Multi-time scale voltage control and reconstruction method and system of novel power system

Abstract

The invention discloses a multi-time scale voltage control and reconstruction method and system for a novel power system, and the method comprises the steps: carrying out the distributed voltage control of a wind power plant at a first time scale, so as to improve the reactive power supporting capability of a fan, and guaranteeing that the node voltage of the wind power plant is maintained in a safe range; under the second time scale, reactive power compensation is carried out on the power distribution network to stabilize node voltage; under the third time scale, performing power distribution network topology reconstruction on the power distribution network to minimize network loss; the time periods of the first time scale, the second time scale and the third time scale are sequentially increased. The invention aims to ensure that all node voltages operate in a feasible range through multi-time scale control and quickly suppress the randomness and volatility of new energy, so as to meet the requirements of stable voltage control and economical and efficient operation of a novel power system (NPS) under the background of distributed new energy access.

Description

Technical Field

[0001] The present invention relates to the technical field of distribution network dispatching, and specifically relates to a multi-time scale voltage control and reconstruction method and system for a new type of power system (NPS). Background Art

[0002] A new type of power system (NPS) refers to a power system that incorporates distributed new energy sources on the basis of a traditional distribution network. The large-scale integration of new energy sources in the new type of power system has made significant contributions to environmental protection and sustainable development. However, due to its inherent randomness, intermittency, and volatility, new energy has brought unprecedented challenges and tests to the traditional distribution network. The large-scale grid connection of a large number of distributed new energy sources has significantly increased the complexity of the power grid, greatly improved the difficulty of management and control, and profoundly affected the regulation mechanism and operating characteristics of the new type of power system.

[0003] To effectively address these challenges and improve the economic efficiency and security of the new type of power system, the industry and academia are widely researching and exploring various technical methods. Voltage control and network reconstruction for improving the economic efficiency and security of the new type of power system have received extensive attention. By implementing refined voltage regulation measures, it is possible to ensure that the grid voltage fluctuates within a reasonable range, effectively avoiding voltage fluctuations caused by the uncertainty of new energy output. By dynamically reconstructing the topological structure of the distribution network, the power flow distribution can be optimized, thereby improving the reliability and operating efficiency of the new type of power system. These strategies not only help to alleviate the pressure brought by the integration of new energy sources but also lay a solid foundation for promoting the transformation towards a green and low-carbon power system.

[0004] Existing distribution network voltage control methods mainly include centralized control, distributed control, and decentralized control. Centralized control of the distribution network voltage sends control commands to each voltage regulation device in the distribution network through a central controller or a master station system to achieve centralized regulation and control of the voltage. However, in the face of a large-scale distribution system, the central processor faces challenges in processing massive real-time data and performing efficient power flow calculations, which may extend the response time of voltage regulation. Distributed control refers to the implementation of independent and coordinated control strategies for voltage through multiple controllers distributed in the distribution network, using electrical quantity information such as voltage and current of local or neighboring devices, as well as possible communication links. Nevertheless, this solution still requires complex communication processes and there are potential competition and interference problems among voltage controllers. The decentralized control method only adjusts its own reactive power compensation capacity or output power based on local electrical quantity measurement data to approach the global optimal solution without exchanging information with other devices or units, which can effectively improve the calculation speed and reduce the dependence on the communication system.

[0005] Distribution network reconfiguration optimizes the operation mode of the distribution system by adjusting the switch states in the network, thereby enhancing the reliability of the distribution system, reducing line losses, achieving load balance, and improving the power supply voltage quality. The network reconfiguration methods mainly include mathematical programming (MP), heuristic algorithms, and switch swapping methods. An effective single-commodity flow constraint is proposed for radial constraints, and the restoration and reconfiguration problems of the distribution system are optimized based on mixed-integer second-order cone programming. Considering the uncertainty of distributed generation, a distributionally robust chance-constrained MILP method based on the Wasserstein distance is proposed for the topological reconfiguration of three-phase unbalanced distribution networks, and the switch cost and expected power supply cost are optimized. Although the MP method can obtain the global optimal solution, its real-time control ability is limited when facing large-scale distribution systems. A multi-objective differential evolution algorithm is proposed for the conditions of various loads and distributed power fluctuations in unbalanced distribution networks, comprehensively optimizing the three-phase unbalanced factor and the number of switches. The prior art also proposes a three-stage switch swapping method. First, all switches are closed, then the switches are sequentially opened to form a radial network, and finally the network structure is further optimized based on the switch swapping strategy. This strategy ensures that the network remains connected during the reconfiguration process and gradually approaches the optimal configuration. To further speed up, the prior art also records using current as an indicator to approximately estimate the change in network losses. This method can quickly respond to the requirements of distribution network reconfiguration and obtain an effect close to the global optimum in an extremely short time. However, the above research does not consider the combined effect and collaborative optimization of a large number of controllable power resources and topological reconfiguration methods on the new power system. Summary of the Invention

[0006] Technical problems to be solved by the present invention: Aiming at the above problems of the prior art, a multi-time scale voltage control and reconfiguration method and system for a new power system are provided. The present invention aims to ensure that the voltages of all nodes operate within a feasible range and quickly suppress the randomness and volatility of new energy through multi-time scale control, so as to meet the voltage stability control and economic and efficient operation requirements of the new power system (NPS) under the background of distributed new energy access.

[0007] To solve the above technical problems, the technical solution adopted by the present invention is as follows: A multi-time scale voltage control and reconfiguration method for a new power system includes the following steps: At the first time scale, decentralized voltage control is performed on the wind farm to enhance the reactive power support ability of the wind turbines and ensure that the node voltages of the wind farm are maintained within a safe range; at the second time scale, reactive power compensation is performed on the distribution network to stabilize the node voltages; at the third time scale, distribution network topology reconfiguration is performed on the distribution network to minimize network losses; the time periods of the first time scale, the second time scale, and the third time scale increase in sequence.

