A power generation coordination control method and system of a new energy power system

Abstract

The application provides a power generation coordination control method and system of a new energy power system, and belongs to the technical field of power system operation and control. A node voltage deviation gradient and an inverter adjustable reactive power space are synchronously acquired, a voltage disturbance vector and a gain matrix are generated through weighted mapping and normalization. Based on the Lagrange dual principle, optimal reactive power distribution weights of each node are solved with the minimum network loss as the target. Then the weights are converted into current reference values, the response speed is matched through first-order inertia filtering, the current error is compensated through a proportional resonant algorithm, and finally PWM signals are generated to drive the inverter to act in coordination, so as to realize power distribution network voltage stability and power generation coordination. The application realizes consistent adjustment pace of multiple point distributed power sources, effectively suppresses voltage interaction feedback oscillation and reduces the total network loss.

Description

Technical Field

[0001] This application belongs to the field of power system operation and control technology, and in particular relates to a power generation coordination control method and system for a new energy power system. Background Technology

[0002] With the high penetration rate of distributed photovoltaic (PV) systems, distribution networks are gradually evolving into active networks with source-load interaction. Generation coordination and control methods are becoming crucial for maintaining grid voltage levels and ensuring power quality. These methods have broad application prospects in addressing voltage exceedance issues caused by renewable energy fluctuations and improving the operational stability of distribution networks.

[0003] Existing technologies typically employ voltage-reactive power droop control strategies based on local information, or utilize centralized management units to uniformly calculate and issue reactive power commands based on steady-state snapshots. These methods primarily attempt to maintain node voltage within a reasonable range by adjusting the reactive power output of grid-connected inverters.

[0004] However, the above methods often overlook the differences in voltage fluctuation gradients at different nodes and the inconsistency in the adjustment response speeds of each inverter, leading to potential misalignment during multi-point concurrent adjustment. This lack of dynamic coordination in adjustment easily causes mutual restraint and repeated oscillations in voltage between nodes, and makes it difficult to optimize global network losses. Therefore, existing technologies suffer from insufficient coordinated control of power generation due to mismatched multi-point adjustment responses. Summary of the Invention

[0005] The purpose of this application is to provide a method and system for coordinated control of power generation in a new energy power system, so as to solve the problem of insufficient coordinated control of power generation in the prior art.

[0006] To address the aforementioned technical problems, in a first aspect, this application provides a method for coordinated control of power generation in a new energy power system, comprising:

[0007] Simultaneously acquire voltage deviation gradient data of target load nodes in the distribution network and the adjustable reactive power space of inverters at each distributed power source grid connection point under output conditions;

[0008] By weighting and aggregating the voltage deviation gradient data and associated geographic location information, a voltage disturbance vector is obtained. Then, the rated capacity and reactive power space of each distributed power node are normalized and weighted to obtain a gain matrix.

[0009] Using voltage disturbance vector as guiding parameter, gain matrix as constraint coefficient, minimizing global network loss of distribution network as objective function, and voltage deviation of each target load node within safe range as constraint boundary, the reactive power allocation weight vector of each distributed power node is obtained by iterative solution through Lagrange duality.

[0010] The product of the weight vector and the per-unit voltage value of each distributed power node is used as the initial reference value of reactive current. The initial reference value is then filtered by a first-order inertial element to match the adjustment response speed of the inverter, thus obtaining the reference base value.

[0011] At the resonant pole of the grid fundamental frequency, the reference voltage vector is obtained by performing error integral compensation on the difference between the reference reference value and the real-time feedback current using quasi-proportional resonance. The reference voltage vector is then converted into a pulse sequence signal through pulse width modulation mapping to control the inverter for state switching.

[0012] Optionally, by performing weighted mapping and vector aggregation on the voltage deviation gradient data and associated geographic location information, a voltage disturbance vector is obtained, including:

[0013] Based on the geographical location information, determine the spatial coordinates of each target load node relative to the preset equilibrium point, and calculate the straight-line distance between each target load node and the preset equilibrium point to obtain the mapping weight of each target load node;

[0014] The disturbance intensity coefficient of each target load node is obtained by multiplying the voltage deviation gradient data with the corresponding mapping weight.

[0015] By using the disturbance intensity coefficient to vector scale the spatial coordinates, the disturbance vector components of each target load node are obtained. The voltage disturbance vector is obtained by superimposing and summing all the disturbance vector components.

[0016] Optionally, the rated capacity and reactive power space of each distributed power node are normalized and weighted to obtain a gain matrix, including:

[0017] Calculate the ratio of the rated capacity of each distributed power node to the total rated capacity of all distributed power nodes to obtain the power weight of each distributed power node.

[0018] Calculate the ratio of reactive power space to the corresponding rated capacity to obtain the adjustment gain coefficient of each distributed power node;

[0019] The power weight is calculated by multiplying it by the corresponding adjustment gain coefficient to obtain the adjustment coefficient of each distributed power node. All adjustment coefficients are then mapped to a preset multidimensional diagonal matrix according to the position index of each distributed power node to obtain the gain matrix.

[0020] Optionally, using the voltage disturbance vector as the guiding parameter, the gain matrix as the constraint coefficient, minimizing the global network loss of the distribution network as the objective function, and ensuring that the voltage deviation of each target load node is within a safe range as the constraint boundary, the reactive power allocation weight vector of each distributed generation node is obtained through iterative solution using Lagrange duality, including:

[0021] The reactive power adjustment direction of each target load node is determined based on the guiding parameters, and the objective function and constraint boundary are correlated and mapped using the reactive power adjustment direction and constraint coefficients to obtain the mapping correlation matrix;

[0022] The augmented dual function is obtained by weighting and superimposing the mapping incidence matrix using preset multiplier adjustment parameters and penalty adjustment parameters.

[0023] By using progressive optimization to perform gradient iteration on the augmented dual function, the weight vectors of each distributed power source node are obtained.

[0024] Optionally, gradient iteration of the augmented dual function is performed using progressive optimization to obtain the weight vector of each distributed power node, including:

[0025] By utilizing the augmented dual function in the search guidance vector of each distributed power node, a step-by-step recursive mapping is performed on the preset initial weight distribution to obtain an optimization path sequence composed of multiple candidate solution vectors.

[0026] Calculate the magnitude of the difference between two adjacent candidate solution vectors in the optimization path sequence to obtain the corresponding deviation value;

[0027] Candidate solution vectors whose deviation values ​​are within the preset steady-state threshold range are determined as convergence target solutions. Numerical components corresponding to each distributed power node are extracted from the convergence target solutions and numerically mapped to obtain the weight vector of each distributed power node.

[0028] Optionally, the product of the weight vector and the per-unit voltage value of each distributed power generation node is used as the initial reference value of the reactive current. A first-order inertial filter is applied to the initial reference value to match the inverter's regulation response speed, resulting in a reference base value, including:

[0029] The ratio of the weight vector to the corresponding voltage per-unit value is calculated to obtain the initial reference value of the reactive current of each distributed power generation node.

[0030] The amplitude change rate of the initial reference value is limited by the preset time constant parameter in the first-order inertial element, and the current transition component is obtained.

[0031] A reference value is obtained by sampling and mapping the current transient component in the time domain.

[0032] Optionally, at the resonant pole of the grid fundamental frequency, the reference voltage vector is obtained by performing error integration compensation on the difference between the reference reference value and the real-time feedback current using quasi-proportional resonance, including:

[0033] The difference between the reference value and the real-time feedback current is calculated to obtain the current deviation component;

[0034] By accumulating the time-domain values ​​of the current deviation component corresponding to the fundamental frequency of the power grid using the resonant poles, the resonant compensation vector is obtained.

