CN-121192831-B - Frequency supporting method for grid-structured converter multi-machine parallel system based on improved particle swarm

CN121192831BCN 121192831 BCN121192831 BCN 121192831BCN-121192831-B

Abstract

The invention discloses a frequency support method of a grid-structured converter multi-machine parallel system based on an improved particle swarm, which comprises the steps of maximizing a damping ratio of a system dominant oscillation mode to obtain an optimization target, solving an optimal virtual inertia initial value and an optimal damping coefficient initial value of each VSG unit by adopting an improved particle swarm optimization algorithm, adjusting balance global exploration and local convergence by nonlinear decremental of inertia weights, strengthening individual cognition at the initial stage of iteration and strengthening group cognition at the later stage by dynamic switching of learning factors, keeping population diversity by a mixing mechanism of Gaussian variation and Kexi variation, and monitoring the running state of the system to obtain the angular frequency variation and the angular frequency variation rate of the system when the system is disturbed, and cooperatively adjusting virtual inertia and damping coefficients of each VSG unit to inhibit frequency overshoot and accelerate oscillation convergence. The method solves the problems of serious frequency overshoot, slow oscillation convergence and stability caused by multi-machine interaction.

Inventors

LIU HUIQIANG
MENG QINGTIAN
GUO QI
MU TENG
Xing huadong

Assignees

内蒙古电力(集团)有限责任公司内蒙古电力科学研究院分公司

Dates

Publication Date: 20260505
Application Date: 20251125

Claims (8)