[0008] Optionally, under the first time scale, decentralized voltage control of the wind farm to enhance the reactive power support capacity of the wind turbines and ensure that the node voltage of the wind farm is maintained within a safe range includes: S101, collect the maximum reactive power capacity, reactive power measurement values, and terminal voltages of all wind turbines; S102, establish the objective function for decentralized voltage control of the wind farm: , where, is the objective function, is the number of wind turbines, is the wind turbine voltage change, is the weight matrix of voltage, is the wind turbine reactive power change, is the weight matrix of reactive power; S103, calculate the gradient of the objective function according to the following formula: , where, and are the gradients of the objective function at time , and are the reactive power and reactive power change at time , is the voltage change at time , is the transformation matrix of voltage and reactive power, is the total transformation matrix; S104, diagonalize the matrix of the voltage and reactive power relationship expression in the gradient of the objective function according to the following formula: , , , , where, and are the reactive powers of the wind turbine at times k + 1 and k respectively, and are intermediate variables, is the step coefficient, is the weight coefficient, is the decoupling coefficient between the wind turbine and other wind turbines, is the wind turbine The change in reactive current at time k, where the superscript is the time interval, is the inertia time constant of the outer loop of the wind turbine power, is the amplitude of the grid phase voltage, and are the th diagonal element of the matrix ; is the inverse of the scaling matrix ; is the total transformation matrix, and the scaling matrix is composed of the scaling coefficients of each wind turbine; S105. Introduce the scaling matrix to construct the decentralized voltage control problem shown below to accelerate the convergence speed: , where is the inverse of the scaling coefficient of the wind turbine in the scaling matrix ; S106. Based on the decentralized voltage control problem, introduce the updated step size coefficient and weight coefficient of each wind turbine, and then calculate the reactive power of each wind turbine at time k + 1 as the reactive power reference value : , where is the step size coefficient of the wind turbine , is the weight coefficient of the wind turbine .

[0009] Optionally, the functional expressions for updating the step size coefficient and weight coefficient of each wind turbine in step S106 are: , , , where is an intermediate variable, is the step size coefficient of the wind turbine , is the weight coefficient of the wind turbine , is the number of wind turbines in the wind farm, is the step size coefficient, is the weight coefficient, is the maximum reactive power capacity of the wind turbine .

[0010] Optionally, the reactive power compensation for the distribution network to stabilize the node voltage at the second time scale includes: S201, collecting the voltages and reactive powers of all nodes in the distribution network; S202, calculating the sensitivity of the node voltage to power according to the node voltage, solving with the minimization of voltage deviation as the optimization objective to obtain the reactive power output when the objective function is minimized, and performing reactive power compensation on the distribution network using a static var generator (SVG) according to the solved reactive power output; and the function expression of the objective function is: , where is the objective function of reactive power compensation, is the number of nodes in the distribution network, is the set of predicted voltage deviation values, and there is: , where ~ are the predicted voltage deviation values of the 1st to nth buses at moment respectively, and the calculation function expression of the predicted voltage deviation value of any ith bus at the kth moment is: , where is the voltage of the ith bus at the initial moment , is the voltage of the ith bus, is the reactive power output, is the sensitivity of the voltage of the ith bus to the reactive power , is the reference voltage.

[0011] Optionally, the original topology of the distribution network consists of 1 slack bus and N PQ buses ; the function expression for calculating the sensitivity of the node voltage to power according to the node voltage is: , , , , where is the indicator function and the indicator function is 1 when and 0 otherwise; represents taking the negative of the indicator function multiplied by the imaginary unit j, is the conjugate of the voltage of the ith bus For the active power of the th bus, being the slack bus and the union of PQ buses , being the element in the th row and th column of the nodal admittance matrix, being the complex form of the voltage of the th bus, being the conjugate of the voltage of the th bus, For the reactive power of the th bus, For the sensitivity of the voltage of the th bus with respect to the active power of the th bus, taking the real part, being the phase angle of the th bus, For the sensitivity of the phase angle of the th bus with respect to the active power of the th bus, For the sensitivity of the voltage of the th bus with respect to the reactive power of the th bus, For the sensitivity of the phase angle of the th bus with respect to the reactive power of the

[0012] Optionally, the distribution network topology reconstruction for minimizing network losses in the third time scale includes: S301, collecting the distribution network topology structure, voltage, active power, and reactive power; S302, calculating with the objective function of minimizing network losses and determining the optimal switch state configuration of the distribution network and performing distribution network topology reconstruction on the distribution network. The functional expression of the objective function of the network losses is: , where is the objective function of network losses, is the number of nodes in the distribution network, and are respectively the voltages of the th and th nodes after the distribution network topology reconstruction, is to take the real part, is the node admittance matrix after the distribution network topology reconstruction, After the distribution network topology reconstruction, the th and the th node voltage phase angle difference, and there is: , where, and are respectively the voltage phase angles of the th and the th nodes after the distribution network topology reconstruction.

[0013] Optionally, the calculation function expressions of the voltage and voltage phase angle of the nodes after the distribution network topology reconstruction are: , , where, is the voltage set after the distribution network topology reconstruction, ~ are respectively the voltages of the 1st to nth nodes after the distribution network topology reconstruction, , is the voltage phase angle set after the distribution network topology reconstruction, ~ are respectively the voltage phase angles of the 1st to nth nodes after the distribution network topology reconstruction, is the initial value of the voltage of the node; is the binary variable indicating whether the branch i acts, 1 for acting and 0 for not acting; is the equivalent active power change amount injected at the two ends of the branch i when the branch i acts, is the equivalent reactive power change amount injected at the two ends of the branch i when the branch i acts; is the initial value of the voltage phase angle of the node, is the sensitivity of the voltage of the node to the active power , is the sensitivity of the voltage of the node to the reactive power , is the sensitivity of the voltage phase angle of the node to the active power , is the sensitivity of the voltage phase angle of the node to the reactive power , is the branch set of the distribution network.

[0014] In addition, the present invention also provides a multi-time scale voltage control and reconstruction system for a new type of power system, including a microprocessor and a memory connected to each other, and the microprocessor is programmed or configured to execute the multi-time scale voltage control and reconstruction method for the new type of power system.

[0015] In addition, the present invention also provides a computer-readable storage medium, in which a computer program or instruction is stored, and the computer program or instruction is programmed or configured to execute the multi-time scale voltage control and reconstruction method for the new type of power system through a processor.

[0016] In addition, the present invention also provides a computer program product, including a computer program or instruction, and the computer program or instruction is programmed or configured to execute the multi-time scale voltage control and reconstruction method for the new type of power system through a processor.