[0035] The current deviation component is linearly scaled to obtain the proportional feedback component, and the proportional feedback component and the resonant compensation vector are superimposed and summed to obtain the reference voltage vector.

[0036] Secondly, this application provides a power generation coordination and control system for a new energy power system, comprising:

[0037] The acquisition module is used to synchronously acquire the voltage deviation gradient data of the target load node in the distribution network and the adjustable reactive power space of the inverter at each distributed power source grid connection point under the output operating condition.

[0038] The generation module is used to obtain the voltage disturbance vector by weighting and vector aggregation of voltage deviation gradient data and associated geographical location information, and to normalize and weight the rated capacity and reactive power space of each distributed power node to obtain the gain matrix.

[0039] The solution module is used to obtain the reactive power allocation weight vector of each distributed power generation node by using the voltage disturbance vector as the guiding parameter, the gain matrix as the constraint coefficient, the minimization of the global network loss of the distribution network as the objective function, and the voltage deviation of each target load node within a safe range as the constraint boundary, through Lagrange duality iterative solution.

[0040] The generation module is also used to take the product of the weight vector and the per-unit voltage value of each distributed power node as the initial reference value of reactive current, and to use a first-order inertial element to perform first-order inertial filtering on the initial reference value to match the adjustment response speed of the inverter and obtain the reference base value.

[0041] The control module is used to find the resonant pole at the fundamental frequency of the power grid. It uses quasi-proportional resonance to perform error integral compensation on the difference between the reference value and the real-time feedback current to obtain the reference voltage vector. The reference voltage vector is then converted into a pulse sequence signal through pulse width modulation mapping to control the inverter to switch states.

[0042] Thirdly, this application provides an electronic device, comprising:

[0043] Memory, used to store computer programs;

[0044] A processor is used to execute the computer program to implement the steps of the power generation coordination control method for a new energy power system as described in the first aspect above.

[0045] Fourthly, this application provides a computer-readable storage medium storing a computer program that, when executed by a processor, can implement the steps of the power generation coordination control method for a new energy power system as described in the first aspect above.

[0046] The generation coordination control method for new energy power systems provided in this application first aggregates the voltage deviation gradient and geographical location into a voltage disturbance vector, and combines it with the gain matrix of the actual capacity of the inverter. Then, it uses the Lagrange duality principle to calculate the optimal reactive power allocation weight while pursuing the minimization of global network loss, effectively solving the problem of lack of global vision in single-point independent regulation.

[0047] Subsequently, a first-order inertial element is used to match the response speed of the weighted reference value, smoothing the adjustment pace between different nodes and avoiding voltage mutual restraint and repeated oscillations caused by response differences. Finally, precise integral compensation for feedback errors is achieved through quasi-proportional resonant control, ensuring synchronous coordination and stable operation of each distributed power source at the physical execution level. Therefore, this application effectively overcomes the problems of insufficient power generation coordination control and high network losses caused by multi-point adjustment response mismatch in the prior art.

[0048] Furthermore, this application first establishes the spatial coordinates of each load node based on its geographical location and calculates the distance weights relative to a preset equilibrium point, thereby achieving a quantitative representation of the spatial distribution characteristics of different nodes in the distribution network and compensating for the inability of a single electrical quantity to reflect spatial differences. Subsequently, by combining voltage deviation gradient data with this mapping weight to generate a disturbance intensity coefficient, and performing vector scaling and aggregation on the spatial coordinates, it is possible to accurately map the dispersed and fluctuating local voltage states of the entire network into a dominant voltage disturbance vector.

[0049] This vector, serving as the guiding core for global regulation, can effectively identify the center of gravity region of voltage instability, thereby guiding distributed power sources to work together in the most needed direction and avoiding regulation direction conflicts caused by local perspectives. Therefore, this application effectively improves the accuracy of global coordinated control for multi-point voltage fluctuations by constructing a vector model that integrates spatial and electrical aspects. Attached Figure Description

[0050] To more clearly illustrate the technical solutions of the embodiments of this application or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0051] Figure 1 A flowchart illustrating a power generation coordination control method for a new energy power system provided in this application embodiment;

[0052] Figure 2 A flowchart illustrating a method for generating weight vectors provided in an embodiment of this application;

[0053] Figure 3 A flowchart illustrating a method for generating a reference base value provided in an embodiment of this application;

[0054] Figure 4 A schematic diagram of the structure of a power generation coordination control system for a new energy power system provided in this application embodiment;

[0055] Figure 5 This is a schematic diagram of the hardware structure of an electronic device provided in one embodiment of this application. Detailed Implementation

[0056] To address the voltage oscillations and high network losses caused by inconsistent regulation of distributed power sources in active distribution networks, existing technologies often rely on local droop control or static centralized command issuance, which often ignores the spatial differences in voltage deviation gradients between different load nodes and the temporal differences in response speeds of various inverters.

[0057] This single regulation mode, which lacks global dynamic coordination, makes it easy for each access point to generate a seesaw effect of mutual constraint when dealing with voltage fluctuations. That is, regulation in one place triggers a reverse fluctuation in another place, resulting in repeated voltage oscillations across the entire network and failing to effectively reduce line losses. This makes it difficult to meet the dual requirements of stability and economy for high-penetration new energy distribution networks.

[0058] To address the aforementioned challenges, this application proposes a power generation coordination control method based on Lagrange duality optimization and dynamic weight coordination. Its core lies in vectorizing and aggregating the spatial gradient of voltage deviation with the inverter's regulation potential to construct guiding parameters reflecting global disturbance characteristics. Specifically, this method utilizes the Lagrange duality principle to calculate the optimal reactive power allocation weights while minimizing network losses. It then filters the commands using a first-order inertial element to match the physical response speed of each inverter, and, in conjunction with quasi-proportional resonant control, achieves precise compensation for current errors.

[0059] This scheme abandons the fragmented regulation approach and, through a strategy of global optimization allocation and dynamic speed matching, ensures that all distributed power sources can respond in a consistent pace and phase, fundamentally eliminating the risk of voltage oscillation caused by multi-point regulation and achieving simultaneous improvement in distribution network voltage stability and operational energy efficiency.

[0060] To enable those skilled in the art to better understand the present application, the present application will be further described in detail below with reference to the accompanying drawings and specific embodiments. Obviously, the described embodiments are merely some embodiments of the present application, and not all embodiments. Based on the embodiments in this application, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this application.

[0061] To address the problems of existing technologies, embodiments of this application provide a method, apparatus, equipment, computer storage medium, and computer program product for coordinated control of power generation in a new energy power system. The method for coordinated control of power generation in a new energy power system provided in this application embodiment will be described first below.

[0062] Figure 1 A flowchart illustrating a power generation coordination control method for a new energy power system according to an embodiment of this application is shown. Figure 1 As shown, the method includes:

[0063] S101. Synchronously acquire voltage deviation gradient data of target load nodes in the distribution network and the adjustable reactive power space of inverters at each distributed power source grid connection point under output conditions.

[0064] Target load nodes refer to critical monitoring buses within the distribution network area that are sensitive to voltage fluctuations or located at the end of feeders; their voltage quality directly reflects the stability of the regional power grid. Voltage deviation gradient data refers to the rate of change of the difference between the real-time effective voltage value and the rated reference value of these nodes over time, indicating the urgency and trend of voltage deviation from steady state.