1. The utility model provides a grid-structured converter multimachine parallel system frequency support method based on improvement particle swarm, which is characterized by comprising the following steps: The optimal virtual inertia initial value and the optimal damping coefficient initial value of each VSG unit are solved by adopting an improved particle swarm optimization algorithm, wherein the improved particle swarm optimization algorithm adjusts the balance global exploration and local convergence through nonlinear decremental of inertia weights, strengthens individual cognition at the initial stage of iteration and group cognition at the later stage through dynamic switching of learning factors, and maintains the diversity of the groups through a mixed mechanism of Gaussian variation and cauchy variation; When the system is disturbed, monitoring the running state of the system to obtain the angular frequency variation delta omega and the angular frequency variation rate domega/dt of the system, and cooperatively adjusting the virtual inertia and damping coefficient of each VSG unit according to the dynamic states of delta omega and domega/dt so as to inhibit frequency overshoot and accelerate oscillation convergence; The improved particle swarm optimization algorithm comprises the steps of adjusting inertia weight from large to small according to a nonlinear rule along with iteration times, balancing global exploring capability of multi-machine virtual inertia and damping coefficient at the initial stage of iteration and local fine convergence capability at the later stage, dynamically switching learning factors along with the iteration process, wherein the learning factors mainly strengthen cognition of particles to self historical optimal parameters at the initial stage of iteration and mainly strengthen learning of the particles to group global optimal parameters at the later stage; Before solving the optimal virtual inertia initial value and the optimal damping coefficient initial value, constructing a closed loop small signal model of the VSG multi-machine parallel system, namely selecting a dq reference coordinate axis of one VSG unit as a public synchronous coordinate system, unifying state variables under the coordinate systems of the other VSG units to the public synchronous coordinate system through coordinate transformation, introducing loads at a grid connection point to construct a coupling relation between multiple machines and a power grid, and forming a closed loop state space model comprising the state variables of the multiple machines VSG, the current components of the power grid and the angle difference between the power grid and the public synchronous coordinate system.
2. The improved particle swarm-based frequency support method for a grid-structured converter multi-machine parallel system, according to claim 1, characterized by comprising the following steps: The cooperative adjustment of the virtual inertia and damping coefficient of each VSG unit according to the dynamic states of delta omega and domega/dt specifically comprises the following steps of determining the adjustment direction according to the positive and negative combinations of delta omega and domega/dt: When delta omega and domega/dt are negative, virtual inertia is increased, and a damping coefficient is increased in the later adjustment stage; Δω is negative and dω/dt is positive, decreasing the virtual inertia and increasing the damping coefficient; Delta omega and domega/dt are positive and simultaneously increase virtual inertia and damping coefficient; Decreasing the virtual inertia and increasing the damping coefficient when Δω is positive and dω/dt is negative; and the adjustment of the virtual inertia and the damping coefficient are limited by combining a preset threshold value so as to avoid frequent parameter switching, wherein the virtual inertia is adjusted in intervals according to the absolute value of dω/dt and the product sign of Δω and dω/dt, and the damping coefficient is adjusted in intervals according to the absolute value of Δω.
3. The improved particle swarm-based frequency support method for the grid-connected converter multi-machine parallel system is characterized in that the range of values of virtual inertia and damping coefficients is determined according to the operating capacity, rated angular frequency and maximum angular frequency change rate of the grid-connected converter multi-machine parallel system, the combination of the virtual inertia and the damping coefficients is required to enable the system damping ratio to be in an ideal interval of 0.7-1.0, and the system damping ratio is calculated according to the characteristic value of a VSG multi-machine parallel system small signal model state matrix and is used for quantitatively evaluating the damping speed of system response oscillation.
4. The frequency supporting method of the grid-structured converter multi-machine parallel system based on the improved particle swarm is characterized in that the grid-structured converter multi-machine parallel system comprises a plurality of VSG converter units, a power grid access component and grid-connected point loads, and each VSG converter unit is connected to an external power grid through a power transmission line after being converged to the grid-connected point through the power grid access component.
5. A grid-tied converter multi-machine parallel system frequency support system based on improved particle swarm for implementing the method of claim 1, comprising: The VSG converter units are used for outputting electric energy to a power grid; the monitoring module is used for collecting and outputting the angular frequency variation delta omega and the angular frequency variation rate dω/dt of the system when the system is disturbed; the control module is respectively in communication connection with the VSG converter unit and the monitoring module, and is configured to: The damping ratio of a system dominant oscillation mode is maximized as an optimization target, an improved particle swarm optimization algorithm is adopted to solve the optimal virtual inertia initial value and the optimal damping coefficient initial value of each VSG unit, the improved particle swarm optimization algorithm adjusts the balance global exploration and the local convergence through nonlinear decremental of inertia weights, enhances individual cognition at the initial stage of iteration and enhances group cognition at the later stage through dynamic switching of learning factors, and population diversity is maintained through a mixing mechanism of Gaussian variation and cauchy variation; and receiving the delta omega and the domega/dt output by the monitoring module, and sending control signals to each VSG unit according to the dynamic states of the delta omega and the domega/dt so as to cooperatively adjust the virtual inertia and damping coefficient of each VSG unit, inhibit frequency overshoot and accelerate oscillation convergence.
6. The grid-connected converter multi-machine parallel system frequency support system based on the improved particle swarm, according to claim 5, is characterized in that the control module is used for solving an optimal virtual inertia initial value and an optimal damping coefficient initial value and cooperatively adjusting virtual inertia and a damping coefficient under disturbance, the system further comprises a grid access component and a grid-connected point load, the grid access component is used for realizing electric energy transmission of a VSG converter unit and the grid-connected point, the grid-connected point load is used for constructing a coupling relation between a plurality of VSGs and a grid, and the auxiliary control module is used for realizing closed-loop control of the system.
7. A computer device comprising a processor, a memory and a computer program stored on the memory and operable on the processor, the computer program when executed by the processor implementing the method of any of claims 1-4, comprising maximizing a damping ratio of a dominant oscillation mode of the system as an optimization objective, solving an optimal virtual inertia initial value and an optimal damping coefficient initial value for each VSG unit using an improved particle swarm optimization algorithm, and adjusting the virtual inertia and damping coefficients in cooperation according to the monitored Δω and dω/dt when a disturbance occurs to the system.
8. A non-transitory computer readable storage medium, wherein the storage medium has stored thereon a computer program which, when executed by a processor, implements the method of any of claims 1-4, comprising improving parameter adaptation of a particle swarm optimization algorithm, hybrid variation mechanism execution, and dynamic co-adjustment of virtual inertia and damping coefficients under disturbance.