[0017] Compared with the prior art, the present invention mainly has the following advantages: The method of the present invention includes, at the first time scale, performing decentralized voltage control on a wind farm to enhance the reactive power support ability of wind turbines and ensure that the node voltage of the wind farm is maintained within a safe range; at the second time scale, performing reactive power compensation on a distribution network to stabilize the node voltage; at the third time scale, performing distribution network topology reconstruction on the distribution network to minimize network losses; the time periods of the first time scale, the second time scale, and the third time scale increase in sequence. The present invention ensures that all node voltages operate within a feasible range and quickly suppresses the randomness and volatility of new energy through multi-time scale control, so as to meet the voltage stability control and economic and efficient operation requirements of a new type of power system (NPS) under the background of distributed new energy access. BRIEF DESCRIPTION OF THE DRAWINGS

[0018] Figure 1 It is a schematic diagram of the basic process of the method in the embodiment of the present invention.

[0019] Figure 2 It is a schematic diagram of the control process at the first time scale in the embodiment of the present invention.

[0020] Figure 3 It is a loop closing operation model in the embodiment of the present invention, where (a) is the network after the switch loop closing operation, (b) is the network before the switch loop closing operation, and (c) is the compensation network.

[0021] Figure 4 It is a 24-hour curve of the load demand and wind power output in one day in the embodiment of the present invention.

[0022] Figure 5 It is the amount of reactive power injected by SVG into the distribution network in the embodiment of the present invention.

[0023] Figure 6 Comparison of the presence or absence of voltage in the distribution network and the control effect of topological reconstruction in the embodiments of the present invention. Specific implementation manners

[0024] In order 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 in conjunction with the accompanying drawings in the embodiments of the present invention.

[0025] As Figure 1 shown, the multi-time scale voltage control and reconstruction method of the new power system in this embodiment includes the following steps: at the first time scale, decentralized voltage control is performed on the wind farm to enhance the reactive power support ability of the wind turbines and ensure that the node voltage of the wind farm is maintained within a safe range; at the second time scale, reactive power compensation is performed on the distribution network to stabilize the node voltage; at the third time scale, distribution network topological reconstruction is performed on the distribution network to minimize network losses; the time periods of the first time scale, the second time scale, and the third time scale increase in sequence.

[0026] Under the constraint that the time periods of the first time scale, the second time scale, and the third time scale increase in sequence, the three required time scales can be selected according to actual needs. For example, as an optional implementation manner, the first time scale, the second time scale, and the third time scale in this embodiment are the second-level time scale, the minute-level time scale (for example, 15 minutes), and the hour-level time scale (for example, 1 hour) respectively. The multi-time scale voltage control and reconstruction method of the new power system in this embodiment uses the variable gradient projection (VGP) method to perform decentralized voltage control in the wind farm (WFs) at the second-level time scale, effectively enhancing the reactive power support ability of the wind turbines and ensuring that the node voltage of the wind farm is maintained within a safe range; at the minute-level time scale, a static var generator (SVG) is used to perform reactive power compensation in the distribution network to flexibly respond to voltage fluctuations and further adjust and stabilize the node voltage; at the hour-level time scale, distribution network topological reconstruction is performed to minimize network losses and maximize economic benefits.

[0027] For the first time scale, this embodiment proposes a voltage control method based on variable gradient projection (VGP) to enhance the safety performance of the NPS. By implementing local fully decentralized gradient projection and adaptive gain adjustment in the wind farm, this method significantly improves the reactive power output of the wind turbines, eliminates communication delays, and speeds up the calculation speed. After stabilizing the terminal voltage of the wind turbines, the SVG is relied on to perform refined voltage regulation on the distribution network.

[0028] Taking a wind farm based on a permanent magnet synchronous generator (PMSG) as an example, it is elaborated in detail. The PMSG mainly consists of three core components: a turbine, a generator, and a converter. The converter includes a generator-side converter and a grid-side converter (GSC). The core function of the grid-side converter is to maintain the stability of the DC-link voltage and provide necessary reactive power support for the wind farm collector system. Since the time constant of the outer power control loop of the GSC is longer than that of its inner current control loop, the dynamic response characteristics of the inner current control loop can usually be approximately described by a first-order lag function. Based on the characteristic analysis of the reactive power control loop of the grid-side converter, the state equation of the reactive power increment of the PMSG wind turbine can be derived as follows: , where, is the state variable is the first derivative of , is the reactive power increment of the grid-side converter, is the integral of the difference between the expected value and the measured value of the reactive power, is q the change in the shaft current, is the control variable, , is the reactive power reference value, is the state matrix, is the input matrix, and there are: , , where, is the inertia time constant of the outer loop of the fan power, is the amplitude of the grid phase voltage, and are the PI controller parameters respectively, is the time constant of the inner loop of the fan current. Setting the sampling interval time to , the discretization of the state equation of the reactive power increment of the PMSG wind turbine can be obtained as follows: , where, and are the state variables at the (k + 1)-th and k-th moments respectively, is the discretized state matrix, is the discretized input matrix, is the control variable at the k-th moment, and there are: , .

[0029] In a wind farm, each wind turbine can be regarded as a current source, responsible for delivering power to the collector system. Given that the line loss is very small compared to the power flow, in the LinDistFlow model, the non-linear equation originally used to describe the relationship between the node voltage and the power injection of the wind farm can be simplified to a linear equation to more effectively show the relationship between the two: , , , where and represent the active power flow and reactive power flow between nodes ij; is the node the set of all directly connected nodes downstream; and are the active injection and reactive injection of the wind turbine on node j; is the voltage at node j; and are the impedances on line ij. For a system with N+1 nodes, the incidence matrix is defined as: , where is related to the slack node, represents the power flow direction of other nodes, and the functional expression of any element in the incidence matrix is:

[0030] Therefore, the LinDistFlow equation can be rewritten in matrix form as: , where represents the sensitivity of the node voltage to the variable, represents the influence of the active power on the voltage. represents the injected active power and reactive power; and are diagonal matrices, representing the resistance and reactance parameters of the system respectively. In a radial wind farm, the voltages of all nodes must meet the requirements of the matrix form of the LinDistFlow equation. In addition, limited by the output capacity of each wind farm converter, the amount of reactive power output needs to meet the following constraints: , where is the reactive power output vector of the WT at node k; and are the upper and lower limits of reactive power respectively. The problem to be solved by voltage control is to reduce the voltage fluctuations at all nodes and increase the reactive power capacity margin of wind turbines. Therefore, the objective function of decentralized voltage control in a wind farm can be expressed as: , where, is the objective function, and are the voltage and reference voltage respectively, and are the reactive power and reference reactive power respectively, is the weight matrix of voltage, is the weight matrix of reactive power; To maintain the maximum reactive power capacity, the rated power is defined as follows . For this quadratic optimization problem, the GP method can be used to find the optimal solution, which requires obtaining the gradient information of the objective function. Therefore, this quadratic optimization problem can be rewritten as: , where, is the objective function, is the number of wind turbines, is the wind turbine 's voltage change amount, is the weight matrix of voltage, is the wind turbine 's reactive power change amount, is the weight matrix of reactive power, and there are: , , where, is the wind turbine 's voltage change amount at time k, is the wind turbine 's reference value of voltage change amount at time k, is the wind turbine 's reactive power change amount at time k, is the wind turbine 's reference value of reactive power change amount at time k, and are the initial voltage measurement value and reactive power measurement value of the wind turbine at the initial time , and are the voltage and reactive power reference values of the wind turbine .

[0031] As shown Figure 2 in the figure, under the first time scale, the decentralized voltage control of the wind farm to improve the reactive power support capacity of the wind turbines and ensure that the node voltage of the wind farm is maintained within a safe range includes: S101, collecting the maximum reactive power capacity, reactive power measurement value, and terminal voltage of all wind turbines; S102, establishing the objective function for the decentralized voltage control of the wind farm: , where, is the objective function, is the number of wind turbines, is the wind turbine voltage change, is the weight matrix of voltage, is the wind turbine reactive power change, is the weight matrix of reactive power; S103, taking , the gradient of the objective function can be calculated according to the following formula: , where, and are the gradients of the objective function at time , and are the reactive power and reactive power change at time , is the voltage change at time , is the transformation matrix of voltage and reactive power, is the total transformation matrix (non-singular matrix); S104, diagonalize the matrix of the voltage and reactive power relationship expression in the gradient of the objective function according to the following formula: , , , , where, and are the reactive powers of the wind turbine at times k + 1 and k respectively, and are intermediate variables, is the step coefficient, is the weight coefficient, , , is the wind turbine Decoupling coefficient with other wind turbines , is the reactive current change of the wind turbine at time k, and the superscript is the time interval, is the inertia time constant of the outer power loop of the wind turbine, is the amplitude of the grid phase voltage, and are the th diagonal element of the matrix is the inverse of the scaling matrix , is the total transformation matrix, and the scaling matrix is composed of the scaling coefficients of each wind turbine; S105, introduce the scaling matrix to construct the decentralized voltage control problem shown in the following formula to accelerate the convergence speed: , where, is the scaling matrix with respect to the wind turbine scaling coefficient inverse; the scaling matrix can be expressed as: , where, ~ are the scaling coefficients of the 1st to Nth wind turbines respectively; and there is: , where, is the reactive power after scaling ; Define to represent the projection onto the constraint set : , where, and are the upper and lower limits of reactive power respectively.

[0032] Thus, the variable iteration equation considering the dynamic characteristics of the decentralized voltage control problem can be obtained: , where, is the step size coefficient of the wind turbine , is the weight coefficient of the wind turbine ; S106. Introduce the updated step size coefficient and weight coefficient of each fan based on the decentralized voltage control problem, and then calculate the reactive power of each fan according to the following formula at the (k + 1)-th moment as the reactive power reference value : , where is the step size coefficient of fan , is the weight coefficient of fan . Through the above method, a generalized prediction (GP) method considering the dynamic response characteristics of wind turbines is constructed, which can quickly adjust the voltage and then improve the control efficiency of wind turbines. The above formula shows that in each step of the iteration, according to the local feedback information of the i-th fan, including the reactive power and the current change , the change trend of the global gradient can be reflected. In this way, the objective function can gradually approach a solution close to the global optimum, and only the local measurement data of each wind turbine is needed to achieve this process.

[0033] To ensure excellent operating performance, the voltage control of the wind farm is given top priority. This means that the weight W U needs to be greater than the weight W q , which results in relatively weak control performance of the reactive power capacity. To enhance the voltage margin, the VGP method designed in this embodiment can achieve effective voltage regulation while ensuring the maximum voltage margin. Specifically, the gain coefficient in the decentralized voltage control problem should match the reactive power support ability of the wind turbines. To achieve the maximum reactive power capacity of each wind turbine, the functional expressions for updating the step size coefficient and weight coefficient of each fan in step S106 of this embodiment are as follows: , , , where is an intermediate variable is the step size coefficient of fan , is the weight coefficient of fan , is the number of fans in the wind farm is the step size coefficient is the weight coefficient is the maximum reactive power capacity of fan .

[0034] In this embodiment, on the second-level time scale, each wind turbine is equipped with a variable gradient projection (VGP) controller, which has a two-layer control structure. At the upper control level, a variable coefficient calculation controller is responsible for collecting the maximum reactive power capacity information of each wind turbine, calculating the corresponding variable coefficients, and then transmitting these coefficients to the optimization controller at the lower level. The optimization controller calculates the optimal reactive power reference value using the VGP method based on the local reactive power measurement and the terminal voltage of the wind turbine. Finally, the optimization controller updates the local reactive power reference value according to the local measurement and the variable coefficients transmitted from the upper layer. This strategy not only significantly reduces the fluctuations of the bus voltage but also maximizes the voltage margin of each wind turbine in the wind farm. Each wind turbine can achieve a performance coordination effect close to the global optimal solution only through local measurements, ensuring the efficiency and stability of the overall operation of the wind farm. The root causes of the voltage fluctuations in the wind farm mainly lie in two major factors: one is the random change of wind resources, resulting in the intermittent and fluctuating output power of wind power; the other is external voltage disturbance events, such as grid faults or sudden changes in load. These factors act together, making the voltage fluctuation problem inside the wind farm extremely complex. In response to this challenge, the method of this embodiment proposes a decentralized voltage fluctuation control strategy for the second-level time scale based on VGP technology. This strategy uses the VGP method to control the voltage of the wind farm, realizing a fully decentralized control architecture. Its core is to adjust the terminal bus voltage within the feasible range while fully considering the dynamic response characteristics of the wind turbine and the grid side converter (GSC). In addition, considering the uneven distribution of the available power of each wind turbine in the wind farm, an adaptive variable gain strategy is designed for the gradient projection method. This strategy can be adjusted in real time to ensure that during the operation of the wind farm, each local wind turbine can make full use of its MRPC (multi-model predictive control), thereby optimizing the overall performance. By using the above variable gradient projection technology, a distributed optimization control strategy for the new energy wind farm connected to the urban distribution network on a short-term time scale has been successfully realized. This strategy effectively reduces the volatility of the bus voltage, maximally explores the reactive power output potential of the wind turbines, and at the same time ensures that the strategy can flexibly adapt to the rapid fluctuation characteristics of wind power. Under the precise control of the variable gradient projection method, the reactive power support ability of the wind turbines is fully and efficiently utilized within the safe operation boundary. Considering the dynamic characteristics of the wind turbines, the proposed method significantly improves the dynamic response efficiency of voltage regulation, ensuring the immediate stability of the power system. In addition, this method also inherits the advantages of the gradient projection method, including fast calculation and convergence stability, thus maintaining the efficiency and reliability of the control strategy while ensuring the optimization of system performance.