[0065] A distributed generation grid-connected inverter refers to a power electronic interface device that connects photovoltaic and other new energy generator sets to the power grid. Output condition refers to the active power output status of the unit at the current moment. Adjustable reactive power space refers to the upper and lower limits of the dynamic range within which the inverter can generate or absorb reactive power, provided that the inverter's apparent power capacity constraint and current active power output are met.

[0066] In practice, the real-time voltage amplitude of each target load node is first collected synchronously using wide-area measurement terminals or smart meters deployed at each node of the distribution network. The difference between this voltage amplitude and the rated voltage is calculated, and the difference is then differentiated or differentially calculated to obtain a voltage deviation gradient data set. ,in Indicates the first The rate of change of voltage fluctuation at each node and .

[0067] At the same time, the rated capacity of each distributed power inverter is read in real time through the communication link. and current active power Based on the power circle constraint principle, using the formula... Calculate the reactive power output limit of each inverter to determine the adjustable reactive power space set. ,in Indicates the first The maximum inductive or capacitive reactive power margin that the inverter can currently provide and .

[0068] S102. By weighting and vector aggregating the voltage deviation gradient data and associated geographical location information, the voltage disturbance vector is obtained. Then, the rated capacity and reactive power space of each distributed power node are normalized and weighted to obtain the gain matrix.

[0069] Optionally, the process of obtaining the voltage disturbance vector by weighting and vector aggregation of the voltage deviation gradient data and associated geographic location information in step S102 may specifically include:

[0070] S1021. Determine the spatial coordinates of each target load node relative to the preset equilibrium point based on the geographical location information, and calculate the straight-line distance between each target load node and the preset equilibrium point to obtain the mapping weight of each target load node.

[0071] The preset balance point refers to the physical center or electrical reference point in the distribution network topology, typically selected from the substation busbar location or the geometric center of the power grid. Spatial coordinates refer to the position vectors of each target load node in a Cartesian or polar coordinate system with the preset balance point as the origin. Mapping weights are numerical factors assigned based on the physical distance between the node and the balance point, used to quantify the attenuation or enhancement of the spatial location's impact on voltage stability. The spatial locations of the preset balance point and target load nodes are shown in Table 1 below.

[0072] Table 1: Comparison of Spatial Locations of Preset Equilibrium Points and Target Load Nodes

[0073]

[0074] As shown in Table 1, Table 1 presents the spatial relationship data between the preset equilibrium point and some target load nodes. The table lists the equilibrium point O, which serves as the reference origin, and the geographical parameters of two typical target load nodes, A and B. The horizontal and vertical coordinates are in kilometers (km), and the relative distances are Euclidean distances calculated from the coordinates.

[0075] In practice, the coordinates of the preset equilibrium point are first determined based on the geographic information system (GIS) data of the power distribution network. And extract each target load node. absolute geographic coordinates Then, the relative spatial coordinates are calculated. Next, according to Table 1, the Euclidean distance formula is used. Calculate the straight-line distance between each node and the equilibrium point. .

[0076] Finally, based on the physical characteristics of far-end nodes having higher impedance and being more sensitive to voltage fluctuations, the straight-line distance... Perform proportional mapping to obtain the mapping weights for each target load node. For example, suppose node A in Table 1 is far from the equilibrium point. Node B is far from the equilibrium point The weight calculation logic is as follows: , ,in The scaling factor is a constant, thus constructing a set of mapping weights. .

[0077] S1022. Calculate the product of the voltage deviation gradient data and the corresponding mapping weight to obtain the disturbance intensity coefficient of each target load node.

[0078] The disturbance intensity coefficient is a scalar value that combines the voltage change rate over time with the geographical location weight in the spatial dimension to represent the degree of influence of a single node on the overall voltage stability of the power grid. The larger the coefficient, the more severe the voltage fluctuations at that node and the higher its regulation priority in terms of spatial distribution.

[0079] In practice, the voltage deviation gradient data set obtained in step S101 is first called. and the mapping weight set generated in step S1021 Then, the voltage deviation gradient at the same node is correspondingly... With mapping weights Perform multiplication to obtain the corresponding disturbance intensity coefficient. Right now Finally, by traversing all target load nodes, a set of disturbance intensity coefficients is obtained. .

[0080] S1023. The spatial coordinates are vector-scaled using the disturbance intensity coefficient to obtain the disturbance vector components of each target load node. The voltage disturbance vector is obtained by superimposing and summing all the disturbance vector components.

[0081] The disturbance vector component refers to the directed line segment obtained by applying the disturbance intensity coefficient as a magnitude scaling factor to the node's spatial coordinate vector, representing the direction and intensity of the voltage instability caused by that node across the entire network. The voltage disturbance vector is the composite result of the disturbance vector components of all nodes in the vector space.

[0082] In practice, the relative spatial coordinates of each target load node are first converted into position vectors. Next, the disturbance intensity coefficient calculated in step S1022 is used. Perform a scalar multiplication on this position vector, that is... The disturbance vector components of each node are obtained. Finally, for all areas within the distribution network... The disturbance vector components of each target load node are geometrically superimposed using the formula. The final voltage disturbance vector is calculated. .

[0083] Optionally, the process of normalizing and weighting the rated capacity and reactive power space of each distributed power node in step S102 to obtain the gain matrix may specifically include:

[0084] S1024. Calculate the ratio of the rated capacity of each distributed power node to the total rated capacity of all distributed power nodes to obtain the power weight of each distributed power node.

[0085] Power weight refers to the share of the inverter capacity of a single distributed power source in the total capacity of all distributed power sources participating in regulation within the entire distribution network area. It is used to measure the benchmark proportion that the node should bear in the global regulation task.

[0086] In practice, the rated capacity of each distributed power inverter read in step S101 is first... The total capacity is obtained by summing the results. Then, the rated capacity of each distributed power node is... Divide by total capacity The dimensionless power weights are obtained. .

[0087] S1025. Calculate the ratio of reactive power space to the corresponding rated capacity to obtain the adjustment gain coefficient of each distributed power node.

[0088] The adjustment gain coefficient refers to the ratio between the inverter's currently available reactive power margin and its total rated capacity, reflecting the equipment's reactive power adjustment margin or potential utilization rate under current operating conditions. The larger the coefficient, the stronger the inverter's reactive power support capability without affecting active power output.

[0089] In specific implementation, the adjustable reactive power space set determined in step S101 is first obtained. and known rated capacity Next, for each distributed power node, its reactive power spatial value is calculated. Compared with the average rated capacity of the entire network The ratio of the two values ​​yields the corresponding adjustment gain coefficient. .

[0090] S1026. Calculate the product of the power weight and the corresponding adjustment gain coefficient to obtain the adjustment coefficient of each distributed power node, and map all the adjustment coefficients to a preset multidimensional diagonal matrix according to the position index of each distributed power node to obtain the gain matrix.

[0091] The regulation coefficient is a comprehensive performance index that takes into account both the static capacity of the inverter and its current dynamic adjustment margin. The gain matrix is ​​a square matrix with the regulation coefficients of each inverter as its diagonal elements, used to define the output boundaries of each node.

[0092] In specific implementation, the power weight obtained in step S1024 is first... The adjustment gain coefficient obtained in step S1025 Perform corresponding multiplication to obtain the adjustment coefficient. Then, build a Diagonal matrix of dimension All calculated adjustment coefficients Fill the matrix along its main diagonal according to the index order of the distributed power nodes, and pad the remaining elements with zeros. The resulting gain matrix... The format is as follows:

[0093]

[0094] For example, suppose the regulation coefficient of inverter C is... The adjustment coefficient of inverter D is Then the gain matrix is

[0095] This embodiment realizes the vectorized guidance of the network-wide regulation demand, quantifies the dynamic regulation potential of each node, and effectively avoids the risk of equipment overload while improving voltage quality.