Description

Frequency supporting method for grid-structured converter multi-machine parallel system based on improved particle swarm Technical Field The invention relates to the technical field of operation and control of power systems, in particular to a frequency supporting method of a grid-structured converter multi-machine parallel system based on improved particle swarms. Background The modern power system is accelerating to evolve, and the characteristics of continuously improving the permeability of new energy and deepening the degree of power electronization are presented. However, the access of large numbers of power electronics to traditional grids poses new stability problems, particularly the inherent low inertia, low damping characteristics of current transformers, which make the frequency and voltage support capabilities of the grid challenging. Because of its ability to independently support system frequency and voltage, the network-built control has been extensively studied by students at home and abroad, and among them, virtual synchronous generator (Virtual Synchronous Generator, VSG) control in network-built converter control has received extensive attention. The method has the core concept that a model of the synchronous generator is introduced into a control algorithm of the inverter, so that the inverter can simulate the external characteristics of rotor inertia, damping, primary frequency modulation, reactive voltage regulation and the like of the synchronous generator, and necessary inertia and damping support are provided for a power grid in an 'active' mode. The distributed power supply can participate in frequency modulation and voltage regulation of the power grid like a traditional synchronous generator, and becomes an effective solution for realizing friendly access of new energy and improving the stability of the power grid. However, limited by the physical characteristics of the power electronics itself, the immunity and overload tolerance levels of VSG systems remain a significant gap from conventional synchronous generators. At present, students at home and abroad have completely studied the small disturbance stability of the single grid-connected system of the grid-structured converter, and have studied to establish a state space model of the single grid-connected system of the grid-structured converter, which comprises a power ring, a voltage ring, a delay link and a circuit part. The harmonic linearization method is adopted in the literature to establish a sequence impedance model of the grid-structured converter, and the difference of sequence impedance characteristics of the grid-structured converter and the following-grid converter is compared and analyzed. There are also documents that use the impedance ratio matrix formed by the grid impedance and the converter impedance to obtain a stability criterion and indicate that the grid-built converter may cause low-frequency oscillation of the system when accessing a strong grid. There are also literature on transient stability analysis of grid-structured converters employing single-loop voltage amplitude control by combining voltage amplitude dynamics with power control loop dynamics using a phase diagram method. There are methods in which the frequency oscillation process during a fault is analyzed using the equal area rule. In summary, many researches are performed on the controller design and modeling, parameter adaptive optimization, and the like of the VSG stand-alone system at present. When a plurality of grid-structured converters are connected to the same grid-connected point through connecting impedance, complex interaction can be generated between the grid-structured converters and the multi-machine parallel system, and the multi-machine interaction can easily cause new stability problems. The stability analysis of the multi-machine system surrounding the grid-structured inverter is studied primarily, and the stability of the frequency small signal of the multi-machine parallel system of the grid-structured inverter is studied in literature. The influence of the virtual inertia and the line impedance matching relation on the system stability is analyzed in literature, and an adaptive compensation method is provided, but the dynamic performance optimization effect is poor. There is a literature that proposes an adaptive virtual impedance control strategy suitable for multiple VSGs in parallel, to achieve the effects of dynamic adjustment of virtual impedance and output reactive power sharing and reduction of circulating current, but the dynamic characteristics are not considered as well. The VSG double-machine parallel small signal model is deduced according to the VSG double-machine parallel system, the response rule of parameters to the system is analyzed, and an inertia matching control strategy and a parameter self-optimizing control strategy are respectively provided to improve the frequency stability performance of the