[0035] In this embodiment, on the minute - scale time frame, a static var generator (SVG) is configured with a quadratic programming (QP) controller. This controller calculates the sensitivity coefficient of voltage to reactive power and collects information such as voltage, active power, and reactive power in the distribution network. Subsequently, with the goal of minimizing voltage deviation, the QP controller calculates the value of reactive power that the SVG needs to provide to the new power system, adjusting the voltage of the distribution network nodes within the normal operating range. For a large - scale distribution network system with new energy generation, in addition to pursuing effective control of wind farm voltage fluctuations in a short time, it is also necessary to comprehensively dispatch reactive power compensation equipment in the distribution network to ensure that the distribution network can maintain voltage stability within a preset normal range under the condition of fluctuating new energy output. The voltage regulation strategy on the minute - scale time frame designs and implements a voltage control method based on sensitivity analysis for a radial distribution network with new energy generation access. When the voltage of the distribution network fluctuates, this method dynamically compensates the required reactive power by precisely regulating the output of the SVG, thus ensuring that the voltage level of the distribution network always remains within the specified normal operating range.

[0036] Assume that the distribution network has N nodes (including N S slack nodes and N PQ PQ nodes), and A and B represent the sets of slack nodes and PQ nodes respectively, that is and . The following definitions are made for all nodes of the distribution network, , , where, and are the voltage and voltage phase angle respectively, , and are the apparent power, active power, and reactive power respectively. The relationship between the node voltage and power injection is as follows: , where and represent the conjugates of and respectively; represents the admittance matrix.

[0037] In order to derive the sensitivity of the voltage (voltage magnitude) and voltage phase angle related to power injection, it is necessary to calculate the partial derivatives of the voltage with respect to the active power and reactive power of the node, and they satisfy the following formulas: , , In these two formulas, the first formula is with respect to and is linear. The second formula is relative to and is linear. These two formulas have a unique solution for the radial network. Once , , , are obtained, the voltage magnitude and phase angle sensitivities can be calculated. In this paper, SVG reactive power compensation is set for the distribution network in the medium time scale, and only the sensitivity relationship between the voltage magnitude and the reactive power is considered: , The problem to be solved by voltage control is to reduce the voltage fluctuations of all nodes and maintain the voltage of each node within the normal operating range. Therefore, the objective function is proposed as: , where is the objective function of reactive power compensation, is the number of nodes in the distribution network, is the set of predicted voltage deviation values. Specifically, in this embodiment, under the second time scale, reactive power compensation for the distribution network to stabilize the node voltage includes: S201, collecting the voltage and reactive power of all nodes in the distribution network; S202, calculating the sensitivity of the node voltage to power according to the voltage of the node, solving with the minimization of the voltage deviation as the optimization goal to obtain the reactive power output when the objective function is minimized, and performing reactive power compensation on the distribution network by using a static var generator SVG according to the obtained reactive power output; and the functional expression of the objective function is: , where is the objective function of reactive power compensation, is the number of nodes in the distribution network, is the set of predicted voltage deviation values, and there is: , where ~ are the predicted voltage deviation values of the 1st to nth busbars at moment respectively, and the calculation functional expression of the predicted voltage deviation value of any ith busbar at the kth moment is: , where is the voltage of the th busbar at the initial moment , is the The voltage of the is the reactive power output, is the th bus voltage sensitivity to reactive power . is the reference voltage. By solving the optimal value of this objective function, the reactive power compensated by the SVG to maintain the voltage stability of the distribution network can be obtained.

[0038] In this embodiment, the original topology of the distribution network consists of 1 slack bus and N PQ buses . The functional expression for calculating the sensitivity of the node voltage to power according to the node voltage is: , , , , where is an indicator function and the indicator function is 1 when and 0 otherwise; represents taking the negative after multiplying the indicator function by the imaginary unit j, is the conjugate of the voltage of the th bus to the active power of the th bus, is the union of the slack bus and the PQ buses , is the element in the th row and th column of the nodal admittance matrix, is the complex form of the voltage of the th bus, is the conjugate of the voltage of the th bus to the reactive power of the th bus, is the sensitivity of the voltage of the th bus to the active power of the th bus, is the sensitivity of the phase angle of the th bus to the active power of the For the reactive power of the th busbar, where is the phase angle of the th busbar, the sensitivity of the reactive power of the th busbar is

[0039] In this embodiment, the reactive power output by the SVG is controlled based on the analysis sensitivity method on a minute - level time scale, the node voltage of the distribution network is adjusted, the power flow distribution is improved, and the overall operation stability and reliability of the distribution network are enhanced, which is beneficial to the subsequent operation of the distribution network topology reconstruction optimization. To optimize the large - scale distribution network structure and improve the economic performance, a fast topology reconstruction method considering new - energy access is proposed on the third time scale in this embodiment. This method first collects the reactive - power control information of wind turbines and SVG at the second - level and minute - level, and then based on this information, a two - stage sensitivity - based switch - swapping method is adopted. An integer programming model is established during each switch - swapping process, and the branch - and - bound method is used to quickly solve it to rapidly determine the optimal topology reconstruction scheme. On the hour - level time scale, a topology reconstruction controller is globally deployed in the entire distribution network, and a two - stage reconstruction strategy is adopted. When the load or power injection conditions in the distribution network change, the controller systematically collects parameters such as network topology, voltage, current, and power. Then, using the sequential switch - opening and closing strategy, the controller calculates and determines the optimal state of the switches in the distribution network to minimize the network loss. This process effectively solves the distribution network reconstruction problem in a relatively short time and ensures the efficient and stable operation of the distribution network. Specifically, in this embodiment, on the third time scale, the distribution network topology reconstruction to minimize the network loss includes: S301, collecting the distribution network topology structure, voltage, active power, and reactive power; S302, calculating and determining the optimal switch - state configuration of the distribution network with the objective of minimizing the objective function of network loss and performing distribution network topology reconstruction. The function expression of the objective function of network loss is: , where is the objective function of network loss, is the number of nodes in the distribution network, and are the voltages of the th and th nodes after the distribution network topology reconstruction respectively, is to take the real part, is the node admittance matrix after the distribution network topology reconstruction, the th and The voltage phase angle difference of the nodes, and there is: , Wherein, and are respectively the voltage phase angles of the th and the th nodes after the distribution network topology reconstruction.