[0096] S103. Using the voltage disturbance vector as the guiding parameter, the gain matrix as the constraint coefficient, the minimization of the global network loss of the distribution network as the objective function, and the voltage deviation of each target load node within a safe range as the constraint boundary, the weight vector of reactive power allocation of each distributed power generation node is obtained by iterative solution through Lagrange duality.

[0097] Optionally, step S103, which uses the voltage disturbance vector as the guiding parameter, the gain matrix as the constraint coefficient, the minimization of the global network loss of the distribution network as the objective function, and the voltage deviation of each target load node within a safe range as the constraint boundary, and iteratively solves the problem through Lagrange duality to obtain the reactive power allocation weight vector of each distributed power generation node, may specifically include:

[0098] S1031. Determine the reactive power adjustment direction of each target load node based on the guiding parameters, and use the reactive power adjustment direction and constraint coefficient to perform correlation mapping between the objective function and the constraint boundary to obtain the mapping correlation matrix.

[0099] Reactive power regulation direction refers to the operational polarity determined based on the voltage fluctuation trend indicated by the voltage disturbance vector, which dictates whether to inject inductive reactive power into the grid (i.e., boost voltage) or absorb capacitive reactive power (i.e., reduce voltage). The objective function is a mathematical expression aimed at minimizing active power losses in distribution network lines while satisfying power system operating constraints.

[0100] Constraint boundaries refer to the inequality condition that the voltage of each node must be kept within the allowable deviation range specified by national standards, such as ±7% of the rated voltage. The mapping correlation matrix is ​​a comprehensive coefficient matrix that integrates the reactive power regulation direction sign, the regulation capability coefficient in the gain matrix, and network topology sensitivity information. It is used to connect the control variable (reactive power output) with the state variables (node ​​voltage and network loss).

[0101] In specific implementation, the voltage disturbance vector obtained in step S1023 is first analyzed. The voltage collapse region is defined as a sector in the coordinate system pointing towards the region with the highest concentration of negative voltage deviation gradients. Calculation... Direction angle And combine it with the geographical azimuth of each feeder branch of the distribution network. Perform association mapping. If If the location is close to that of a certain feeder branch, the area is determined to be a key area for voltage instability, and [further details will be provided]. The modulus length is used as the demand intensity factor for global regulation to determine the direction indicator of reactive power regulation across the entire network. ,like If it points in the direction of decreasing voltage, then It is +1 if it is +1, and -1 otherwise.

[0102] Next, a reactive power output vector of each distributed power source is constructed. The objective function for minimizing the global network loss for the independent variable is shown in the following formula (1):

[0103] (1)

[0104] in For the first The resistance of the line, Let be the current function flowing through this line. Simultaneously, establish voltage constraint boundaries: .in As the reference voltage vector, This is the voltage reactive power sensitivity matrix. This is the lower limit vector of voltage amplitude. This is the upper limit vector of voltage amplitude.

[0105] Specifically, this sensitivity matrix is ​​extracted by calculating the inverse of the Jacobian matrix of the distribution network power flow equations. First, based on the node admittance matrix of the distribution network... Construct the power balance equations using the current node voltage phasors:

[0106]

[0107]

[0108] in, and Representing nodes respectively The active power injection setpoint and reactive power injection setpoint are the generator output minus the load power. and These represent the voltage amplitude at the current node, respectively. Voltage phase angle and admittance matrix electrical conductivity and susceptance Calculated nodes The calculated values ​​of active power and reactive power.

[0109] Subsequently, the Jacobian matrix was obtained using the Newton-Raphson method. Its block form is , where submatrix Under the decoupling assumption, neglecting the effect of active power on voltage, the submatrix is ​​calculated. inverse matrix and will As the voltage reactive power sensitivity matrix ,Right now Finally, adjust the direction indicator. as well as The normalized component is applied to the sensitivity matrix to generate a mapping correlation matrix. .

[0110] S1032. The mapping correlation matrix is ​​weighted and superimposed using preset multiplier adjustment parameters and penalty adjustment parameters to obtain the augmented dual function.

[0111] The preset multiplier adjustment parameter refers to the constant factor used in the Lagrange multiplier method to initialize the dual variable, i.e., the Lagrange multiplier step size or scaling ratio, and affects the algorithm's sensitivity to constraint violations. The penalty adjustment parameter refers to the weight coefficient of the quadratic penalty term introduced in the augmented Lagrange function, used to enhance the convexity of the objective function, accelerate convergence, and prevent oscillations at the feasible region boundary during iteration. The augmented dual function is the comprehensive evaluation function that combines the original objective function, the Lagrange term of the linear constraints, and the quadratic penalty term; it is the direct object of iterative optimization.

[0112] In practice, the mapping-related matrix is ​​first calculated. infinite norm and network node size Using formulas Calculate the multiplier adjustment parameter This ensures that the update step size of the dual variable matches the magnitude of the constraint gradient. Using the formula... Calculate penalty adjustment parameters ,in It is an acceleration factor with a value ranging from 5 to 10. These are the diagonal elements of the gain matrix. For a preset non-zero small constant, such as This is to prevent the denominator from being zero, thereby ensuring that the penalty term can effectively suppress constraint violations without destroying the convexity of the objective function.

[0113] Next, we introduce the Lagrange multiplier vector. The voltage constraints correspond to each node. The mapping correlation matrix obtained in step S1031 is used... and the original objective function Constructing augmented dual functions As shown in the following formula (2):

[0114] (2)

[0115] in This is the voltage constraint boundary vector.

[0116] S1033. Use progressive optimization to perform gradient iteration on the augmented dual function to obtain the weight vector of each distributed power node.

[0117] Progressive optimization refers to a numerical calculation strategy that approximates the optimal solution step by step, typically including prediction, correction, and step size adjustment. The weight vector refers to the final calculated reactive power allocation ratio vector that each distributed power source should bear, which directly determines the operational range of each inverter.

[0118] In practice, interior-point methods or gradient descent methods are used to apply the augmented dual function. Perform iterative solution. First, initialize the reactive power output vector. In the first In this iteration, the function is calculated with respect to... gradient Update along the negative gradient direction. Subsequently, after multiple iterations until the convergence condition is met, such as the gradient magnitude being less than 1 / 2, the process continues. The optimal reactive power output vector is obtained. Finally, extract directly. The elements in the vector are used to obtain the reactive power allocation weight vector for each distributed power node. .

[0119] This embodiment realizes the global optimal solution search under multiple objectives, ensuring the rapid convergence and numerical stability of the algorithm under complex distribution network topology. The final output weight vector can accurately guide each distributed power source to output power on demand and according to energy, taking into account both regulation effect and economic operation.

[0120] Optionally, step S1033, which uses progressive optimization to perform gradient iteration on the augmented dual function to obtain the weight vector of each distributed power node, may specifically include:

[0121] Figure 2 A flowchart illustrating a method for generating weight vectors according to an embodiment of this application is shown. Figure 2 As shown, in the optimization process, starting from a preset initial weight distribution, an iterative gradient descent strategy based on the augmented Lagrange multiplier method is adopted, i.e., a progressive optimization process. Relying on the search guidance vector of the augmented dual function in each distributed power node dimension, a step-by-step recursive mapping operation is performed. This operation gradually generates a series of candidate solution vectors along the gradient descent direction of the objective function, thereby constructing a complete optimization path sequence.

