed based on Lyapunov stability theory to estimate the system voltage state and unknown disturbances in real time and output disturbance estimates. The second-layer optimization module is used to perform rolling optimization based on the disturbance estimate, prediction optimization model and constraint model, using feedforward strategy, time-domain rolling strategy, target-weighted rolling strategy and terminal constraint strategy, with the second-layer sampling period as the step size, and output the second-layer control command, which includes the reactive power regulation command of the photovoltaic inverter and the power regulation command of the Type II electric vehicle. The instruction issuance and feedback module is used to issue the first-level control instruction and the second-level control instruction to the corresponding controllable devices for execution, collect the actual voltage output of the system after execution and feed it back to the disturbance observation module to form a closed-loop adaptive voltage control.

9. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores a computer program, which, when executed by a processor, implements the steps of the adaptive voltage control method for an active distribution network under strong disturbance conditions as described in any one of claims 1 to 7.

10. A computer program product, characterized in that, The system includes a computer program that, when executed by a processor, implements the steps of the adaptive voltage control method for an active distribution network under strong disturbance conditions as described in any one of claims 1 to 7.

Images