[0040] In this embodiment, the calculation function expressions of the voltage and voltage phase angle of the nodes after the distribution network topology reconstruction are: , , Wherein, is the voltage set after the distribution network topology reconstruction, ~ are respectively the voltages of the 1st to nth nodes after the distribution network topology reconstruction, , is the voltage phase angle set after the distribution network topology reconstruction, ~ are respectively the voltage phase angles of the 1st to nth nodes after the distribution network topology reconstruction, is the initial value of the voltage of the node; is the binary variable indicating whether the branch acts, 1 for acting and 0 for not acting; is the equivalent active power change amount injected at the two ends of the branch when the branch acts, is the equivalent reactive power change amount injected at the two ends of the branch when the branch acts; is the initial value of the voltage phase angle of the node, is the voltage of the node to the sensitivity of the active power , is the voltage of the node to the sensitivity of the reactive power , is the voltage phase angle of the node to the sensitivity of the active power , is the voltage phase angle of the node to the sensitivity of the reactive power , is the branch set of the distribution network.

[0041] Figure 3The in-loop operation model in the embodiment, where (a) is the network after the switch in-loop operation, Close represents the in-loop state, (b) is the network before the switch in-loop operation, Open represents the open-loop state, and (c) is the compensation network, where and are the voltages of node T before and after the closed-loop operation respectively. is the voltage difference of node T before and after the closed-loop operation. is the impedance between node i and node T. is node and the impedance between node T. and are the nodes before and after the closed-loop operation respectively voltage. and are the nodes before and after the closed-loop operation respectively voltage. is the node before and after the closed-loop operation voltage difference. is the node before and after the closed-loop operation voltage difference. is the compensation current. is the impedance of the switch action branch. is the voltage difference across the switch; according to the superposition theorem, the voltage and current in the network after the switch in-loop operation consist of two parts. One part is the voltage and current before the operation, as shown in Figure 3 (b) in; the other part is the voltage and current generated by the power source, as shown in Figure 3 (c) in. To ensure that the two networks have the same port voltage-current characteristics, the equivalent voltage source should be equal to the voltage difference across the switch: , where is the voltage difference across the switch, and are the voltages across the switch; any node on the loop except node i and can be designated as node T; how the active and reactive power changes at any nodes and affect the voltages of nodes , and T can be expressed as: , , , where and are nodes and The change in active power and are the changes in reactive power at nodes and respectively. is the apparent power at node , and is the imaginary part of the apparent power at node . The open-loop operation is equivalent to the inverse process of the closed-loop operation, which will not be elaborated here.

[0042] To verify the multi-time scale voltage control and reconfiguration method of the novel power system in this embodiment, a 33-node distribution network containing a wind farm (WF) and a static var generator (SVG) for reactive power compensation is adopted in this embodiment. During the 24-hour period, considering the dynamic changes of network load and power injection comprehensively, the control strategy planning and implementation of the novel power system (NPS) are fully considered. Specifically, the wind farm performs variable gradient projection (VGP) voltage control once per second to ensure the voltage stability of wind turbines. The SVG is used to regulate the voltage of the distribution network once every 15 minutes to keep the voltage within the normal range. In addition, the topology of the distribution network is optimized every hour according to its real-time operating status. It should be noted that the SVG is connected to the 11th node of the distribution network, and the wind farm is connected to the 26th node of the distribution network. The initial switch configuration is set to 32 sectional switches (index numbers from 1 to 32) and 5 tie switches (index numbers from 33 to 37). The effectiveness and practicality of the proposed fast multi-time scale voltage control and topology reconfiguration method are verified by the distribution network. All simulations are implemented on MATLAB R2019a equipped with an Intel Core i7-12700F 2.10 GHz CPU and 32GB RAM processor.

[0043] On the second time scale, considering the fast fluctuation characteristics of wind power, the variable gradient projection control strategy is adopted in this paper to ensure that the voltage of wind turbines is maintained within the normal range and to maximize the voltage margin of each unit in the wind farm. This strategy is implemented by a decentralized control method, effectively avoiding the negative effects that may be caused by excessive communication load and communication delay, and significantly enhancing the reactive power support ability of the wind farm to the power grid. Table 1 shows the effect comparison between the method in this embodiment (VGP method) and the existing centralized model predictive control (CMPC) method for wind farms.

[0044] Table 1: Effect comparison between the method in this embodiment (VGP method) and the existing CMPC method

[0045] As can be seen from Table 1, the method of this embodiment (VGP method) can achieve control performance close to the global optimum only through local measurement. Compared with the CMPC method, the method of this embodiment (VGP method) can also achieve precise regulation of the voltage within the range of 0.999 to 1.001. However, it should be noted that the local controller of the CMPC method takes more than four times as long to calculate as the method of this embodiment (VGP method), and also requires more central controller computing resources. Each local controller of the method of this embodiment (VGP method) only needs to execute a simple calculation process in each control cycle, and can achieve rapid dynamic voltage regulation based on the instant local data and feedback.