[0122] Subsequently, a convergence analysis is performed on the optimization path sequence. Specifically, the deviation value, which quantifies the degree of fluctuation of the solution vector, is obtained by calculating the difference magnitude between two adjacent candidate solution vectors in the sequence. When the deviation value falls within the preset steady-state threshold range, it indicates that the iterative process has stabilized, and the current candidate solution vector is then locked as the convergence target solution.

[0123] Finally, numerical components corresponding one-to-one with each distributed power node are separated from the convergent objective solution, and the weight vector of each distributed power node is finally established through numerical mapping. This method includes:

[0124] S10331. Using the augmented dual function as the search guidance vector in each distributed power node dimension, a step-by-step recursive mapping is performed on the preset initial weight distribution to obtain an optimization path sequence composed of multiple candidate solution vectors.

[0125] The search steering vector refers to the gradient vector formed by the first-order partial derivatives of the augmented dual function with respect to the reactive power output variables of each distributed power source at the current iteration point, indicating the direction of the fastest descent of the function value. The preset initial weight distribution refers to the initial reactive power output values ​​assigned to each distributed power source node before the algorithm starts, which are usually set according to the rated capacity ratio or historical operating data.

[0126] Step-by-step recursive mapping refers to the process of updating the current solution vector to a better new solution vector by following the iterative rules of gradient descent and utilizing the search guidance vector and a set step size parameter. The optimization path sequence refers to an ordered set of intermediate solution vectors generated after multiple recursive mappings from the initial solution. The preset initial weight distribution is shown in Table 2 below:

[0127] Table 2: Initial Weight Distribution Comparison Table

[0128]

[0129] As shown in Table 2, Table 2 illustrates an example of the preset initial weight distribution configuration used in this embodiment, which lists the initial reactive power allocation weights for different types of distributed power nodes. This table provides a starting point for iterative optimization; setting reasonable initial values ​​helps reduce the number of iterations and avoid getting trapped in local optima.

[0130] In practice, the initial solution vector is first constructed based on the actual situation of each distributed power source in the distribution network and the contents shown in Table 2. For example, according to Table 2, if there are three power sources DER1, DER2, and DER3, then the initial solution vector... Next, the augmented dual function is calculated with respect to... The gradient vector, as the first... The search guidance vector for the next iteration .

[0131] Then, using the formula Perform recursive mapping, where This refers to the learning rate or step size. (After...) In the next iteration, a sequence of optimization paths is generated, including multiple candidate solution vectors. .

[0132] S10332. Calculate the difference magnitude between two adjacent candidate solution vectors in the optimization path sequence to obtain the corresponding deviation value.

[0133] The difference magnitude refers to the numerical distance between the solution vector obtained in a later iteration and the solution vector obtained in the previous iteration in the optimization path sequence. It is usually measured using the Euclidean norm or the maximum norm. The deviation value is a quantized scalar of the difference magnitude, used to evaluate the convergence and stability of the current iteration process.

[0134] In practice, the optimization path sequence generated in step S10331 is first extracted. adjacent elements and Next, calculate the difference vector between the two. Then, the magnitude of the difference vector is calculated using norm operations to obtain the first... Deviation value of the next iteration .

[0135] S10333. The candidate solution vector with deviation value within the preset steady-state threshold range is determined as the convergence target solution, and the numerical components corresponding to each distributed power node in the convergence target solution are extracted and numerically mapped to obtain the weight vector of each distributed power node.

[0136] The preset steady-state threshold range refers to the numerical standard interval for determining whether the optimization algorithm has reached convergence, and the solution is considered stable when the deviation falls within this interval. The convergent target solution refers to the solution vector obtained in the last iteration that satisfies the steady-state threshold condition, representing the globally optimal or locally optimal reactive power allocation scheme. Numerical mapping refers to the process of directly assigning values ​​to the values ​​in the convergent target solution or converting them proportionally to the final control weights. The preset steady-state threshold ranges are shown in Table 3 below:

[0137] Table 3: Steady-state threshold range comparison table

[0138]

[0139] As shown in Table 3, Table 3 illustrates an example configuration of the preset steady-state threshold range used in this embodiment, specifying the judgment thresholds under different accuracy requirements. This table ensures that the accuracy of the output results meets the actual needs of the control system, balancing computation time and control accuracy.

[0140] In practice, the judgment threshold should first be determined by referring to Table 3 based on the application scenario. For example, in practical engineering applications, select Next, check the deviation value calculated in step S10332. If satisfied If the iteration stops, the current candidate solution vector is changed. Determined as the convergent objective solution Finally, extract Each component in This value is assigned to the reactive power allocation weight vector of each distributed power node. ,in .

[0141] This embodiment achieves rapid and accurate positioning of the optimal reactive power allocation weights, effectively preventing algorithm oscillations or premature convergence, and ensuring the mathematical convergence and physical feasibility of the final output weight vector.

[0142] S104. The ratio of the weight vector to the per-unit voltage value of each distributed power node is used as the initial reference value of the reactive current. The initial reference value is then subjected to first-order inertial filtering using a first-order inertial element to match the adjustment response speed of the inverter, thereby obtaining the reference base value.

[0143] Optionally, step S104, which uses the ratio of the weight vector to the per-unit voltage value of each distributed power generation node as the initial reference value of the reactive current, and performs first-order inertial filtering on the initial reference value using a first-order inertial element to match the inverter's adjustment response speed, to obtain the reference base value, may specifically include:

[0144] Figure 3 A flowchart illustrating a method for generating reference base values ​​according to an embodiment of this application is shown. Figure 3 As shown, the method includes:

[0145] S1041. Calculate the ratio of the weight vector to the corresponding voltage per-unit value to obtain the initial reference value of the reactive current of each distributed power generation node.

[0146] The per-unit voltage value refers to the ratio of the real-time voltage measurement at the grid connection point of a distributed power source to the reference voltage, used to standardize the numerical dimensions under different voltage levels. The initial reference value of reactive current refers to the reactive current component that the inverter should theoretically inject into the grid to achieve the predetermined reactive power output target under ideal steady-state conditions.

[0147] In practice, the optimal reactive power allocation weight vector of each distributed power node obtained in step S10333 is first called. Simultaneously, the effective voltage values ​​of each distributed power source's grid connection point are collected in real time. and divided by the reference voltage The set of voltage per unit values ​​is obtained. Next, based on instantaneous power theory, using the formula... Calculate the initial reference value of reactive current for each node. Finally, iterate through all nodes to generate an initial reference value vector. .

[0148] S1042. The amplitude change rate of the initial reference value is limited by the preset time constant parameter in the first-order inertial element to obtain the current transition component.

[0149] A first-order inertial element refers to a transfer function module in a control system used to simulate the hysteresis characteristics of a physical object or to smooth sudden changes in input signals. It typically takes the form of... Preset time constant parameter This refers to the key indicator that determines the response speed of this stage; the larger the value, the smoother the output follows the input, used to match the inherent response time of different types of inverters. The current transition component refers to the intermediate-state current command that changes smoothly over time after inertial filtering. The preset time constant parameters are shown in Table 4 below:

[0150] Table 4: Time Constant Parameter Comparison Table

[0151]

[0152] As shown in Table 4, this table illustrates the preset time constant parameters set for different types of inverters in this embodiment, listing typical response time constants for devices such as photovoltaic inverters and energy storage converters. The data in this table ensures that the dynamic characteristics of the control commands match the actual capabilities of the physical equipment, preventing equipment protection tripping due to excessively rapid command changes.

[0153] In practice, first, based on the equipment type and capacity of each distributed power source, refer to Table 4 to determine the corresponding time constant. For example, for Type-A devices, select... Next, the initial reference value obtained in step S1041 is... As input, the current time step is calculated using a discretization algorithm such as the backward difference method. Current transition component The specific iterative formula is as follows: ,in The sampling period is It is based on the time constant and sampling period The calculated filter coefficients and .

[0154] In particular, when selecting a time constant At this time, it is necessary to ensure a balance between filtering smoothing and system stability. For centralized power plants with extremely slow response, a phase lead compensation stage is used in the underlying control algorithm to offset the phase lag caused by the first-order inertial filter at the fundamental frequency.

[0155] S1043. A reference value is obtained by sampling and mapping the current transition component in the time domain.

[0156] Discrete-time sampling mapping refers to the process of sampling and holding the current transient components in the continuous time domain or high-frequency computation domain according to the execution cycle of the inverter's underlying controller, such as the PWM carrier cycle. The reference value refers to the discrete current command that is ultimately input to the current loop controller and remains constant in each control cycle.

[0157] In practice, the sampling frequency of the inverter controller must first be determined. For example, 10kHz corresponds to a sampling period. Then, at each sampling time... Read the real-time current transition component calculated in step S1042. Then, this instantaneous value is locked and remains unchanged within the current control cycle, serving as the reference baseline value for that cycle. .

[0158] This embodiment achieves flexible matching of the regulation speed of each distributed power source, which not only avoids the inverter overcurrent protection from malfunctioning due to sudden changes in commands, but also ensures that the entire network is in sync during the regulation process, effectively eliminating voltage feedback oscillations caused by uneven speeds. At the same time, the phase compensation technology ensures high-fidelity execution of the regulation commands in the time domain, improving dynamic stability.

[0159] S105. At the resonant pole of the grid fundamental frequency, the reference voltage vector is obtained by performing error integral compensation on the difference between the reference reference value and the real-time feedback current using quasi-proportional resonance. The reference voltage vector is then converted into a pulse sequence signal through pulse width modulation mapping to control the inverter for state switching.

[0160] The fundamental frequency of the power grid refers to the standard power frequency in the distribution network, which is usually 50Hz in China. The resonant pole is the frequency point in the QPR transfer function of a quasi-proportional resonant controller where the gain approaches infinity or a maximum. By setting it at the reference frequency, zero steady-state error tracking of AC signals at that frequency can be achieved.

[0161] The reference voltage vector refers to the voltage command signal output by the inverter controller used to drive the power electronic switches, including amplitude and phase information. Pulse width modulation mapping refers to the process of converting the continuously changing reference voltage vector into a high-frequency digital pulse sequence that controls the switching transistors to turn on and off, using space vector pulse width modulation (SVPWM) or sinusoidal pulse width modulation (SPWM) techniques.

[0162] In practice, the resonant frequency of the QPR controller is first set. rad / s. Furthermore, based on the open-loop cutoff frequency of the inverter system... Calculate control parameters based on the phase margin (PM) requirement. First, based on the allowable grid frequency fluctuation range, which is typically [missing information]. Hz, setting the cutoff frequency bandwidth , usually take rad / s.

[0163] Next, based on the desired bandwidth And it is usually taken as 1 / 10 of the switching frequency. rad / s, ignoring the influence of high-frequency resonance terms, using the formula Calculate the proportionality coefficient, where This is the value of the filter inductance. This is the DC bus voltage. Finally, to ensure that the steady-state error at the fundamental frequency is less than a preset value... Using the formula Calculate the resonance coefficient.

[0164] Configure the above parameters Then, the feedback current output by the inverter is collected in real time. And calculate its value compared with the reference base value obtained in step S104. The difference between Subsequently, the difference was decomposed into proportional feedback components. Harmonic compensation component ,in , for After the resonant channel The time-domain response after integral compensation. Then, the two are superimposed and introduced into the real-time acquired grid background voltage feedforward component. To counteract the disturbance of the current loop by the grid electromotive force, the reference voltage vector in the time domain is obtained. .

[0165] Finally, The input is fed into the SVPWM modulation module, which generates six PWM drive signals through sector judgment and action time calculation. These signals control the switching of IGBTs or MOSFETs inside the inverter, thereby ensuring that the actual reactive current output by the inverter accurately follows the instructions given by the upper-level optimization, ultimately achieving coordinated support for the distribution network voltage.

[0166] This embodiment utilizes quasi-proportional resonant control technology to achieve zero steady-state error tracking of the AC reference signal, effectively overcoming the shortcomings of traditional PI control in steady-state error in a stationary coordinate system. This ensures that each distributed power source can perform reactive power regulation tasks with high precision and fast response, guaranteeing the implementation of the power generation coordination control strategy at the actual circuit level.

[0167] Optionally, the process in step S105 of obtaining the reference voltage vector by using quasi-proportional resonance to perform error integration compensation on the difference between the reference reference value and the real-time feedback current at the resonant pole at the fundamental frequency of the power grid can specifically include:

[0168] S1051. Calculate the difference between the reference value and the real-time feedback current to obtain the current deviation component.

[0169] The current deviation component refers to the difference between the expected reactive current command input to the inverter controller and the actual sampled grid-connected current in each control cycle, and represents the instantaneous deviation between the current output state and the target state.

[0170] In practical implementation, firstly, in each PWM control cycle, such as At the start time, read the reference value obtained in step S1043. Simultaneously, the actual output current at the inverter's grid connection point is acquired in real time via a current transformer and an analog-to-digital converter (ADC). Next, use subtraction. Calculate the current time Current deviation component .

[0171] S1052. By using the resonant poles, the time-domain values ​​of the current deviation components corresponding to the fundamental frequency of the power grid are accumulated to obtain the resonant compensation vector.

[0172] A resonant pole is a frequency point in the denominator of the transfer function of a quasi-proportional resonant controller that causes the gain to approach infinity. It is typically located at the fundamental frequency of the power grid, such as 50Hz, and is used for attenuated integration of the error signal at that frequency. The resonant compensation vector is the control quantity component output after processing by the resonant circuit. It specifically compensates for the fundamental AC component of the current deviation to eliminate steady-state error.

[0173] In practice, the resonant frequency of the QPR controller is first set. rad / s, and configure the resonance coefficient and cutoff frequency bandwidth Next, the current deviation component obtained in step S1051 is... Performing a Laplace transform yields .

[0174] Then, the error signal Input to resonant channel transfer function Finally, the transfer function is transformed into a difference equation executable by the digital controller using discretization methods such as bilinear transform or zero-order hold method. The time-domain resonance compensation vector is then iteratively calculated based on the error values ​​at the current and historical moments. .

[0175] S1053. The current deviation component is linearly scaled to obtain the proportional feedback component, and the proportional feedback component and the resonant compensation vector are superimposed and summed to obtain the reference voltage vector.

[0176] The proportional feedback component refers to the control quantity obtained by directly multiplying the current deviation component by a fixed proportional gain coefficient. It is mainly used to improve dynamic response speed and make an immediate response to sudden errors. The reference voltage vector refers to the total control command obtained by synthesizing the fast response component of the proportional channel and the steady-state compensation component of the resonant channel. It directly determines the amplitude and phase of the inverter output voltage.

[0177] In practice, the proportional gain coefficient is first set. Next, the proportional feedback component is calculated. Then, this proportional component is compared with the resonance compensation vector obtained in step S1052. By performing time-domain superposition, the final reference voltage vector is obtained. The formula is: .