[0046] Figure 4 Plotted are the 24-hour curves of the load demand and wind power output for a typical summer day used in this study. On the minute-level time scale, considering the possible dynamic changes in the power generation output and load in the distribution network, sensitivity analysis technology is used to determine the optimal reactive power compensation amount that the SVG should provide to the distribution network when the voltage fluctuates. This distribution network voltage control strategy can quickly respond to voltage fluctuations or disturbances, ensuring that the distribution network operates in a safe, continuous, and stable state. Figure 5 Shows the amount of reactive power that the SVG needs to inject into the distribution network every 15 minutes obtained by applying sensitivity analysis. In the time period from the 0th to the 4th hour, due to the decrease in the system load demand, the reactive power output by the SVG decreases accordingly and reaches its minimum value of 0.0441 MVar at the 5th hour. After that, as the load demand gradually increases, the reactive power output by the SVG also rises accordingly until it reaches its maximum value of 0.0795 MVar at the 20th hour. Figure 6 Compares the voltage deviations of all nodes in the distribution network without voltage and topology reconstruction control with those after the distribution network uses the strategy proposed in the text for voltage control and topology reconstruction and their curves over time. The results show that the distribution network voltage control and topology reconstruction strategy combining the comprehensive sensitivity analysis method and the switch exchange method shows a relatively significant effect. The strategy proposed in this paper effectively limits the overall voltage deviation of the system within the range of 0.0008 to 0.0025. Specifically, at the 20th hour, the optimized voltage deviation reaches its maximum value, but this value is significantly smaller than the deviation level without implementing the voltage control strategy.

[0047] On the hourly time scale, to achieve the goal of minimizing the power grid loss of the distribution network and improve its economic performance, a two-stage topology reconstruction strategy is adopted in this paper to optimize the switch configuration in the distribution network. By simple current calculation, the switch adjustment position can be quickly determined, thus avoiding the cumbersome and complex power flow calculation steps. This strategy not only significantly shortens the calculation cycle but also effectively improves the efficiency of the optimization process. In this embodiment, tests are carried out. Table 2 details the effect of implementing this two-stage topology reconstruction strategy on the switch planning of the distribution network per hour. Table 3 compares the planning effect of the method in this embodiment and the mathematical programming method (MP algorithm) on the distribution network at the 12th hour.

[0048] Table 2: Effect of the two-stage reconstruction method on the switch planning of the distribution network every 1h

[0049] Table 3: Planning effect of the method in this embodiment and the mathematical programming method on the distribution network at the 12th hour

[0050] The results of Table 2 and Table 3 show that the two-stage reconstruction strategy proposed by the method in this embodiment can quickly adjust the switch position, optimize the structure of the distribution network, and reduce the active power grid loss. The mathematical programming method does not consider the reactive power output of the static var generator (SVG) during the optimization process, while the method in this embodiment clearly takes into account the reactive power output of the SVG. Therefore, it shows a more significant effect in terms of network loss optimization. From the branches disconnected from the distribution network, it can be seen that although the two methods show consistency in the results, the method in this embodiment not only effectively reduces the network loss but also significantly shortens the time-consuming of the calculation process.

[0051] In addition, this embodiment also provides a multi-time scale voltage control and reconstruction system for a new type of power system, including a microprocessor and a memory connected to each other. The microprocessor is programmed or configured to execute the multi-time scale voltage control and reconstruction method for the new type of power system.

[0052] In addition, this embodiment also provides a computer-readable storage medium, in which a computer program or instruction is stored. The computer program or instruction is programmed or configured to execute the multi-time scale voltage control and reconstruction method for the new type of power system through a processor.

[0053] In addition, this embodiment also provides a computer program product, including a computer program or instruction. The computer program or instruction is programmed or configured to execute the multi-time scale voltage control and reconstruction method for the new type of power system through a processor.

[0054] Those skilled in the art should understand that the technical solutions provided by the present invention can be in the form of a method, a system, or a computer program product. Therefore, the present invention can be implemented in the form of a complete hardware embodiment, a complete software embodiment, or an embodiment combining software and hardware aspects. Moreover, the present invention can be in the form of a computer program product implemented on one or more computer-readable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) that contain computer-usable program code. The present invention is described with reference to the flowcharts and / or block diagrams of methods, devices (systems), and computer program products according to the embodiments of the present invention. It should be understood that each process and / or block in the flowchart and / or block diagram, and the combination of processes and / or blocks in the flowchart and / or block diagram, can be realized by computer program instructions. These computer program instructions can be provided to the processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing devices to generate a machine, so that the instructions executed by the processor of the computer or other programmable data processing devices generate a device for realizing the functions specified in the process Figure 1 one process or multiple processes and / or blocks Figure 1 one block or multiple blocks. These computer program instructions can also be stored in a computer-readable memory that can direct a computer or other programmable data processing device to work in a specific manner, so that the instructions stored in the computer-readable memory generate a manufactured article including an instruction device, and the instruction device realizes the functions specified in the process Figure 1 one process or multiple processes and / or blocks Figure 1 one block or multiple blocks. These computer program instructions can also be loaded onto a computer or other programmable data processing device, so that a series of operation steps are executed on the computer or other programmable device to generate a computer-implemented process. Therefore, the instructions executed on the computer or other programmable device provide steps for realizing the functions specified in the process Figure 1 one process or multiple processes and / or blocks Figure 1 one block or multiple blocks.

[0055] The above is only the preferred embodiment of the present invention, and the protection scope of the present invention is not limited to the above embodiments. All technical solutions falling within the idea of the present invention belong to the protection scope of the present invention. It should be noted that for those of ordinary skill in the art in this technical field, several improvements and refinements made without departing from the principle of the present invention should also be regarded as within the protection scope of the present invention.

Claims

1. A novel multi-time scale voltage control and reconstruction method for power system, characterized in that: The method comprises the following steps: at a first time scale, performing distributed voltage control on the wind farm to improve the reactive power support capability of the wind turbine and ensure that the node voltage of the wind farm is maintained within a safe range; at a second time scale, performing reactive power compensation on the distribution network to stabilize the node voltage; at a third time scale, performing distribution network topology reconstruction on the distribution network to minimize network loss; the time periods of the first time scale, the second time scale and the third time scale are increased in sequence.