[0178] This embodiment achieves zero steady-state error tracking of AC reactive current commands, effectively solving the problems of steady-state error and phase lag in traditional control in a stationary coordinate system. It ensures that the inverter executes the upper-level optimization strategy accurately and quickly, significantly improving the regulation accuracy and stability of multi-point coordinated control in the distribution network.

[0179] Figure 4 This application provides a schematic diagram of a specific implementation of a power generation coordination control system for a new energy power system, referring to... Figure 4 The system may include:

[0180] The acquisition module 410 is used to synchronously acquire the voltage deviation gradient data of the target load node in the distribution network and the adjustable reactive power space of the inverter at each distributed power grid connection point under the output condition.

[0181] The generation module 420 is used to obtain the voltage disturbance vector by weighting and vector aggregation of voltage deviation gradient data and associated geographical location information, and to normalize and weight the rated capacity and reactive power space of each distributed power node to obtain the gain matrix.

[0182] The solver module 430 is used to obtain the reactive power allocation weight vector of each distributed power generation node by iteratively solving through Lagrange duality, with voltage disturbance vector as the guiding parameter, gain matrix as the constraint coefficient, minimizing the global network loss of the distribution network as the objective function, and voltage deviation of each target load node within a safe range as the constraint boundary.

[0183] The generation module 420 is also used to take the ratio of the weight vector to the per-unit voltage value of each distributed power node as the initial reference value of the reactive current, and use a first-order inertial element to perform first-order inertial filtering on the initial reference value to match the adjustment response speed of the inverter and obtain the reference reference value.

[0184] The control module 440 is used to obtain a reference voltage vector by performing error integral compensation on the difference between the reference reference value and the real-time feedback current at the resonant pole of the grid fundamental frequency using quasi-proportional resonance. The reference voltage vector is then converted into a pulse sequence signal through pulse width modulation mapping to control the inverter for state switching.

[0185] The power generation coordination control system of the new energy power system in this application embodiment is used to implement the aforementioned power generation coordination control method of the new energy power system. Therefore, the specific implementation of the power generation coordination control system of the new energy power system can be found in the embodiment section of the power generation coordination control method of the new energy power system above. The specific implementation can be referred to the description of the corresponding embodiments, which will not be repeated here.

[0186] Figure 5 A schematic diagram of the hardware structure of an electronic device provided in one embodiment of this application is shown.

[0187] The electronic device may include a processor 510 and a memory 520 storing computer program instructions.

[0188] Specifically, the processor 510 may include a central processing unit (CPU), an application-specific integrated circuit (ASIC), or one or more integrated circuits that can be configured to implement the embodiments of this application.

[0189] Memory 520 may include mass storage for data or instructions. For example, and not limitingly, memory 520 may include a hard disk drive (HDD), floppy disk drive, flash memory, optical disk, magneto-optical disk, magnetic tape, or Universal Serial Bus (USB) drive, or a combination of two or more of these. Where appropriate, memory 520 may include removable or non-removable (or fixed) media. Where appropriate, memory 520 may be internal or external to the integrated gateway disaster recovery device. In a particular embodiment, memory 520 is non-volatile solid-state memory.

[0190] Memory may include read-only memory (ROM), random access memory (RAM), disk storage media devices, optical storage media devices, flash memory devices, and electrical, optical, or other physical / tangible memory storage devices. Therefore, typically, memory includes one or more tangible (non-transitory) computer-readable storage media (e.g., memory devices) encoded with software including computer-executable instructions, and when the software is executed (e.g., by one or more processors), it is operable to perform the operations described with reference to the method according to the first aspect of this disclosure.

[0191] The processor 510 reads and executes computer program instructions stored in the memory 520 to implement any of the power generation coordination control methods of the new energy power system in the above embodiments.

[0192] In one example, the electronic device may also include a communication interface 530 and a bus 540. Wherein, such as Figure 5 As shown, the processor 510, memory 520, and communication interface 530 are connected through bus 540 and complete communication with each other.

[0193] The communication interface 530 is mainly used to realize communication between various modules, devices, units and / or equipment in the embodiments of this application.

[0194] Bus 540 includes hardware, software, or both, that couples components of an online data traffic metering device together. For example, and not limitingly, the bus may include an Accelerated Graphics Port (AGP) or other graphics bus, an Enhanced Industry Standard Architecture (EISA) bus, a Front Side Bus (FSB), HyperTransport (HT) interconnect, an Industry Standard Architecture (ISA) bus, an Infinite Bandwidth Interconnect, a Low Pin Count (LPC) bus, a memory bus, a Microchannel Architecture (MCA) bus, a Peripheral Component Interconnect (PCI) bus, a PCI-Express (PCI-X) bus, a Serial Advanced Technology Attachment (SATA) bus, a Video Electronics Standards Association Local (VLB) bus, or other suitable buses, or combinations of two or more of these. Where appropriate, bus 540 may include one or more buses. Although specific buses are described and illustrated in embodiments of this application, any suitable bus or interconnect is contemplated herein.

[0195] The electronic device can execute the power generation coordination control method of the new energy power system in the embodiments of this application, thereby realizing the power generation coordination control method of the new energy power system described in conjunction with the accompanying drawings.

[0196] Furthermore, in conjunction with the power generation coordination and control method for the new energy power system in the above embodiments, this application embodiment can provide a computer-readable storage medium for implementation. This computer-readable storage medium stores computer program instructions; when these computer program instructions are executed by a processor, they implement any of the power generation coordination and control methods for the new energy power system in the above embodiments.

[0197] It should be clarified that this application is not limited to the specific configurations and processes described above and shown in the figures. For the sake of brevity, detailed descriptions of known methods are omitted here. In the above embodiments, several specific steps are described and shown as examples. However, the method process of this application is not limited to the specific steps described and shown. Those skilled in the art can make various changes, modifications, and additions, or change the order of steps, after understanding the spirit of this application.

[0198] The functional blocks shown in the above-described structural diagram can be implemented as hardware, software, firmware, or a combination thereof. When implemented in hardware, they can be, for example, electronic circuits, application-specific integrated circuits (ASICs), appropriate firmware, plug-ins, function cards, etc. When implemented in software, the elements of this application are programs or code segments used to perform the required tasks. Programs or code segments can be stored on a machine-readable medium or transmitted over a transmission medium or communication link via data signals carried on a carrier wave. "Machine-readable medium" can include any medium capable of storing or transmitting information. Examples of machine-readable media include electronic circuits, semiconductor memory devices, ROM, flash memory, erasable ROM (EROM), floppy disks, CD-ROMs, optical disks, hard disks, fiber optic media, radio frequency (RF) links, etc. Code segments can be downloaded via computer networks such as the Internet, intranets, etc.

[0199] It should also be noted that the exemplary embodiments mentioned in this application describe methods or systems based on a series of steps or apparatus. However, this application is not limited to the order of the above steps; that is, the steps can be performed in the order mentioned in the embodiments, or in a different order, or several steps can be performed simultaneously.

[0200] The aspects of this application have been described above with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of this application. It should be understood that each block in the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, or other programmable data processing apparatus to produce a machine such that these instructions, executable via the processor of the computer or other programmable data processing apparatus, enable the implementation of the functions / actions specified in one or more blocks of the flowchart illustrations and / or block diagrams. Such a processor can be, but is not limited to, a general-purpose processor, a special-purpose processor, a special application processor, or a field-programmable logic circuit. It is also understood that each block in the block diagrams and / or flowcharts, and combinations of blocks in the block diagrams and / or flowcharts, can also be implemented by dedicated hardware performing the specified functions or actions, or can be implemented by a combination of dedicated hardware and computer instructions.

[0201] The above provides a detailed description of the power generation coordination control method and system for a new energy power system provided in this application. Specific examples have been used to illustrate the principles and implementation methods of this application. The descriptions of the embodiments above are only for the purpose of helping to understand the method and its core ideas. It should be noted that those skilled in the art can make several improvements and modifications to this application without departing from the principles of this application, and these improvements and modifications also fall within the protection scope of this application.

Claims

1. A method for coordinated control of power generation in a new energy power system, characterized in that, include: Simultaneously acquire voltage deviation gradient data of target load nodes in the distribution network and the adjustable reactive power space of inverters at each distributed power source grid connection point under output conditions; By performing weighted mapping and vector aggregation on the voltage deviation gradient data and associated geographical location information, a voltage disturbance vector is obtained. Then, the rated capacity of each distributed power node and the reactive power space are normalized and weighted to obtain a gain matrix. Using the voltage disturbance vector as the guiding parameter, the gain matrix as the constraint coefficient, minimizing the global network loss of the distribution network as the objective function, and ensuring that the voltage deviation of each target load node is within a safe range as the constraint boundary, the weight vector of reactive power allocation for each distributed power generation node is obtained by iterative solution through Lagrange duality. The ratio of the weight vector to the per-unit voltage value of each distributed power node is used as the initial reference value of the reactive current. The initial reference value is then subjected to first-order inertial filtering using a first-order inertial element to match the adjustment response speed of the inverter, thereby obtaining a reference base value. At the resonant pole of the grid fundamental frequency, the error integral compensation of the difference between the reference reference value and the real-time feedback current is performed by using quasi-proportional resonance to obtain the reference voltage vector. The reference voltage vector is then converted into a pulse sequence signal through pulse width modulation mapping to control the inverter to switch states.

2. The method according to claim 1, characterized in that, The step of obtaining a voltage disturbance vector by weighting and vector aggregation of the voltage deviation gradient data and associated geographic location information includes: Based on the geographical location information, determine the spatial coordinates of each target load node relative to the preset equilibrium point, and calculate the straight-line distance between each target load node and the preset equilibrium point to obtain the mapping weight of each target load node; The disturbance intensity coefficient of each target load node is obtained by multiplying the voltage deviation gradient data with the corresponding mapping weight. The disturbance intensity coefficient is used to vector scale the spatial coordinates to obtain the disturbance vector components of each target load node, and the voltage disturbance vector is obtained by superimposing and summing all the disturbance vector components.

3. The method according to claim 1, characterized in that, The normalization and weighting of the rated capacity of each distributed power node and the reactive power space to obtain the gain matrix includes: Calculate the ratio of the rated capacity of each distributed power node to the total rated capacity of all the distributed power nodes to obtain the power weight of each distributed power node. Calculate the ratio of the reactive power space to the corresponding rated capacity to obtain the adjustment gain coefficient of each distributed power node; The power weight is calculated by multiplying it by the corresponding adjustment gain coefficient to obtain the adjustment coefficient of each distributed power node. All the adjustment coefficients are then mapped to a preset multidimensional diagonal matrix according to the position index of each distributed power node to obtain the gain matrix.

4. The method according to claim 1, characterized in that, The method uses the voltage disturbance vector as the guiding parameter, the gain matrix as the constraint coefficient, the minimization of the global network loss of the distribution network as the objective function, and the voltage deviation of each target load node within a safe range as the constraint boundary. Through iterative solution using Lagrange duality, the reactive power allocation weight vector of each distributed power generation node is obtained, including: The reactive power adjustment direction of each target load node is determined based on the guiding parameters, and the objective function and the constraint boundary are correlated and mapped using the reactive power adjustment direction and the constraint coefficient to obtain the mapping correlation matrix; The augmented dual function is obtained by weighting and superimposing the mapping correlation matrix using preset multiplier adjustment parameters and penalty adjustment parameters. The augmented dual function is iterated by gradient optimization to obtain the weight vector of each distributed power node.

5. The method according to claim 4, characterized in that, The step of performing gradient iteration on the augmented dual function using progressive optimization to obtain the weight vector of each distributed power node includes: By utilizing the search guidance vector of the augmented dual function in each distributed power node dimension, a step-by-step recursive mapping is performed on the preset initial weight distribution to obtain an optimization path sequence composed of multiple candidate solution vectors. Calculate the magnitude of the difference between two adjacent candidate solution vectors in the optimization path sequence to obtain the corresponding deviation value; The candidate solution vector whose deviation value is within the preset steady-state threshold range is determined as the convergence target solution, and the numerical components corresponding to each distributed power node in the convergence target solution are extracted and numerically mapped to obtain the weight vector of each distributed power node.

6. The method according to claim 1, characterized in that, The initial reference value of reactive current is obtained by using the ratio of the weight vector to the per-unit voltage value of each distributed power node as the initial reference value, and by performing first-order inertial filtering on the initial reference value using a first-order inertial element to match the adjustment response speed of the inverter, including: Calculate the ratio of the weight vector to the corresponding voltage per-unit value to obtain the initial reference value of the reactive current of each distributed power generation node; The amplitude change rate of the initial reference value is limited by the preset time constant parameter in the first-order inertial element to obtain the current transition component; The reference value is obtained by sampling and mapping the current transition component in the time domain.

7. The method according to claim 1, characterized in that, The resonant pole at the fundamental frequency of the power grid is used to obtain a reference voltage vector by integrating the error between the reference reference value and the real-time feedback current using quasi-proportional resonance. This includes: The difference between the reference value and the real-time feedback current is calculated to obtain the current deviation component; By using the resonant poles to accumulate the time-domain values ​​of the current deviation components corresponding to the fundamental frequency of the power grid, a resonant compensation vector is obtained. The current deviation component is linearly scaled to obtain a proportional feedback component, and the proportional feedback component and the resonant compensation vector are superimposed and summed to obtain the reference voltage vector.

8. A power generation coordination control system for a new energy power system, characterized in that, include: The acquisition module is used to synchronously acquire the voltage deviation gradient data of the target load node in the distribution network and the adjustable reactive power space of the inverter at each distributed power source grid connection point under the output operating condition. The generation module is used to obtain a voltage disturbance vector by performing weighted mapping and vector aggregation on the voltage deviation gradient data and associated geographical location information, and to normalize and weight the rated capacity of each distributed power node and the reactive power space to obtain a gain matrix. The solution module is used to obtain the reactive power allocation weight vector of each distributed power generation node by iteratively solving through Lagrange duality, with the voltage disturbance vector as the guiding parameter, the gain matrix as the constraint coefficient, the minimization of the global network loss of the distribution network as the objective function, and the voltage deviation of each target load node within a safe range as the constraint boundary. The generation module is also used to take the ratio of the weight vector to the per-unit voltage value of each distributed power node as the initial reference value of the reactive current, and use a first-order inertial element to perform first-order inertial filtering on the initial reference value to match the adjustment response speed of the inverter, so as to obtain a reference reference value. The control module is used to obtain a reference voltage vector by performing error integration compensation on the difference between the reference reference value and the real-time feedback current at the resonant pole of the grid fundamental frequency using quasi-proportional resonance. The reference voltage vector is then converted into a pulse sequence signal through pulse width modulation mapping to control the inverter to perform state switching.

9. An electronic device, characterized in that, include: Memory, used to store computer programs; A processor is configured to execute the computer program to implement the steps of the power generation coordination control method for a new energy power system as described in any one of claims 1 to 7.

10. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores a computer program that, when executed by a processor, enables the generation coordination control method for a new energy power system as described in any one of claims 1 to 7.

Images