2. The multi-time scale voltage control and reconstruction method of the novel power system according to claim 1 is characterized in that: Under the first time scale, the distributed voltage control of the wind farm to improve the reactive power support capacity of the wind turbine and ensure that the node voltage of the wind farm is maintained within a safe range includes: S101, collecting the maximum reactive power capacity, reactive power measurement value and terminal voltage of all wind turbines; S102, establish the objective function of the distributed voltage control of the wind farm: ， in, is the objective function, is the number of fans, For fans The voltage change, is the voltage weight matrix, For fans The reactive power change, is the weight matrix of reactive power; S103, calculate the gradient of the objective function according to the following formula: ， in, and for The gradient of the objective function at time , and for Reactive power and reactive power change at the moment, for The voltage change at a given moment, is the transformation matrix of voltage and reactive power, is the total transformation matrix; S104, diagonalizing the matrix of the voltage and reactive power relationship expression in the gradient of the objective function according to the following formula: ， ， ， ， in, and Fan The reactive power at time k+1 and time k is, and is the intermediate variable, is the step size coefficient, is the weight coefficient, For fans and the decoupling coefficients of other fans, For fans The change in reactive current at time k, is the interval time, is the inertia time constant of the fan power outer loop, is the grid phase voltage amplitude, and For the matrix No. diagonal elements, is the scaling matrix The inverse, is the total transformation matrix, the scaling matrix It is composed of the scaling factors of each wind turbine; S105, introduce scaling matrix The decentralized voltage control problem is formulated as follows to speed up convergence: ， in, is the scaling matrix About Fan The inverse of the scaling factor of ; S106, based on the decentralized voltage control problem, introduce the updated step coefficient and weight coefficient of each wind turbine, so as to calculate the power consumption of each wind turbine according to the following formula: Reactive power at time k+1 As reactive power reference : ， in, For fans The step size coefficient, For fans The weight coefficient of .

3. The multi-time scale voltage control and reconstruction method of the novel power system according to claim 2 is characterized in that: The function expression for updating the step coefficient and weight coefficient of each wind turbine in step S106 is: ， ， ， in, is the intermediate variable, For fans The step size coefficient, For fans The weight coefficient of is the number of wind turbines in the wind farm, is the step size coefficient, is the weight coefficient, For fans Maximum reactive power capacity.

4. The multi-time scale voltage control and reconstruction method of the novel power system according to claim 3 is characterized in that: The performing reactive power compensation on the distribution network to stabilize the node voltage at the second time scale includes: S201, collecting voltage and reactive power of all nodes in the distribution network; S202, calculating the sensitivity of the node voltage to the power according to the node voltage, solving the problem with minimizing the voltage deviation as the optimization goal to obtain the reactive power output that minimizes the objective function, and using the static VAR generator SVG to perform reactive power compensation on the distribution network according to the reactive power output obtained by the solution; and the functional expression of the objective function is: ， in, is the objective function of reactive power compensation, is the number of nodes in the distribution network, is a set of voltage deviation prediction values, and has: ， in, ~ The 1st to nth buses are The voltage deviation prediction value at the moment, and the calculation function expression of the voltage deviation prediction value of any i-th bus at the moment k is: ， in, For the The busbar at the initial moment The voltage, For the The voltage of the busbar, is the reactive power output, For the The voltage of the busbars to the reactive power The sensitivity of is the reference voltage.

5. The multi-time scale voltage control and reconstruction method of the novel power system according to claim 4 is characterized in that: The original topology of the distribution network consists of 1 slack bus and N PQ buses Composition; the function expression for calculating the sensitivity of the node voltage to the power according to the node voltage is: ， ， ， ， in, is the indicator function and the indicator function is 1 when it is, otherwise 0; It means that the indicator function is multiplied by the imaginary unit j and then becomes negative. is the conjugate of the voltage of the ith bus For Active power of busbars The sensitivity of To relax the busbar and PQ bus The union of is the node admittance matrix Line Column elements, For the The complex form of the voltage of the bus, is the conjugate of the voltage of the ith bus For Reactive power of busbars The sensitivity of is the voltage of the ith bus For Active power of busbars The sensitivity of To obtain the real part, To take the imaginary part, is the phase angle of the ith busbar For Active power of busbars The sensitivity of is the voltage of the ith bus For Reactive power of busbars The sensitivity of is the phase angle of the ith busbar For Reactive power of busbars The sensitivity of .

6. The multi-time scale voltage control and reconstruction method of the novel power system according to claim 5 is characterized in that: The step of reconfiguring the distribution network topology to minimize network loss at the third time scale includes: S301, collecting the topology, voltage, active power and reactive power of the distribution network; S302, with the objective function of minimizing network loss as the goal, calculate and determine the optimal switch state configuration of the distribution network and reconstruct the distribution network topology. The function expression of the objective function of network loss is: ， in, is the objective function of network loss, is the number of nodes in the distribution network, and They are the first and The voltage of the node, To obtain the real part, is the node admittance matrix after the distribution network topology is reconstructed, After the distribution network topology is reconfigured and The voltage phase angle difference of the nodes is: ， in, and They are the first and The voltage phase angle of each node.

7. The multi-time scale voltage control and reconstruction method of the novel power system according to claim 6 is characterized in that: The calculation function expressions of the node voltage and voltage phase angle after the distribution network topology is reconfigured are: ， ， in, is the voltage set after the distribution network topology is reconstructed, ~ are the voltages of the 1st to nth nodes after the distribution network topology is reconfigured, , is the voltage phase angle set after the distribution network topology is reconstructed, ~ are the voltage phase angles of the 1st to nth nodes after the distribution network topology is reconfigured, is the initial value of the node voltage; For branch i A binary variable for action or not, 1 for action and 0 for no action; For branch i The equivalent active power change injected into the nodes at both ends of branch i during operation, For branch i When in action, on branch i The equivalent reactive power change injected into the two end nodes; is the initial value of the node voltage phase angle, The voltage of the node For active power The sensitivity of The voltage of the node Reactive power The sensitivity of is the voltage phase angle of the node For active power The sensitivity of is the voltage phase angle of the node Reactive power The sensitivity of It is a collection of branches of the distribution network.

8. A novel multi-time scale voltage control and reconstruction system for a power system, comprising interconnected microprocessors and memories, characterized in that: The microprocessor is programmed or configured to execute the multi-time scale voltage control and reconstruction method of the novel power system as claimed in any one of claims 1 to 7.

9. A computer-readable storage medium having a computer program or instruction stored therein, characterized in that: The computer program or instruction is programmed or configured to execute the multi-time-scale voltage control and reconstruction method of the novel power system described in any one of claims 1 to 7 through a processor.

10. A computer program product comprising a computer program or instructions, characterized in that The computer program or instruction is programmed or configured to execute the multi-time-scale voltage control and reconstruction method of the novel power system described in any one of claims 1 to 7 through a processor.

Images